I have failed to find the parasitic draw on an RV (slower than this one, it takes a couple weeks to kill the battery). I finally got a "Top Post Battery Master Disconnect Switch", which is a little dinglebob that goes between the battery post and cable and lets you easily break the circuit. I highly recommend the $12 solution for an intermittently used vehicle.
You can also grab a trickle-charger and use that when it's at rest, which can (on modern vehicles) prevent the computer from going nuts.
If you're fancy, you can even make it a quick-disconnect setup so that if you drive away with it still on, it won't break anything (or do the poor-man's version where you have the wire connected to the wheel chocks).
I thought about this but the problem with newer cars (mine at least) is killing the power wipes the memory so it won't pass an emissions check (or something like that). After my battery died I had to find an excuse to drive an extra 100 miles so I could pass emissions.
On my car the emissions laws definitely caused more emissions than they saved that day.
I hear this all the time, but I have never understood it. If I clear my drive cycle monitors, they are all passing again just a few miles later. Do some vehicle manufacturers just put insane constraints on the drive cycle monitors?
There is a drive cycle set of requirements and it’s usually possible to clear them in under 20 miles. Trying to figure which part of the cycle isn’t accomplished yet is the challenge. I’ve chased it down on my wife’s Honda when we were trying to get it to go “Ready” the same afternoon. Usually just not doing anything unusual for a day or so of normal use is enough, but you can often query the OBD2 system to get the portion of the cycle that’s not yet complete. Just doing a long highway drive isn’t going to do more than a moderate highway drive already did.
> Do some vehicle manufacturers just put insane constraints on the drive cycle monitors?
Yes. Combine with the fact some only pass after certain conditions (eg. over 3000 revs for 10 mins, or idle for 5 mins, or only after 3 starts, only when the outdoor temperature is within this range, etc). Sometimes those conditions aren't stated in the manuals.
That makes the checks passing seem more like a random process and driving more will usually make them pass, eventually.
And drivers without a code reader won't know when it has passed, so they are usually instructed by mechanics to drive a long way to be sure before taking it in for a test (don't want to have to reschedule the test just because you didn't drive it enough).
I recently noted in my OBD2 scanner that some vehicles support the measure of miles since check engine lights were cleared or something to that extent. It made me ponder at some point in the future there is going to be regulation mandating that to pass an emissions test you must pull up to the test station with a minimal threshold mileage on the car to vet no check engine lights have tripped in that interval.
I installed a battery disconnect switch on the dashboard of my jeep. There's a spot that seems made for it. It was a $27 solution, but I don't have to pop the hood. And it solved my charge parasite problem instantly.
I'm reminded how often people come to the RV forum trying to figure out parasitic draw. The answer almost always ends up being the propane detector. People forget it's even there.
Less of an issue on newer RVs that come with at least one basic solar panel installed from the factory.
I'm surprised it didn't already have a master shut off switch next to the battery. You should always shut off / disconnect the battery when not in use because, as someone else pointed out, the propane detector will drain it.
Well this is timely. I have a parasitic draw on my NC Miata. The first mechanic just replaced the battery. The second told me it's normal for the battery to die if you don't drive it for 3 days (because "cars have computers now"). It seems like mechanics really don't like digging into the electrical part of the car... but these days that encompasses more and more of what can go wrong.
Without time to dig into it myself I've just been parking it with a battery tender every time I come home.
> it's normal for the battery to die if you don't drive it for 3 days
I've heard this nonsense a couple times now. I was incredulous at first but everyone seems to say this. What are anyone's suggestions for a weekend-only car? (I bike to work so I don't need a car on weekdays.) Trickle charge the car battery on weekdays?
Most new-ish cars also store the dinamic engine calibration data in ECU RAM so if you disconnect the battery you can find your engine running harsher next time. So are the values of the headlights auto leveling system and other stuff.
You're also risking to reset the radio/media unit security code (on cars that aren't too new but also not super old), which if you got the car second hand and the former owner didn't save the piece of paper it came with and give it to you, tough luck unlocking it.
Thinking more, given how common parasitic drains are, carmakers should have all non-essential fuses in a separate subpanel section that can be easily switched off/diagnosed.
There are systems that can be attached to save the parameters from the car before you disconnect the battery or even to provide electricity to the car while the battery is being checked in another room. Not common to find and most people don't need them, but they exist.
It's nonsense! If a car is dying after not being driven for several months, ok fair enough. If after several days a particular year make model dies it deserves a recall!
I have this problem, but it really only becomes an issue when a few days becomes a few weeks. Do that a few times and the battery will lose capacity too.
Weeks, sure. 2 weeks is probably on the lower end of acceptable though. My car is often left for 2 weeks and is completely fine, but most of the time it's weekly.
Yes, they make battery tenders that are specifically for this. They'll trickle charge when needed, but also not charge at all when the battery is at the right voltage.
You can drive a car from the 90s before they started shipping alarms and keyfobs in every vehicle. My 1995 Miata draws just enough power to maintain the clock in the radio, and it starts just fine after sitting for 6 months (it's not a winter car).
My 2000 Ford Explorer has all kinds of electronics and modules from the factory. And even my crap I added in... I leave my Qi charger, ham radio, Bluetooth FM transmitter all running and plugged in, and I can come out a week later and the battery is still not dead. And when I'm camping I run a 12v van all night long. Truck starts right up every time. I don't get what would be draining more power than that in a more modern car. You'd think they'd be more efficient at not wasting juice.
There's battery kill switches you can install onto the battery pole to effectively disconnect everything from the battery. This should work for older cars. For newer ones you may experience hiccups like reverse camera not working for a few hours until it has re-paired itself with the head unit.
Also the engine and transmission computers may go into "learn" mode if they are cold-starting having had no power for a while, and so you may have stuff like a wandering idle speed for a little while.
This really wouldn't work for any car made in the last 20 years. Most of them have anti-theft systems that require you to put in a PIN to use your stereo if it ever gets disconnected from the battery. I suppose you could type that in every single time, but there are other computers like your automatic transmission management that learn how you drive, etc, and might not like to get reset every day.
This is false. I have owned many different cars from the last 20 years and none of them did this when you disconnected the battery. A quick google search shows that it's Hondas that do this, not "any car made in the last 20 years".
Really nonsense. It's true it might be more common due to all the additional modules and computers cars have now. If one goes crazy, it might drain your battery. But this is only in case of a failure. Edit: cold weather can accelerate that.
If it's 'weekend only', battery should be no problem. If sometimes you leave it a couple months unused, that might be a problem.
Depending on how ready you would like to be when you're going to use it, or the access you have to your car (is it far away or in your garage?) you could add a connector for a battery charger/maintainer (some of them have accessories for plugs you can leave permanently, see the NOCO GC002 for an example), and/or add a battery switch to easily disconnect the battery.
Warning: your radio might ask for a security code if it loses battery connectivity. Be sure you have it.
I don't have extensive experience with my current car bought last summer but 2 weeks in warmer weather definitely wasn't an issue.
I agree colder weather is always a bigger issue and at 1-2 months I'd definitely be thinking of a trickle charger if at all possible. An older car I didn't drive in the winter would definitely end up with a dead battery if I didn't keep it on a charger.
AGM batteries are a lot more durable than regular lead-acid. They can cost a bit more but for me the longer life, more resistances to issues if you do run them all the way down and their sealed, maintenance free nature makes them the first thing I swap out. Optima used to be the only commonly available batteries but Costco, for example, has AGM equivalents for just about every battery out there - you may have to order it and wait a few weeks.
They were also the original batteries used in Miata's since they give off a lot less hydrogen when charging. Since the Miata battery is in the trunk, and all lead acid batteries give off hydrogen when charging, and hydrogen in an enclosed space can also be rightfully called a bomb, having less of that is a good thing!
I must have had a bad run of big H8 AGMs in my Jag and Volvo. I think I've replaced about one a year on average. The Jag is very subject to parasitic drain and sometimes it just didn't come back from running itself out.
Did the Miata not come with a hole in the trunk floor for an exhaust hose for the battery? Both of my cars did...
I didn't drive my 2019 Mercedes for months during the pandemic and the battery was fine. It did give me a warning when I restarted driving it (I guess the battery was starting to get low-ish), but it immediately fixed itself.
You should see if there are carshare options around you.
Even if you’re spending $50/weekend (which is a fair amount of driving — Communauto is as low as $3/hr), that’s still quite a bit cheaper than owning a depreciating asset with lifetime maintenance costs (not to even mention insurance and fuel).
Oh do I wish there were carshare options near me. Just moved from Philadelphia where ZipCar seems to be the only remaining contender, and you're lucky if there's one within a half hour's walk in the part of the city I was in. It's even worse in the town I moved to. There used to be a few at the nearby train station, but those seem to have disappeared. (This reminds me, I should go cancel my membership.)
I'm jealous of the carshare options that seem to be available elsewhere. If it were as easy here as in this NJB video, it'd be a no-brainer: https://youtu.be/OObwqreAJ48
This must be the new astroturfing attack against concepts like public transportation. I've seen this phrase pop up too frequently at this point for it to be organic.
The simple may see it as astroturfing, others see it as an observation by those paying attention. I'm now living in a house I own without a mortgage because 20 years ago I bought a house and didn't rent. Utilities and property taxes are my only expenses. Yes, there is maintenance but if you don't think that's also factored into rent you are beyond delusional. Landlords aren't going to lose money out of the goodness of their hearts. I haven't had a car payment in over a decade because I take care of my stuff. The car goes in the garage; I don't fill my garage with a bunch of useless stuff I will likely never touch again (or pay someone else to store it!).
Common sense stuff like the above, sadly, isn't common sense any more - so hence pithy phrases like own nothing and like it. A modern day emperor has no clothes, if you will.
> In 2016, Auken published an essay originally titled "Welcome to 2030. I own nothing, have no privacy, and life has never been better",[2] later retitled "Here's how life could change in my city by the year 2030". It described life in an unnamed city in which the narrator does not own a car, a house, any appliances, and any clothes, and instead relied on shared services for all of their daily needs.
From your link. I'll let others decide whether or not that is "lying".
I did that for the exact same problem on the same model car earlier this year, but gave up after trying a few places that all said the same thing (including the dealer). Glad to see this post though… now I know I’m not the crazy one.
I would expect a _good_ car (no maintenance issues, new gas, newer battery) to go at least a couple months without being driven and not have any issues upon starting it up.
The average car in average condition should go around a month without being driven. Anything less than that either means a battery nearing the end of its life, not getting charged, or a parasitic draw somewhere.
We have a 10 year old Honda Odyssey that we keep just for long road trips and it has gone 3-4 months without being driven with no issues. I usually do try to fire it up every 1-2 months just to make sure the battery gets a charge and to avoid long term storage issues around oil and belts and other parts that degrade.
Thanks, that's good advice. Luckily we've burned a tank or so every road trip (every few months) but I was wondering about that and whether it was good for gas to sit so long.
I park up my hilux for a year at a time, sometimes, and it Just Starts. Sure, it’ll take three attempts with a cold, damp engine and a sluggish battery, but it has yet to let me down.
Had one of those, too, and used to hide it in a pint of Guinness for a laugh. Dropped it in the sea a few times. Accidentally melted it slightly in a small housefire. It lived on after I moved on to a t68i as an SMS gateway for my home server for my one-man proto-Twitter.
My wife drives about once a week, probably less on average. She has a 2013 Honda CR-V, only 15,000 miles on it. We live in Minnesota, so have cold temps, but the car in a garage. The car always starts. We've replaced the battery once just due to longevity.
15k miles on a 2013 CR-V? Enjoy the last car you'll ever own! I assume you know this, but some of the maintenance items for that car are 8 years I think; worth checking the manual to find out!
Haha, yep. She refers to it as the literal "just driven to the grocery store by a little old lady" car (because it's basically that and Home Depot). She even got the lowest trim level (on purpose) so it doesn't have anything fancy (no BT radio, no keyless start, no big display, etc.).
Good point on the maintenance. I do get the oil changed, again more by time than distance, and I'll look up the other stuff now that you've mentioned it, thanks.
I try to keep my cars on trickle chargers (Noco Genius 5 in my case) when not driven regularly, because regularly killing $300 AGM batteries gets expensive.
That said, I'd expect most cars should be able to manage, say, 2-3 months. It's when I exceed 4 months with some regularity that battery life starts getting sketchy.
If you're driving the car every weekend, I'd expect that should be plenty. Note that fancier cars with more electronic features don't do well when only driven short distances. If you only drive 2-3 miles at a time, you might be discharging the battery faster than you're charging it.
That said, at between like 0F and -10F, I'd be intentional about starting that car every couple days. Below -10F to -20F, I'd start it each day. Below -20, I'd consider bringing the battery inside for the night - even if you have a battery blanket. Once a lead-acid car battery has frozen, it will die very quickly from then on - I'd believe within a few days or less even in normal weather. AGM batteries are awesome alternatives for the cold weather lifestyle, in terms of their resistance to freezing damage.
I have a week-end only car (2002 Mini) and I've had no issue letting sit for a week or two. Including opening/closing it without starting the engine to retrieve stuff in that timeframe.
And a week or two is on the low-end of how much you can let it sit. My SO car (1st gen citroen C3) often don't run for one or two month at a time and there is absolutely no issue with the battery.
On the other hand, if you have parasitic drain and access to an electric socket, there are battery chargers that can keep the car topped up. Or without an electric socket, just unplug the battery when the car isn't used (though that may disable some things, for example in my car the hatch is inoperable without power).
It's fairly straightforward if the car is unmodified, as everything should be fused. Multi-meter between positive terminal and positive lead and pulling each fuse to see when the drop goes away will isolate the circuit. Then you've narrowed it down pretty far and can go from there.
The horrible thing is if the car is modified, and you discover an un-fused circuit.
I’ve had unknown draws before and many times they’re not constant draws. A constant draw is easy and quick to find. A draw that only happens when some module or other wakes up is a lot harder to find.
As an example, my 2017 Corolla runs the fuel pump for a while 5 hours after shutdown to test fuel system pressurization. I didn’t even know until I recently started parking it inside and I finally heard it running.
Yeah that's pretty normal across different manufacturers, and you only hear it if you park in a quiet garage and happen to be in the garage at the moment it turns on to do it's thing. It's testing the seal of the fuel cap to make sure you aren't just venting gasoline fumes.
I learned about this with my 2013 Scion iQ. I was living with my brother who didn't allow weed smoking in or near the house so I would go hotbox my car.
Weirded me out the first time I'm sitting there in silence smoking a bowl and suddenly fuel pump noises. Which were easier than normal to hear because I took the back seats (the only sound deadening in the back) out of that car so I'd have room for a full size spare tire instead.
With my current car (2020 Kia Sportage) I've noticed the head unit is waking up before I even start the car. As soon as you unlock the car I can see my flash drive light up with activity as the head unit gets ready to pickup music from where it left off.
I had to sleep in a work truck once and every hour or so it would modulate the shutters in the air vents to refresh the air in the cab or something. The click and buzz whir kept waking me up, it was quite annoying.
I recently watched a YouTube video from someone who'd tracked a nasty intermittent parasitic draw [0]. The problem only happened after turning the ignition on & back off (so the standard test of disconnecting the battery & reconnecting via the ammeter wouldn't show it), and the draw cycled between 3.6A and 0.3A.
Also, after turning the ignition on & back off there were a lot of (normal) transient draws (from things like the dome light that don't turn off immediately). Even with the problem circuit disconnected, it drew around 6A (!) immediately after the ignition was switched off, dropped to 0.4A after about a minute, and sat at that level for another 9 minutes before dropping again to 0.06A. That means if he hadn't waited ~10 minutes per test, he'd have been chasing draws that were actually normal.
Combine an intermittent fault with intermittent normal behavior, and you've got a troubleshooting nightmare.
Yes. I had a Dodge Charger where the parasitic draw wouldn't really become problematic until 3 days after you re-attached the battery. At which point it would be so bad it would discharge from full to empty overnight. It could even out-draw a trickle charger.
I took it to several mechanics who couldn't find anything - because they'd unplug the battery to add a meter into the circuit. They weren't willing to wait 3 days afterwards to come back to it.
I ended up disconnecting the battery overnight. Worked fine then. Traded it in; I feel a bit bad for the next buyer.
Ye olde ammeters were secretly voltmeters with a low-value but high-precision shunt resistor in series with the circuit to be measured.
Current passing through the resistor causes a tiny voltage drop, which is measured by the meter. As the voltage is proportional to the current, the scale painted on the cardboard behind the needle did the actual conversion.
In the author's setup the fuse takes the shunt's place; its resistance is apparently a known value that can be gathered from a datasheet.
The absolute current value probably doesn't matter - in fact I very much doubt a hobbyist-level multimeter is even capable of accurately measuring the (fractions of?) millivolts across a fuse.
It's more important as a boolean signal - "is there current being drawn on this fuse?", and then usually even those fractions of millivolts will generally be enough to make your el-cheapo multimeter register 1mV and tell you something is drawing current, where as a fuse with no voltage drop across it at all will reliably show 0mV (just like it would if the probes were shorted) on even a cheap meter.
Same idea really, just that you have to bother disconnecting something and try one by one vs just measuring it fuse by fuse.
Other method is getting DC clamp meter with low range (say 2A). Then just clamp in on wires. Bit fiddly as they work on magnetic field so they need to be reset before measure (And can work as compass in a pinch...)
I know what you're trying to say... but this reads like very dangerous advice and you may want to revise.
Ammeter between a battery's terminal and the car's electrical system.
Also note that you'll probably create sparking when doing this, and if the battery has recently been charging this can be dangerous (hydrogen outgassing).
I think the mistake may have actually been writing ammeter instead of voltmeter. Ammeter is not safe as general advice as a typical one won’t be anywhere near large enough capacity for working with a car battery and should come with a lot of disclaimers about how to hook it up. Measuring voltage drop would work but only for fairly high loads.
> Yah, that doesn't really make sense for small parasitic loads.
If your multimeter has a sub-mV range -- which a decent meter should -- you can measure a voltage drop across the battery ground lead. I've successfully done that to diagnose a battery draw in my car.
I'm not talking about measuring the terminal voltage of the battery. What I mean is using the ground lead as a current shunt by connecting one meter lead to the negative battery terminal and the other to where the ground lead connects to the vehicle chassis. The voltage measured across that shunt is proportional to the amount of current being drawn from the battery.
It's the same principle as measuring the voltage across a fuse, except it ends up measuring the draw for the whole car, not just one circuit.
No. Not if the meter is in "Ammeter" mode (or is purely an ammeter only). An ammeter presents a very low resistance to the circuit you are attaching it to, low enough that for almost all circuits, it is effectively a "short".
If one connects an ammeter across a low resistance voltage source (i.e., across the car battery) the ammeter will appear as a short circuit to the battery, and one or more bad things will happen. The 'least bad' will be blowing a rather expensive fuse in the ammeter that protects it from this kind of accidental use. Several of the "most bad" will involve hot molten metal and/or extreme heat.
> If one connects an ammeter across a low resistance voltage source (i.e., across the car battery) the ammeter will appear as a short circuit to the battery
Yes. My first comment on this thread says not to do that, and suggesting an alternative is "Ammeter between a battery's terminal and the car's electrical system." instead. https://news.ycombinator.com/item?id=35514652
I was replying to "anywhere near large enough capacity" in the next post. The typical 10A range is massive overkill for typical parasitic loads in a car, and even things like the fuel pump deciding to kick on and run in most cars.
Using an ammeter as you describe would work fine, but the topics this thread including mine should come with a footnote[1] at least. Especially when this thread shows a simple mistake can cause a lot of confusion about what to do. My thought process with my comment about capacity was that muscle memory makes it likely you'll eventually start the car if you're touching the ignition switch/button during troubleshooting, although on further thought and reading the other comments I'd do the following: pull the starter fuse, then use a 10A ammeter function of a common multimeter, connected inline between the negative harness and the negative terminal and not on the positive side.[1]
[1]This is just what I would do. Only do this based on a reputable guide if you don't have experience working around car batteries.
>It seems like mechanics really don't like digging into the electrical part of the car.
If you've ever seen a car's complete wiring harness you'll know why. It's not only because there's three gazillions of wires - they're also heavily insulated with tape that makes it a nightmare to access those wires in the first place.
> It seems like mechanics really don't like digging into the electrical part of the car...
Tell me about it. I'm in a similar boat. My audi a4 has had a parasitic draw for the past 10 years. My mechanic didn't want to take the time to address it because that sort of thing takes a serious amount of time and he's always backed up. I got out my fluke and narrowed the cause to what I believe is the comfort control unit. I can't replace it myself because it's a coded part. My mechanic won't do it because he won't confirm that's the cause.
None of the other mechanics in my area have the equipment. Even the Audi dealership won't touch it because they're backed up too. So, I keep a battery jumper in the glove box and hook the battery up to a tender when I can (outdoor parking).
I don't know what I'm going to do for my next car. Pretty much every car made after 2010 has coded parts and are more computer than mechanical. It does make me tempted to learn the electronics part and open a garage that focuses specifically on fixing electronic issues in modern cars.
I hear people say that about my BMW, and realistically the hassle of some reliability issues is nothing compared to driving the lifeless husks of cars they recommend instead.
(there are reliable sports cars, but it significantly limits the pool you can select from)
That's fair, but it's funny because that's how I always felt about BMWs. My buddy loves them but I always felt like they were so... sterile. Unless you are at a track or love cutting off people in traffic it feels really hard to have any fun in one.
Sterile compared to what? Is there another manufacturer of acceptably-tuned rear-drive sedan chassis out there at a semi-reasonable price point that I should know about?
(I'm talking about pre-2010 BMW here; I would totally agree with most new ones being sterile)
As a current owner of a 90s Jaguar and an 05 Volvo, I've been looking at low-mileage E90 era BMWs for my next daily driver. Not too scared of moderately expensive maintenance, but would like it to be more reliable than the Volvo has been.
As a former E46 and current E83 owner, my advice is go for the earliest chassis code you can find while keeping mileage low. You need to find a unicorn at this point. (Unless you like turbos - then you have to go E90 or newer.)
BMW really ramped up the electronic complication starting with my E83, and that shit wears out right alongside all the typical flimsy BMW plastics and rubbers. Plus the delightful parasitic drains from tfa.
They also ramped up their maintenance-unfriendly designs around the E46 era. On an E39, a rear spring swap is easy. Try it on a newer one, and you will regret the attempt.
I agree it may be 90% as fun to drive (which is subjective), but you're comparing FWD to RWD.
I'd much rather have a 80's-90's Nissan or Subaru, especially built, than a modern BMW, though. Much more soul in vintage cars imo.
Some vintage BMWs were pretty reliable, a lot of modern Nissans have awful reliability.
It's AWD, but the MS6 has about 100% the repair bills of any old BMW, this person just has no idea what they're talking about. Maybe they owned one when they were new and doesn't realize what long term reliability is like
Agreed - the last 5 that felt good to drive was the e39, and the rest probably peaked with the e90 or e8x. I felt the same way with the corvette - my c5 was the most fun to drive. C6 was an upgrade in comfort but significantly number, and the C7 was horrific. I took a C7Z for a spin and it felt slower and more boring than my c5 despite having 2x the HP.
Agreed. I beat my C5 and C6 pretty hard, but aside from (some) of the GM/F cars you’re very limited in choice. Especially if you’re looking for a lightweight sedan.
Anecdote. I've had a BMW for the past 9 years now, and not once have I had to open the hood of the car or so much as replace my wiper fluid. I don't even know where the lever for any of that is. I pretty much take it in to a dealership when a light on the dashboard lights up, they drive me to work, pick me up after and my car is good to go.
After my Audi I learned I'm never buying another German car. Hard to work on myself, absurdly expensive for the shops to work on them, parts are very expensive. Since most owners seem to be "of a demographic" that does not mod/repair their vehicles, the online community pales compared to some other brands where this is a bigger part of the ownership culture, IME. Yes, I know there are passionate Audi modders and repair forums out there, but I'm referring to a percentage of the ownership base and in comparison to communities around other makes.
The Audi was great fun to drive though, no denying that. Just different priorities in what I want out of a car.
The problem with specializing in these issues is that most customers will be dissatisfied when you charge them thousands of dollars to reset a minor module on their car.
How many hours of shop time are you willing to pay for with no guarantee of a solution, and very few (if any) parts replaced?
Dude's been tending his issue for a decade, I bet he'd have at least those like him as a customer. In a metro he might have a pretty lucrative business. On the other hand in the metro there is probably someone who already does this work, but they are very hard to find!
Ford lied to its own dealership when my coworker went to them about a transmission slip with his 2012 Ford Fiesta.
He had the people at the dealership drive the car. They acknowledged the problem and got on a call together with Ford. Ford refused to even entertain the idea there was a flaw but less than a year later they came up with a "software patch".
Pete would bring it up with anyone he met. Do not buy a Ford Focus / Ford Fiesta with automatic transmission.
> How many hours of shop time are you willing to pay for with no guarantee of a solution, and very few (if any) parts replaced?
Imagine opening a computer software support business where you bring in your old proprietary binary programs that you got from who knows where, and the business is to fix bugs you've found.
Maybe some really big enterprises would pay for that sort of thing.
But I doubt it would be feasible for consumer stuff.
It would be a crazy/fun kind of business to run. I'm sure it probably exists.
I've toyed with that kind of thought. The technical part seems fun, finding customers (especially the first customers) seems difficult :p I'd really like to work at such a company though.
There are companies around that offer VIN re-coding by mail order if you need to swap one of the computers, but you'd have to be in a fairly big metro area to profitably run a company just doing electronic trouble shooting. It's easier just to "fire the parts cannon" after the usual suspects (dead sensors, shorted wiring) have been accounted for.
Having had a 2006 model car in my workshop forever rebuilding the engine, I have though it might be worthwhile starting a business doing that kind of trouble shooting but nobody will want to pay, because the vehicles involved are inevitably old and worthless. It's fun to do it for yourself, but as a business your typical customer who needs you won't be able to justify paying you.
My other thought was to possibly set up the tools and whatnot to support local businesses in doing that kind of troubleshooting on older cars but building the knowledge and tools to distribute would be a very expensive exercise.
I love this phrase, and it reminds me that the behavior exists in the computer world too. From the Jargon File:
field circus: [a derogatory pun on `field service'] n. The field service organization of any hardware manufacturer, but especially DEC. There is an entire genre of jokes about DEC field circus engineers:
Q: How can you recognize a DEC field circus engineer with a flat tire?
A: He's changing one tire at a time to see which one is flat.
Q: How can you recognize a DEC field circus engineer who is out of gas?
A: He's changing one tire at a time to see which one is flat.
Actually had this problem in PC repair. A lot of times I'd spend hours/days diagnosing an issue just so I could understand it inside and out and have a quick solution for the future. I could almost never bill customers for it since it would've cost hundreds of Euros (easily going into 1000+) for what ended up being a fairly simple fix (the $10.000 to hit a pipe with a hammer phenomenon). I'm just lucky I had enough slack to be able to do in-depth diagnoses, because it saved me fuckloads of time later on.
Then there were the cases with arbitrary random bluescreens that were impossible to diagnose where I'd upfront tell them I could only do the bare minimum because otherwise they might have to pay 500 Euros for a solution that probably wouldn't work.
Pulling fuses is a pretty reliable way to find the relevant fuse in the first place. If the battery doesn't run down while a particular group of fuses is pulled, the one you want to pull is in that group. Start with half the fuses, then narrow down.
You might have to reprogram your radio channels...
I can recommend searching for "South Main Auto parasitic draw" on YouTube. The guy is a genius at electrical troubleshooting.
His strategy isn't to pull fuses, it's to set the car sit for at least 30 minutes or so (with key off) and then check each fuse with a multimeter to see which has current on it, and then check everything on that circuit. Arm yourself with schematics and wiring diagrams, otherwise it'll end up being something of a wild goose chase.
(The idea being if you start pulling fuses, you can "reset" various computers in the car, which may show up as a false positive. It can take up to 30 minutes or so for all the various computers in a car to all go to sleep, although it's usually only a few minutes for most cars.)
It would take someone more skilled than I, or with better tools, to check current in a fuse box without removing fuses. Maybe I should watch that guy and get better...
You can check continuity across the contact points, but measuring a 75mA current via the voltage drop across a 0.0034ohm resistance (give or take) means reading 0.2mV, which is… borderline for a consumer multimeter.
I wonder if one couldn't identify the time it takes for the battery to get to some low point (50%?) and then determine the order of magnitude draw they would be looking for. 75mA would take a long time (I think) to drain the battery to where starter can't turn over (like a week?).
If it were me, I would pull the fuse, then jumper the block over to a different fuse on a circuit that is off when the ignition is off. Then you never have to think about it again.
I wired in an aftermarket DVD player in an old car and tapped into a fuse that was supposed to not draw power when off. Of course, it did draw power so the DVD player never shut off. It wasn't too much of a draw but I didn't want to end up with a dead battery situation because of it so my solution was to just plug the fuse tap in when the kids were in the car and wanted to watch something.
I'm sure there was a more elegant solution (find a different fuse, rig up a button, etc.) but this was the path of least resistance and wasn't really annoying to do.
Aftermarket head units should have two power inputs - one for standby, and one that only gets power when "accessory power" is on. Every car I've owned I've been able to wire it up so that works, then it only turns fully on when the car is on.
I avoid very new cars like the plague though, so not sure how different they are.
It makes me anxious thinking about all the bullshit in new cars - touch screens, coded parts, 4G connections with GPS tracking, etc etc.
This was an rear-seat overhead DVD player, not a console one. Although, I probably would have been better off wiring to the head unit instead. At the time, I still had the factory one in there so it's not quite as expandable and more complicated to add stuff to.
Yeah, I had an old car where the useless tape deck developed a short (I think) that would drain the battery after about a day. I didn't know it was the stereo until I played around with removing fuses one by one though.
At least! When my engine finally needed a rebuild, so did the transmission, steering box, CV joints, and center differential. And you're right, it needed new bushings, shocks, springs, and balljoints.
Yeah, I recently bought a car (I had driving license for 15 years before that but for 14 of them I didn't drive coz I moved to the big city), and basically bought one of cheapest, had to replace dampers last year, this year one of springs broke, one of calipers have some problems and apparently one of previous owners installed lambda probe emulator to hide dead catalytic converter...
Tho a bunch of that might be because I bought "sporty car" and those rarely have easy life in the first place...
I drive a 1998 Jeep Grand Cherokee. I get asked if I'm interested in selling it at least once or twice a month, nowadays. The sightlines are amazing compared to modern cars, its AC is OP, and the AMC 4 liter straight six is one of the most durable car engines ever made. Mileage is not great, but it's paid for and dirt cheap to insure, that buys a lot of gas right there.
Inline 6 has to be one of the best engines. I got 3 cars with them in 4L, gas injected, gas carbureted and diesel. They go so well.
The only computer in the diesel one is the glow plug timer and there is no computer in the carbureted one (both 1991 Nissan Patrols).
The injected one is a 2005 ford falcon and is starting to get all sorts of interior electrical niggles. Only wiring issues in the patrols is a left headlight short which was easily routed around using relays.
Nice, I had a 97 Jeep Grand Cherokee and it was a great vehicle. I did have a problem with the doors falling off, which I think was a defect at the time.
> My mechanic won't do it because he won't confirm that's the cause.
My experience is that a lot of independent mechanics will do what you ask them to do, so long as they think it’s not unsafe and have confidence that you’ll pay the bill without complaint when the task is done whether or not the problem is fixed.
Probably not a good “this is my first time meeting you, but please replace my flux capacitor” but if you’ve got a history with the shop, I’m surprised you couldn’t talk him into it. (I worked in a shop briefly in college. We’d do what the customer wanted, including installing parts they bought, but the only warranty was on a “we spent an hour; you paid for an hour; thus ends the transaction” basis.)
Yes. I had a problem with a leaking transmission once and the shop refused to accept my diagnosis because "in all my years servicing transmissions, I never heard of that failing."
Finally, the service manager agreed to check it out with the understanding that if I was wrong, I still had to pay for the time they spent investigating.
This was me with a Suzuki vstrom with a bad solenoid. Mechanic ran 12 volts through it on the bench and refused to accept it wasnt working in the bike.
Bought my own solenoid online and surprise the bike started working again.
Debugging spider sense is definitely an upside of spending years working with terrible code.
I once moved into a place that was great, except the shower would start out hot, quickly drop to warm, and then stay warm for as long as we cared to try it. I hated this, so every time I showered I'd spend some time trying to debug it, despite knowing nothing about plumbing.
Eventually I asked myself: why do normal hot water heaters work the other way, where they stay hot a long time and then get pretty cold? Clearly, there's some way of keeping the incoming cold water separate from the already-heated hot water, like by putting the output pipe at the top and the input pipe at the bottom. But if you reverse them, you might get what we had.
I leapt out of the shower and felt the hot water heater pipes and sure enough, reversed. My landlord came over shortly thereafter and fixed it, and I felt very smug when I had my first properly hot shower.
I once spent over a year debugging a furnace. It would run fine, except every now and then on it would just refuse to turn on for a day or two. Eventually noticed it was only rainy days, and was due to it thinking the exhaust was blocked because there wasn't enough pressure differential between input air and exhaust.
Turned out the installer had failed to put in the high altitude kit which told it to expect a smaller pressure differential. Mostly it was fine, but on rainy days with low barometric pressure, the difference would drop below threshold. But I had to be home for enough rainy days to figure it out.
I've done that twice with some shops I didn't have any history with.
Once I got them to replace my gearbox oil even though they said it wasn't needed. With another one I brought them a third-party pedal and asked them to install it. In both cases it did solve my issue by the way, even though I might have been wrong.
They don't care as long as you're ok to pay for the hours regardless of whether it solves your problem or not.
> Even the Audi dealership won't touch it because they're backed up too.
Getting an appointment at my Audi dealer is always 6+ week minimum wait. I don't think I've ever dealt with a dealer this consistently backed up. And my issues have been recalls and brake problems; but simple oil change is a similar wait time.
Not really, When powered down(battery diconnected) the computer then has to re learn parameters/pass internal validity checks. On a well preforming system this will be done in as little 1 or 2 rides but on a marginal system this can take many rides. the computer may not enter closed loop mode until these checks pass.
closed loop mode is where the computer set it's output values(fuel air ratio, timing etc) based on it's input sensors. open loop mode is where it sets it's outputs on known safe working values, but it may not run particularly well. i think this is related to the so called "limp mode"
Note My knowledge on this subject is very out of date. I had to figure it out for my old 2001 car. Newer vehicles may actually have flash storage and the computer parameters survive power out.
Yes, my Audi did not like being disconnected from power. If I remember correctly, it didn't go into limp mode but it was in some partial alarm state when it reconnected such that one had to mimic the Contra cheat code with their fob to get it to shut up. I jest, but yeah. Battery dead was not a condition with a transparent recovery.
Yeah, there are lots of ways cars could store such settings that wouldn't be cleared when main battery power is disconnected (which is something that regularly needs to be done when the car is serviced). Flash storage would work too.
They use a battery backup. The main symptom of having a dead cmos battery is that your computer starts loosing it's cmos settings, it's a feature not a bug. I have a system where it keeps the cmos config on flash. the only system I managed to brick(almost). I put a bad config in(never disable usb on a laptop where everything is usb) and there was no easy way to reset it. I had to buy a sioc clip and a chip programmer to reflash the thing the hard way.
The point is all cars have a wonderful battery backup system, and resetting the computer when power is lost is also a nice feature.
This isn't a good idea for an Audi. The computer doesn't like going without electricity. It runs a bit rough for a day or two afterwards and sometimes will "forget" how much gas is in the tank.
Some newer vehicles get upset when you do that, but older ones generally don't. You may lose radio programming though. I've got an old '95 Caravan that's going to be getting a quick disconnect like this one pretty soon though; the drain takes a few weeks to kill the battery, but we don't drive it very often either. Why do we still own it? With the back seats removed, you can easily put 8'x4' sheets of plywood in it.
I did this with an old vehicle taht was pretty basic without a lot of electronics.
the actual problem was sort of electronic-releated - leave lights on, drain battery to zero.
I put on a battery-cutout-monitor-thing. If battery drained to x%, it would disconnect the battery. to start the car, I think you had to stand on the brake pedal (add a load - like brake lights?) and it would reset and reconnect the battery - then you could start it.
this sounds like the sort of perfectly reasonable system that the original manufacturer could install in order to save the world from buying several hundred thousands replacement car batteries every year.
I believe this means that the part's computer has the vehicle's VIN loaded into it with a computer that can only be purchased/used by a certified technician. If he was to buy a new part, it would have a blank VIN; and if he pulled one from a junker, it would have that car's VIN.
(edit: should also say, I think a VIN mismatch would cause the ECU to refuse to work with that part and shut down)
It gets even more fiendish than that - you can have entire linked dependencies of locked modules. My old Saab’s key was locked to the security computer, to the column immobilizer, then to the engine computer.
Lost one key? Bummer. Lost all your keys? $2500, several hours of reprogramming time at the dealer, and a bunch of new parts shipped over from Trollhattan.
Thankfully the hacking scene has managed to bypass lots of this as Saab no longer exists to make parts. But this was the state of the art in 2003. I’m sure it’s even worse now.
It's going to get worse - on most cars the CAN bus is not encrypted and the messages are not signed. There was a comment here recently where a tech in the auto industry claimed that one manufacturer was planning to start signing the CAN bus packets. That will mean it likely becomes impossible to do things like re-code the PCM for a motor with a different VIN so you can re-use the PCM in another vehicle.
I wonder if this didnt start with 2005 7 series when it received FLIR Night Vision system. That thing was locked down due to ITAR and gave BMW excuse to pair with ECU.
Good luck in 5 more years when Tesla decide that your car is now unsupported.
I wonder if the Tesla mechanics are measurably better than the franchise dealers? I doubt it. I bet they have fancy testers, but when the tester is not showing 100% the right answer I doubt they are any use at all.
I've had mine (Model 3) for about 5 years now, so far they updated the hardware for free and it's gotten way better software for free over the air. It's a much better car now then when I bought it (so at least so far, the opposite of what you suspect is true).
Service has also been way better for me than when I had cars serviced by franchised dealers, but that may be because I'm in SF. I had an issue and two days later a mobile Tesla service car came to my house and fixed it in my driveway for free.
> I can’t replace it myself because it’s a coded part.
Can you please explain this? Why does it need ‘coding’? It’s not a VIN specific part I’d think - i.e. engine controller.
Also have you explored Ross Tech VCDS tool? You might be able to identify the coding on your part and use it on a new part. Replacing an ecu should be straightforward imo.
I’d say an engine controller which needs a specific vehicle information to work is going to be difficult.
But an ecu that controls say your headlights, depending on the oem it is just a part with unique part# that can be swapped. I work for one of the American OEMs and have replaced a ton of controllers on vehicles (engineering vehicles and customer vehicle). never had an issue. That being said I know German vehicles are next level pita since I do own an Audi. I just wasn’t aware about the coding part for the Audis.
In VAG cars, the ECU and dashboard are often paired to each other for anti-theft purposes, since the immobilizer is part of the ECU itself. To stop people from simply pulling the ECU and replacing it with one keyed to the immobilizer key chip they have - the ECU is paired to the dashboard, which often requires a lot more time and effort to remove.
For other modules they're often used across many different cars/engine/transmission/body type and regulatory markets. "Coding" a module tells it what accessories/options it's connected to, what else is in the car, what regulatory market it is in, and so on.
There's also "coding" for how the module behaves with certain features. For example, body control modules on US cars won't allow you to use the remote to roll up the windows, but by flipping one of the coding bits, you can enable it. There's also silly stuff like enabling / disabling gauge sweep on the dash.
(Explaining this mostly for others, you seem aware of at least some of this)
Yes I work in automotive and have routinely replaced ECUs. I work for an American OEM, I guess we don’t pair ecus to each vehicle. Unless its ecm/tcm/radio/security gateway and probably the key fob ecu. Everything else is just a part number that you can easily swap out. The few screws holding the controller and the wiring connector(assuming the controller is easily accessible)…
Though I own an Audi and use VCDS I was not familiar with pairing an ecu to the vehicle, however I was aware of coding in the VCDS. (I have not yet needed to swap an ecu on my car)
It's with the price just to see the looks on the faces of stranded motorists you help jump start in one or two minutes after they've been trying for an hour to jumpstart a minivan with a hatchback.
I had similar with my ‘15 Amarok. Battery would reliably be flat after parking for 24h, and I was either told there was no issue or it was unfixable, or I would need a new ECU, a new engine, a new truck.
After a short and inconclusive session looking for leaks through the fuse box, and then many, many hours of crawling around under the truck with a voltmeter, I discovered a frayed and heavily corroded +12v lead associated with the tow harness rubbing against the chassis. Fixed it. Issue resolved.
Unfortunately, a majority of mechanics simply aren’t competent, and the remainder don’t want to touch jobs which are mysteries, as they’ll just end up with a customer refusing to pay.
My last car, a 2007 Acura TSX, had the same problem--it wouldn't start after 4 or 5 days (maybe 2 or 3 during the winter). This model was from the first wave of touchscreens and bluetooth. Mechanics shrugged, pointed to the tiny, underpowered battery, and told me to drive it more.
Thankfully there are a lot of car enthusiast forums out there that have been plugging away on vBulletin for years. Someone on an Acura forum had figured out that the bluetooth module was always on and looking for a connection. I tried disconnecting it, and the problem was fixed.
God bless online car forums. Usually just awesome, well-intentioned, and helpful people that participate, with decades of knowledge indexed. There was a year way back there when the usefulness of the forums surpassed the Chilton's manual for me entirely - probably something to do with the addition of YouTube.
If you like watching amateur YouTubers rip out their hair chasing electrical problems, I recommend checking out Tavarish and Samcrac. It turns out cheap flooded auction cars have tons of those
lol @ Tavarish and Samcrac mentioned here, they are TV entertainers who are more concerned about drama. If you want to learn how to repair stuff, these guys are where its at ^
Do you mean mustie1? I used to watch that channel all the time, super entertaining. Shows you all the cool technical bits of engine repair and maintenance without boring you to death.
I don't often have to write the channel name out, I think you may be right. I enjoy his genuine and unstressed demeanor towards the camera. Just a guy wrenching on some hardware and talking about it.
> It seems like mechanics really don't like digging into the electrical part of the car.
That is certainly correct. Troubleshooting electrical issues takes a different set of skills than most mechanics are trained to deal with. Most mechanics are so busy that throwing parts at a problem is the best way to deal with issues where the cause isn't readily apparent. It's not worth their time to really dig into a difficult issue because if they spend 8 hours of shop time and try to bill that to a customer, the customer is going to just balk anyway. Better to throw a part at it, bill 1.5 hrs, send it out the door, and let the next guy deal with it. And then move on to some more profitable muffler and brake repair jobs.
This is thankfully not ALL mechanics, but it certainly is a large percentage of them, IMO. (Yep, even dealerships.)
This sounds like a majority of software engineers (and their direct managers) asking to add more CPU and RAM, when software is slow, instead of analysing a problem.
I guess mechanic's time is more valuable than parts.
In the military, the folks tasked with Mechanical maintenance of vehicles are an entirely different branch with different training, infra, etc than those tasked with Electrical maintenance of vehicles. Perhaps we need a similar dichotomy on the civilian side.
> It seems like mechanics really don't like digging into the electrical part of the car.
this is like saying that computer programmers don't like digging into the proprietary and binary blobs from other vendors.
I am trying to say it's not fair to blame mechanics because even if they want to dig into the electronics, the car manufacturers take active measures to prevent 'random mechanics' from digging around.
I suppose there are many rational and reasonable way to justify why this is so; probably all having some form of "because safety" and "because IP of vendors".
3 days is definitely not good for an NC. That said, mine is usually around 2 months before the battery is drained. That should give you some better expectations.
2006 NC owner here with a reasonably new battery. I keep it parked in an attached garage, so no extreme cold or heat. Mine will always start after leaving it for 1 month. I probably left it for 2 months at some point in the last two years, and it started, but that feels like pushing it.
The next time a mechanic tells you cars can't sit for a few days without a battery drain problem, ask how people who park at the airport get home after a two-week trip.
my battery is in fairly good shape. when my nc was sitting for long periods (months), the old battery eventually gave out. I had to have it jumped once, then bought a li-on battery to jump it with just in case.
the real key to long storage (beyond a trickle charger) is clean gas.
right now, though, it only has to handle 2 weeks of sitting at most (in an attached garage).
I've been to multiple mechanics, talked to roadside assistance people, posted in multiple forums, and talked to family about this problem with an old car of mine... and they all swore up and down that you should start a car if it sits longer than a weekend, maybe a week tops. They made me feel insane about it, it's one of those widely accepted things that's just not true most of the time.
There are other reasons to do that unrelated to the electrical system of the car. You're keeping the engine and any other moving parts lubricated. You're burning fuel, so that you don't wind up with as much degraded oil.
Sure but that's never mentioned... this is always talked about as common sense "this is how you make sure your battery doesn't die." It seems as though problems with parasitic drain are so common that this has become standard advice?
I think a cheap thermal camera would be very helpful in these situations with a relatively fast drain. A car battery is ~1 kWh so anything that drains it in a few days is dissipating on the order of 10 watts. In my experience that's very easy to see in a thermal camera. Just leave the car in a nice thermally-stable place for a day and then go hunting.
Imagine as a software developer you tracked your time working tickets, your pay was hourly, and depending on the ticket type you were prescribed how long that ticket should take and paid those set hours whether you went over/under the set hours. This sounds completely insane but this is how car mechanics work (as I understand a lot of shops do this). Replacing a battery might be 1/4 hour labor, replacing brakes on car model XYZ is 1.25 hours, etc. This is why mechanics won't really want to spend time on issues like this. It could be a wild goose chase tracking down every electrical sub-system in the car only to find something chewed thru a wire somewhere under a panel. The fix might be $15 in parts and 20 hours of actual labor but their system might quote replacing a wire as 1 hour labor.
Refinement: That's kind of how dealer mechanics work when they know that book time they are billing is more than the time they actually need for the task. For all the other situations, there is book time like "diagnosis, complexity high, diagnostic equipment involved" billed per 6min unit.
> It seems like mechanics really don't like digging into the electrical part of the car... but these days that encompasses more and more of what can go wrong.
Having worked as a mechanic in my youth, and worked as a software engineer in SV, the kinds of troubleshooting skills and detail-oriented attention span involved in diagnosing these issues are far more common in the latter than the former industries.
My impression is it's rare for someone with such abilities to stay a mechanic, they can earn far more money in tech, with less exposure to hazards.
On the subject of mechanics failing to diagnose electrical issues, I have my own story too:
Decades ago, back in IL, a friend inherited a low-mileage minimalist Ford Escort hatchback, manual trans, crank windows, it was a great little econobox to inherit, on paper. He kept having the battery die on him. Not being a mechanic or even a hobbyist gearhead himself, he kept bringing it to shops. They replaced the alternator, the wiring harness, the battery, the starter, they just kept throwing parts at the car. This is what most "mechanics" do nowadays; a poorly informed process of elimination via new parts, on your dime.
I hadn't been in contact with this friend for years when I heard about this "cursed" low-mileage car sitting in his garage, full of new parts with invoices totaling well over $1k. He was car-less at the time because of this situation. I offered to fix it for him, but he didn't have any confidence left in the vehicle or my ability to fix it, he was understandably fatigued by the whole thing. So I offered something like $250 and took the car off his hands.
15 minutes with a voltmeter revealed a huge voltage drop across the negative battery terminal and the chassis. Followed the negative strap to where it attached to the chassis and the area was corroded (recall it's an IL car). Removed the rusty bolt, wire-brushed the unibody steel behind it, the bolt and cable lug, slapped dielectric grease on everything and reassembled.
The car charged the battery fine and ran like a champ. I ended up selling it back to him a year or two later for basically what I paid plus a few hundred for my trouble/towing etc. He wouldn't even take it back until I had driven it for years to prove it was fixed. It was that brutal an experience for him, dealing with "expert" "mechanics" bleeding him dry.
Eric O from the South Main Auto youtube channel has some great videos on tracking down parasitic drains on modern-ish cars. From his videos I learned that often cars won't fully sleep for up to 45 minutes depending on the model and manufacturer, as well as checking for the voltage drop across fuses to figure out which circuit is draining the battery. With this I was able to track down a parasitic drain in my 2000 Beetle (failing door lock). I'm sure he mentioned other tips as well but those are the ones that stick out to me.
> The second told me it's normal for the battery to die if you don't drive it for 3 days
No, this is not normal. When the cars are manufactured, they often sit for more than 3 days just to get on a train for delivery. During transport they often sit for more than 3 days.
Heh, I also have an NC2 Miata, and I don't drive it for 3-6 month periods, and it still starts on the first/second engine spin.
If the battery is over 3 years old, I would replace it and start anew simply because of the life/abuse it'd had so far.
And then I'd start at the obvious ones - radio, aftermarket radio, alarm, etc. You can literally put a multimeter along the fusebox to see what is pulling voltage after - but most likely it will be something along 12V connected switch.
Wiring isn't problematic, but if you have a hitch installed it may be wired to always be on.
I also have a parasitic draw on my NC Miata, and ALSO use a battery tender every time I park it! If you find out the issue, let me know, would be interesting to see if it’s the same!
I'm sure if the mechanic was paid for it hourly they'd gladly tear your car apart trying to find the drain, but I think most people can't stomach that.
If anyone takes a car to a mechanic and asks for the estimate and the estimate is (100$ an hour until I find the leak and it may take a whole day or two), then most people are gonna gag.
For me it lasted close to 4 years. I had installed a tracker that uses a SIM that you have to keep topped up to receive the messages. No issues where I live so over time I stopped and didn't even think about it. Fast forward, I went through 3 batteries, couple mechanics, bought a portable jumper and a physical battery disconnector and even got a voltage meter and was mucking around with the fuse box etc and just grew to live with it until one day I was cleaning the car, saw the thing and just disconnected it and then a few weeks later I noticed that leaving the car on weekends the battery was fine the Monday (usually had to jumpstart as it would drain). Turns out the tracker was trying to send messages almost every minute THAT it couldn't send messages! So it had been doing that for years and was the tiny leech on the battery, not large enough to be detected but enough to slowly drain a brand new battery over the course of a few days.
I once had a parasitic drain on my BMW 3-series (e46). The dealership diagnosed the problem appropriately. It turned out to be the GROM Audio module I installed to replace the CD changer. It would drain the battery in about a week. So back to burning CDs and the issue was resolved.
Our 1990 Miata has a parasitic drain too. We installed a battery disconnect switch that takes about 5 seconds to connect or disconnect whenever we use the car. It's a little bit of an extra hassle, but it was an easy solution.
Let me translate the second mechanic's advice into our language: "Have you tried turning it on and off again?"
How much of your clients' time (and therefore money) would you spend chasing down a bug in software with an unknown root cause, which never appears as long as operations schedules a restart every couple of days? Software engineers are all too happy to dive into those rabbit holes, but that's because we don't usually have to look working stiffs in the eye and hand them the bill for our services.
That's likely not an incompetent or lazy mechanic. It's an honest one who doesn't believe they can solve your problem at a price you'd accept.
The first mechanic, on the other hand, I'm a lot more suspicious of. Did they bother testing the battery before replacing it?
That's probably what the author was thinking in February 2022 when they removed the batteries from their keyfob and only used the metal backup key, only to have the problem resurface less than a year later. It's probably safe to say that without batteries, the fobs were having no effect on the keyless system housed within the car.
Maybe the keyless entry computer stays awake to try to sync with a fob and then sleeps on success? If a user chooses not to use a fob at all, maybe it stays awake indefinitely trying to find a key to sync with.
Yeah sleep-wake problems often are related to things missing or broken in the (poorly constructed) system.
If the software on the car was poorly written, it may not have had a good back-off on the fob check, or if the fob itself was malfunctioning it may have left the car in a weird state.
While I wouldn't expect a fob to drain a car, it certainly doesn't seem implausible, especially given what the fix is.
Thinking this through a little bit further, I am wondering if subsystems can wake up if someone else uses their fob but the key signal doesn't match. Surely, at some level, something has to wake up to validate this.
Back when I was commuting there was an older subaru I always parked near in the parking garage that, whenever I unlocked my newer subaru with my fob, it’s alarm would go off. I sat there one day and did it over and over again, so it was definitely casual.
My conjecture: there is ambient RF noise on key fob's frequency (most likely 315MHz), which is causing the key fob receiver circuit to remain active and process any signals that it receives, rather than going to sleep and periodically waking up to check for incoming transmissions.
Had the same problem recently with a 2011 Mercedes. Didn’t figure it out until eventually the Keyless Go module (which allows push to start and automatic door unlocking) died. This fixed the parasitic drain.
Apparently when these sorts of modules are dying they fail to turn off and drain the battery by constantly searching for the key fobs. Incidentally while the battery drain was going on the key fob batteries were dying extremely quickly as well, also suggesting a lot more communication than normal was occurring.
While not a fix for this problem, a nice form of insurance is a Lithium Ion jump start battery[1]. I'm amazed that a hand-held battery can discharge fast enough to jump start my car, but ... it works great.
This is type of thing I've seen Eric O. over at South Main Auto[1] fix in dozens of videos. Mad troubleshooting skills, works hard to find root causes of things. He's pretty amazing, hidden away in some no-account town in upstate NY.
I really enjoy his channel- another good one is the Pine Hollow Auto Diagnostics channel, which has dozens of really involved parasitic draw diagnoses, have learned a great deal from that one.
Although I perform as much of my own auto maintenance as I can, I watch Eric O because he's entertaining, and all of the technical wisdom and advice just comes along for the ride.
Man, I would love for him to look at my 2003 Ford Ranger. I turned on the headlights and the anti-lock breaks engaged and the gauges randomly fluctuated. That's the most recent 'fun' story I have with it... Had a parasitic drain in this thing for 5+ years, but it operates 100% fine as long as I keep the battery charged with a tickle charger and can avoid jumping it. I've mentioned this to nearly every mechanic in the last 5+ years, no one has ever found the issue. Three said replace the battery, 1 said get a new alternator. Seems like those 2 would be the go-to for mechanics these days.
Check your ground wires, especially when you flip the light, honk the horn, or anything that is relatively high current. Corrosion on a wire can be fixed with Deoxit spray.
I used to be a computer engineer working on embedded systems for automobiles. Quiescent current is what the normal proper draw is called for these systems when the car is off. We worked very hard getting these numbers in spec but it was hard to catch everything especially in this case where the issue is probably due to software missing the sleep state for that module. This could be from bad code or your CAN/LIN bus is messed up in your car. 99% of mechanics (and engineers) have no idea where to start with debugging these issues and the answer will be "replace the module".
Anecdata: Some Hondas (e.g . 2010 Element) exert enough parasitic drain to put the car into a chronically undercharged state if a door is left ajar overnight.
The wrong thinking turned out to be: Car is safe in garage, interior lights go off, and my hands are full now -- why bother going back to the door to shut it completely?
And pro-tip if you need to go down the line past the fuse box: if you have a DC clamp-on ammeter, you can do multiple windings to 2x or 3x the current readings for smaller draws that the clamp-on might not pick up on.
Whoa, thank you. I was disappointed to read in my multimeter's manual (klein CL800) that the resolution for DC (up to 60A, if I'm reading "range" in the chart correctly) was only 10mA.
I had a parasitic drain on my Tacoma from my Automatic device (OBD-II reader / GPS tracker). I parked in a garage with no cell service and it drained the battery by constantly trying to connect. It was a non-issue when I commuted daily but started killing the battery when COVID happened and I switched to WFH.
Hard for mechanics to diagnose because they had no idea what that device was and also one of their first steps was to unplug it, put it in a cupholder, and plug their own OBD-II reader in.
Now the Tacoma is parked again most of the time because I got a Tesla and I'm noticing a voltage drop. I need to do more research, I don't know if this is normal for being parked or if there's a drain.
My father (a professional auto mechanic) took about 6 months to find a random engine shutoff during harder cornering in his Toyota, the culprit was a damaged inner insulation layer for a few wires within the wrapped wiring harness for the ECU, engine moving on the engine mounts would sometimes cause a slight harness shift and a short would cause a hard reset of the ECU.
Another story I have is my computer geek friend finding an issue with my Subaru Legacy, when no other mechanic could (I tried 3). The issue was that the car was old, first gen OBD, he downloaded some legacy asm ECU reader and made a diagnostic cable from spare parts, the issue was that the car would not start when cold, or be extremely hard to start. The culprit was damaged wire that tells the ECU that the engine is in cranking condition, therefore needs a different fuelling mode, without seeing this mode there was not enough fuel to start when cranking. Eventually this was found by hooking up an oscillograph to see the injector impulse length (duty cycle), as the old gen. diagnostic didn't show any errors, he compared a working car injectors to mine when cranking cold and found this. 10 years later I'm still amazed by his skill and dedication, I'd have scrapped the car otherwise. The car was sold eventually to another person who restored it (rust repair mostly) and his wife still drives it to this day, a 1994 Subaru Legacy.
Your friend’s fuel issue seems a lot like the intermittent no-start, low fuel pressure issue I have been debugging (3-series) for over a year now. It has defeated three dealerships and two independent mechanics so far, so I’m taking over the diagnosis now. No clue yet but today’s cars have too damn many computers in them.
which engine is that? I had an N54 335 for 5 years too, even E36 era BMWs have vastly superior error logging compared to JDM from that vintage. I hope you find your issue, but if it defeated dealerships it sounds troublesome. Log some live data with INPA if it's an older one, I have no experience with BMWs post EXX series to drop any advice...
Yea, N54. INPA helps confirm things, but just the idea that you have to pour over logs to diagnose a system that should be simple and straightforward... sigh. My last car was from the '80s and you could wrench pretty much anything you needed to without having to negotiate with a dozen computers running the show.
Eric regularly takes the viewer through the entire process including sometimes measuring electronic components with diagnostic tools that include oscilloscope functions.
Pine Hollow Auto Diagnostics goes even deeper. And focuses much more on the diagnostic. These two channels are good together in that they show different sides of the auto repair industry.
Ivan has done some parasitic drain diagnosis in the last few videos, including a new custom made logger that he had sent to him for tracking drain over long duration (like overnight or a couple days).
If you like South Main Auto and you havent' seen Pone Hollow Auto Diagnistics, I think you'll like it.
Skill and dedication indeed. The cost of being able to determine the root cause of the car problem probably costs more than the car itself. By a factor of ten.
Not just for highly computerized cars, I have a 97 Ranger where literally everything is manual and I don't even have a dome light anymore. There's a parasitic drain I can't quite figure out so I just make sure to start it every once in a while. Since I'm a year round bike commuter the truck can sit around for a long time, especially in winter when there's no house and garden projects. I've considered adding a manual battery disconnect, the only thing I'd lose are the pre-programmed radio stations.
After the pandemic started I was using my car about once a month or two and the battery was getting drained enough that the car would not start. This solar panel fixed that problem:
It is powerful enough to actually charge the battery instead of just maintain it, and it has a charge controller so it won't overcharge the battery. It did fail me once when we had a couple months of rain and overcast weather, but kept the car going otherwise. Eventually replaced the battery with an AGM battery and that plus the solar charger has worked very well.
Second this suggestion for anyone working through parasitic draw issues. He has a very solid, logical process for working through this (among other issues). I always enjoy watching the videos because it's like watching a who-dun-it where you're trying to guess the cause as he gathers evidence.
I found that my old (1997) Range Rover would flatten its battery after being parked outside my house over a weekend. If I didn't drive it for two or three days, it would be too flat to start.
It turns out that the BECM (body ECU) wakes up from sleep if it hears anything coming in from the RF keyfob receiver, on 433.92MHz, a standard ISM frequency. If it's not a valid fob code (and there is a surprisingly complex rolling code sequence) it'll go back to sleep after 20 minutes or so.
The oil tank gauge for my central heating had a remote sensor that... yup. Transmitted on 433.920MHz, of course.
The aerial for the keyfob is on the rear right corner of the vehicle, the oil tank for the heating was on the left-hand side of the driveway right at the back of the house, so when I backed into my driveway it was right beside the receiver.
I managed to do just fine with popping the lid off the oil tank once a month and looking inside to see how much fuel was left, having taken the battery out of the tank sender.
Can't recall where, but I recently learned this was a common issue at airports in the early days of wireless key fobs. The radar sweep at the airport would wake up the receivers causing cars parked there for a while to have dead batteries.
Once you find the fuse that passes the highest current, why not just place a switch button in series with it? Turn it off when exiting the car. Add an LED to show when it is on.
Personally, I'd go for a relay on a non-problematic circuit.
But I also get why you might just leave it disconnected - if it's pulling a few more amps with no clear cause, what is going to go wrong with it next, and what vital system will it take with it?
Oh a relay that goes on after you turned the key would be very handy. It is rather easy to find a point where there's only a voltage present after/while starting.
Of course, when the array controlled circuit would be vital for entering or starting, this would not be an option.
It was not all that easy to get to and not all that easy to mount a switch securely near by. May be I just lacked imagination.
The real reason and the part I skipped was the leakage current was getting worse over time. By the time I replaced the MICU it was close to 2A draw. It was just a matter of time before it died.
Had a similar issue with a 1.5-year-old vehicle that would have the battery fully drained in a little over a week if the vehicle wasn't driven. Had the dealership "fix" the problem three times and the problem was never fixed. I used the lemon law to get a brand new vehicle. I won't name the brand, but dealing with the manufacturer for this was rather trouble free.
Wow, I have a 2012 Mazda3 and have the same problem, which only really surfaced during the pandemic when I stopped driving my car regularly. The dealer said it was just a normal level of drain (even though it would die in just 4 days of non-use) and I should get a trickle charger. Which I did, and is annoying, but is something that I can deal with. I’m not sure if I’m willing to disconnect my remote entry to see if this fixes it, but I’m glad others have seen similar behavior because it really bugs me that the mechanics would just shrug and not really diagnose why this was happening.
We have a similar problem with a Lexus rx300 2001. The computer never goes to sleep. I was thinking of looking for (or making) a patch-fuse with a kill switch to stop the drain. Computer sleep depends on the flakey door controller switches.
Well this is entertaining. Our 2012 Mazda CX-7 seems to die if we don't drive it for a week. I've sort of ignored this problem but now I'm going to go check some fuses and see if this is the same problem.
I've never understood why cars don't just cut power before the battery is too low to start. If that's too complicated then just have an option to cut all power after X hours of non-running time.
I bet the old popular theft of car radio is a root cause; cars needed to stay running to keep the alarm active and power the radio continuously so it wouldn't require the unlock code every time you start the vehicle.
It's actually pretty impressive how much power even a 200mA draw at 12V is over a few days; it's about 58 Watt-hours or 3.5 iPhone batteries per day.
> I've never understood why cars don't just cut power before the battery is too low to start.
The design of most cars can't tolerate losing 12V DC power without problems. Clearing superfluous codes after a battery swap is SOP. Cars are build for two simultaneous customers: regulatory regimes and work-a-day 13.5K miles a year workers that start (and thus charge) their cars at least every 48-72 hours. Other use cases aren't really considered.
Some are better than others though. I have a 21 year old Nissan with a factory alarm and key fobs that sits for a month or more and starts fine every time. Heating helps a lot.
Something I forgot to mention here that supports my view that designers of modern cars are not thinking about this problem at all: after battery power is lost cars lose their maps and tables for ignition timing and fuel delivery. After a usually short period of time (20 minutes is typical) that data is lost and the engine etc. has to 'relearn' when the power is restored and the vehicle is operated for a time.
Cutting the power is easy, the hard part is for the car to know when/how to reconnect power when the driver wants to start the vehicle. I guess with a physical ignition switch, the mechanical motion of the tumbler could reconnect power. But, most cars don't have ignition tumblers anymore.
514 comments
[ 5.3 ms ] story [ 481 ms ] threadIf you're fancy, you can even make it a quick-disconnect setup so that if you drive away with it still on, it won't break anything (or do the poor-man's version where you have the wire connected to the wheel chocks).
On my car the emissions laws definitely caused more emissions than they saved that day.
Yes. Combine with the fact some only pass after certain conditions (eg. over 3000 revs for 10 mins, or idle for 5 mins, or only after 3 starts, only when the outdoor temperature is within this range, etc). Sometimes those conditions aren't stated in the manuals.
That makes the checks passing seem more like a random process and driving more will usually make them pass, eventually.
And drivers without a code reader won't know when it has passed, so they are usually instructed by mechanics to drive a long way to be sure before taking it in for a test (don't want to have to reschedule the test just because you didn't drive it enough).
A code reader really pays off sometimes to reset everything :)
I'm reminded how often people come to the RV forum trying to figure out parasitic draw. The answer almost always ends up being the propane detector. People forget it's even there.
Less of an issue on newer RVs that come with at least one basic solar panel installed from the factory.
Without time to dig into it myself I've just been parking it with a battery tender every time I come home.
I've heard this nonsense a couple times now. I was incredulous at first but everyone seems to say this. What are anyone's suggestions for a weekend-only car? (I bike to work so I don't need a car on weekdays.) Trickle charge the car battery on weekdays?
Most new-ish cars also store the dinamic engine calibration data in ECU RAM so if you disconnect the battery you can find your engine running harsher next time. So are the values of the headlights auto leveling system and other stuff.
You're also risking to reset the radio/media unit security code (on cars that aren't too new but also not super old), which if you got the car second hand and the former owner didn't save the piece of paper it came with and give it to you, tough luck unlocking it.
If you want to get fancy you can isolate the exact fuse and switch just that:
https://www.ebay.com/itm/175668177479
Thinking more, given how common parasitic drains are, carmakers should have all non-essential fuses in a separate subpanel section that can be easily switched off/diagnosed.
That's after being off for 30 minutes though. Some things stay on for a while.
If it's 'weekend only', battery should be no problem. If sometimes you leave it a couple months unused, that might be a problem.
Depending on how ready you would like to be when you're going to use it, or the access you have to your car (is it far away or in your garage?) you could add a connector for a battery charger/maintainer (some of them have accessories for plugs you can leave permanently, see the NOCO GC002 for an example), and/or add a battery switch to easily disconnect the battery.
Warning: your radio might ask for a security code if it loses battery connectivity. Be sure you have it.
My Honda HRV dies anytime it gets below -7C. Outside, inside, driven a bunch a day or two before -- once it's cold that battery taps out.
Probably a vampire drain but I suspect it's just a small battery. My old 2013 Mazda 3 starts in -25C without fail.
I agree colder weather is always a bigger issue and at 1-2 months I'd definitely be thinking of a trickle charger if at all possible. An older car I didn't drive in the winter would definitely end up with a dead battery if I didn't keep it on a charger.
They were also the original batteries used in Miata's since they give off a lot less hydrogen when charging. Since the Miata battery is in the trunk, and all lead acid batteries give off hydrogen when charging, and hydrogen in an enclosed space can also be rightfully called a bomb, having less of that is a good thing!
Did the Miata not come with a hole in the trunk floor for an exhaust hose for the battery? Both of my cars did...
Even if you’re spending $50/weekend (which is a fair amount of driving — Communauto is as low as $3/hr), that’s still quite a bit cheaper than owning a depreciating asset with lifetime maintenance costs (not to even mention insurance and fuel).
I'm jealous of the carshare options that seem to be available elsewhere. If it were as easy here as in this NJB video, it'd be a no-brainer: https://youtu.be/OObwqreAJ48
https://en.m.wikipedia.org/wiki/You%27ll_own_nothing_and_be_...
From your link. I'll let others decide whether or not that is "lying".
The average car in average condition should go around a month without being driven. Anything less than that either means a battery nearing the end of its life, not getting charged, or a parasitic draw somewhere.
Every 3 days? No way.
Good point on the maintenance. I do get the oil changed, again more by time than distance, and I'll look up the other stuff now that you've mentioned it, thanks.
That said, I'd expect most cars should be able to manage, say, 2-3 months. It's when I exceed 4 months with some regularity that battery life starts getting sketchy.
If you're driving the car every weekend, I'd expect that should be plenty. Note that fancier cars with more electronic features don't do well when only driven short distances. If you only drive 2-3 miles at a time, you might be discharging the battery faster than you're charging it.
That said, at between like 0F and -10F, I'd be intentional about starting that car every couple days. Below -10F to -20F, I'd start it each day. Below -20, I'd consider bringing the battery inside for the night - even if you have a battery blanket. Once a lead-acid car battery has frozen, it will die very quickly from then on - I'd believe within a few days or less even in normal weather. AGM batteries are awesome alternatives for the cold weather lifestyle, in terms of their resistance to freezing damage.
I have a week-end only car (2002 Mini) and I've had no issue letting sit for a week or two. Including opening/closing it without starting the engine to retrieve stuff in that timeframe.
And a week or two is on the low-end of how much you can let it sit. My SO car (1st gen citroen C3) often don't run for one or two month at a time and there is absolutely no issue with the battery.
On the other hand, if you have parasitic drain and access to an electric socket, there are battery chargers that can keep the car topped up. Or without an electric socket, just unplug the battery when the car isn't used (though that may disable some things, for example in my car the hatch is inoperable without power).
The horrible thing is if the car is modified, and you discover an un-fused circuit.
Weirded me out the first time I'm sitting there in silence smoking a bowl and suddenly fuel pump noises. Which were easier than normal to hear because I took the back seats (the only sound deadening in the back) out of that car so I'd have room for a full size spare tire instead.
With my current car (2020 Kia Sportage) I've noticed the head unit is waking up before I even start the car. As soon as you unlock the car I can see my flash drive light up with activity as the head unit gets ready to pickup music from where it left off.
Also, after turning the ignition on & back off there were a lot of (normal) transient draws (from things like the dome light that don't turn off immediately). Even with the problem circuit disconnected, it drew around 6A (!) immediately after the ignition was switched off, dropped to 0.4A after about a minute, and sat at that level for another 9 minutes before dropping again to 0.06A. That means if he hadn't waited ~10 minutes per test, he'd have been chasing draws that were actually normal.
Combine an intermittent fault with intermittent normal behavior, and you've got a troubleshooting nightmare.
[0] https://www.youtube.com/watch?v=rVScppKsfHs
I took it to several mechanics who couldn't find anything - because they'd unplug the battery to add a meter into the circuit. They weren't willing to wait 3 days afterwards to come back to it.
I ended up disconnecting the battery overnight. Worked fine then. Traded it in; I feel a bit bad for the next buyer.
How does this compare with the way presented in the article based on measuring resistive voltage drop across each fuse in turn?
Current passing through the resistor causes a tiny voltage drop, which is measured by the meter. As the voltage is proportional to the current, the scale painted on the cardboard behind the needle did the actual conversion.
In the author's setup the fuse takes the shunt's place; its resistance is apparently a known value that can be gathered from a datasheet.
"ye olde"? That's how every typical one works, till you go to clamp meter
> its resistance is apparently a known value that can be gathered from a datasheet.
that's... optimistic
The absolute current value probably doesn't matter - in fact I very much doubt a hobbyist-level multimeter is even capable of accurately measuring the (fractions of?) millivolts across a fuse.
It's more important as a boolean signal - "is there current being drawn on this fuse?", and then usually even those fractions of millivolts will generally be enough to make your el-cheapo multimeter register 1mV and tell you something is drawing current, where as a fuse with no voltage drop across it at all will reliably show 0mV (just like it would if the probes were shorted) on even a cheap meter.
Other method is getting DC clamp meter with low range (say 2A). Then just clamp in on wires. Bit fiddly as they work on magnetic field so they need to be reset before measure (And can work as compass in a pinch...)
I know what you're trying to say... but this reads like very dangerous advice and you may want to revise.
Ammeter between a battery's terminal and the car's electrical system.
Also note that you'll probably create sparking when doing this, and if the battery has recently been charging this can be dangerous (hydrogen outgassing).
Yah, that doesn't really make sense for small parasitic loads.
> Ammeter is not safe as general advice as a typical one won’t be anywhere near large enough capacity for working with a car battery
It should be fine with the car not "on", but the general level of danger of doing anything near a car battery requires caution.
If your multimeter has a sub-mV range -- which a decent meter should -- you can measure a voltage drop across the battery ground lead. I've successfully done that to diagnose a battery draw in my car.
It's the same principle as measuring the voltage across a fuse, except it ends up measuring the draw for the whole car, not just one circuit.
No. Not if the meter is in "Ammeter" mode (or is purely an ammeter only). An ammeter presents a very low resistance to the circuit you are attaching it to, low enough that for almost all circuits, it is effectively a "short".
If one connects an ammeter across a low resistance voltage source (i.e., across the car battery) the ammeter will appear as a short circuit to the battery, and one or more bad things will happen. The 'least bad' will be blowing a rather expensive fuse in the ammeter that protects it from this kind of accidental use. Several of the "most bad" will involve hot molten metal and/or extreme heat.
Yes. My first comment on this thread says not to do that, and suggesting an alternative is "Ammeter between a battery's terminal and the car's electrical system." instead. https://news.ycombinator.com/item?id=35514652
I was replying to "anywhere near large enough capacity" in the next post. The typical 10A range is massive overkill for typical parasitic loads in a car, and even things like the fuel pump deciding to kick on and run in most cars.
[1]This is just what I would do. Only do this based on a reputable guide if you don't have experience working around car batteries.
If you've ever seen a car's complete wiring harness you'll know why. It's not only because there's three gazillions of wires - they're also heavily insulated with tape that makes it a nightmare to access those wires in the first place.
My old math teacher would deduct points for not listing the units.
But I've heard so many stories of friends paying hundreds to auto sparkies only for the problem to remain.
Tell me about it. I'm in a similar boat. My audi a4 has had a parasitic draw for the past 10 years. My mechanic didn't want to take the time to address it because that sort of thing takes a serious amount of time and he's always backed up. I got out my fluke and narrowed the cause to what I believe is the comfort control unit. I can't replace it myself because it's a coded part. My mechanic won't do it because he won't confirm that's the cause.
None of the other mechanics in my area have the equipment. Even the Audi dealership won't touch it because they're backed up too. So, I keep a battery jumper in the glove box and hook the battery up to a tender when I can (outdoor parking).
I don't know what I'm going to do for my next car. Pretty much every car made after 2010 has coded parts and are more computer than mechanical. It does make me tempted to learn the electronics part and open a garage that focuses specifically on fixing electronic issues in modern cars.
(there are reliable sports cars, but it significantly limits the pool you can select from)
(I'm talking about pre-2010 BMW here; I would totally agree with most new ones being sterile)
BMW really ramped up the electronic complication starting with my E83, and that shit wears out right alongside all the typical flimsy BMW plastics and rubbers. Plus the delightful parasitic drains from tfa.
They also ramped up their maintenance-unfriendly designs around the E46 era. On an E39, a rear spring swap is easy. Try it on a newer one, and you will regret the attempt.
(Although the Mazdaspeed6 was 90% of a BMW with 10% of the maintenance bill).
I'd love to know which specific model you found sterile (wouldn't happen to be some 4 cylinder would it?)
People who drive boring cars love to parrot magazine reviewers... not realizing their definition of sterile is relative to a 911, not their Accord.
> Unless you are at a track or love cutting off people in traffic it feels really hard to have any fun in one.
I track my cars, but it's not hard to enjoy pleasant driving dynamics without resorting to being an irresponsible driver in a well-sorted car.
I'll defer to you on how fun cutting people off in traffic is though?
The Audi was great fun to drive though, no denying that. Just different priorities in what I want out of a car.
How many hours of shop time are you willing to pay for with no guarantee of a solution, and very few (if any) parts replaced?
Most people will give up, stop researching and buy a new car when the dealer tells them it can't be fixed.
Shoutout to "The Hondew Shop" in Dallas. He does great work on all Honda/Acura products!
He had the people at the dealership drive the car. They acknowledged the problem and got on a call together with Ford. Ford refused to even entertain the idea there was a flaw but less than a year later they came up with a "software patch".
Pete would bring it up with anyone he met. Do not buy a Ford Focus / Ford Fiesta with automatic transmission.
Imagine opening a computer software support business where you bring in your old proprietary binary programs that you got from who knows where, and the business is to fix bugs you've found.
Maybe some really big enterprises would pay for that sort of thing. But I doubt it would be feasible for consumer stuff.
It would be a crazy/fun kind of business to run. I'm sure it probably exists.
Bug Hunters
Extremely valuable service. I bet there'd be a lot of demand.
Having had a 2006 model car in my workshop forever rebuilding the engine, I have though it might be worthwhile starting a business doing that kind of trouble shooting but nobody will want to pay, because the vehicles involved are inevitably old and worthless. It's fun to do it for yourself, but as a business your typical customer who needs you won't be able to justify paying you.
My other thought was to possibly set up the tools and whatnot to support local businesses in doing that kind of troubleshooting on older cars but building the knowledge and tools to distribute would be a very expensive exercise.
I love this phrase, and it reminds me that the behavior exists in the computer world too. From the Jargon File:
field circus: [a derogatory pun on `field service'] n. The field service organization of any hardware manufacturer, but especially DEC. There is an entire genre of jokes about DEC field circus engineers:
Q: How can you recognize a DEC field circus engineer with a flat tire?
A: He's changing one tire at a time to see which one is flat.
Q: How can you recognize a DEC field circus engineer who is out of gas?
A: He's changing one tire at a time to see which one is flat.
Then there were the cases with arbitrary random bluescreens that were impossible to diagnose where I'd upfront tell them I could only do the bare minimum because otherwise they might have to pay 500 Euros for a solution that probably wouldn't work.
You might have to reprogram your radio channels...
His strategy isn't to pull fuses, it's to set the car sit for at least 30 minutes or so (with key off) and then check each fuse with a multimeter to see which has current on it, and then check everything on that circuit. Arm yourself with schematics and wiring diagrams, otherwise it'll end up being something of a wild goose chase.
(The idea being if you start pulling fuses, you can "reset" various computers in the car, which may show up as a false positive. It can take up to 30 minutes or so for all the various computers in a car to all go to sleep, although it's usually only a few minutes for most cars.)
I'm sure there was a more elegant solution (find a different fuse, rig up a button, etc.) but this was the path of least resistance and wasn't really annoying to do.
https://i.pinimg.com/originals/a4/c5/bb/a4c5bbd867b819a687ea...
I avoid very new cars like the plague though, so not sure how different they are.
It makes me anxious thinking about all the bullshit in new cars - touch screens, coded parts, 4G connections with GPS tracking, etc etc.
Completely disconnect battery when not in use.
And honestly on something that used it might be the best, taking it apart again just to replace the thing you thought you might not need to is PITA
Tho a bunch of that might be because I bought "sporty car" and those rarely have easy life in the first place...
The only computer in the diesel one is the glow plug timer and there is no computer in the carbureted one (both 1991 Nissan Patrols).
The injected one is a 2005 ford falcon and is starting to get all sorts of interior electrical niggles. Only wiring issues in the patrols is a left headlight short which was easily routed around using relays.
My experience is that a lot of independent mechanics will do what you ask them to do, so long as they think it’s not unsafe and have confidence that you’ll pay the bill without complaint when the task is done whether or not the problem is fixed.
Probably not a good “this is my first time meeting you, but please replace my flux capacitor” but if you’ve got a history with the shop, I’m surprised you couldn’t talk him into it. (I worked in a shop briefly in college. We’d do what the customer wanted, including installing parts they bought, but the only warranty was on a “we spent an hour; you paid for an hour; thus ends the transaction” basis.)
Finally, the service manager agreed to check it out with the understanding that if I was wrong, I still had to pay for the time they spent investigating.
I was right :-)
Bought my own solenoid online and surprise the bike started working again.
I once moved into a place that was great, except the shower would start out hot, quickly drop to warm, and then stay warm for as long as we cared to try it. I hated this, so every time I showered I'd spend some time trying to debug it, despite knowing nothing about plumbing.
Eventually I asked myself: why do normal hot water heaters work the other way, where they stay hot a long time and then get pretty cold? Clearly, there's some way of keeping the incoming cold water separate from the already-heated hot water, like by putting the output pipe at the top and the input pipe at the bottom. But if you reverse them, you might get what we had.
I leapt out of the shower and felt the hot water heater pipes and sure enough, reversed. My landlord came over shortly thereafter and fixed it, and I felt very smug when I had my first properly hot shower.
Turned out the installer had failed to put in the high altitude kit which told it to expect a smaller pressure differential. Mostly it was fine, but on rainy days with low barometric pressure, the difference would drop below threshold. But I had to be home for enough rainy days to figure it out.
That's a great quote. I think I'll use that instead of "where there's smoke ..." from now on.
Once I got them to replace my gearbox oil even though they said it wasn't needed. With another one I brought them a third-party pedal and asked them to install it. In both cases it did solve my issue by the way, even though I might have been wrong.
They don't care as long as you're ok to pay for the hours regardless of whether it solves your problem or not.
Getting an appointment at my Audi dealer is always 6+ week minimum wait. I don't think I've ever dealt with a dealer this consistently backed up. And my issues have been recalls and brake problems; but simple oil change is a similar wait time.
Assure to write somewhere your radio code password first.
closed loop mode is where the computer set it's output values(fuel air ratio, timing etc) based on it's input sensors. open loop mode is where it sets it's outputs on known safe working values, but it may not run particularly well. i think this is related to the so called "limp mode"
Note My knowledge on this subject is very out of date. I had to figure it out for my old 2001 car. Newer vehicles may actually have flash storage and the computer parameters survive power out.
The point is all cars have a wonderful battery backup system, and resetting the computer when power is lost is also a nice feature.
the actual problem was sort of electronic-releated - leave lights on, drain battery to zero.
I put on a battery-cutout-monitor-thing. If battery drained to x%, it would disconnect the battery. to start the car, I think you had to stand on the brake pedal (add a load - like brake lights?) and it would reset and reconnect the battery - then you could start it.
(edit: should also say, I think a VIN mismatch would cause the ECU to refuse to work with that part and shut down)
Lost one key? Bummer. Lost all your keys? $2500, several hours of reprogramming time at the dealer, and a bunch of new parts shipped over from Trollhattan.
Thankfully the hacking scene has managed to bypass lots of this as Saab no longer exists to make parts. But this was the state of the art in 2003. I’m sure it’s even worse now.
That will be quite hard to overcome.
The dealership model has really been awful for legacy manufacturers.
I wonder if the Tesla mechanics are measurably better than the franchise dealers? I doubt it. I bet they have fancy testers, but when the tester is not showing 100% the right answer I doubt they are any use at all.
Service has also been way better for me than when I had cars serviced by franchised dealers, but that may be because I'm in SF. I had an issue and two days later a mobile Tesla service car came to my house and fixed it in my driveway for free.
>Service has also been way better for me than when I had cars serviced by franchised dealers
Which franchised electric car dealer are you comparing it to?
Can you please explain this? Why does it need ‘coding’? It’s not a VIN specific part I’d think - i.e. engine controller. Also have you explored Ross Tech VCDS tool? You might be able to identify the coding on your part and use it on a new part. Replacing an ecu should be straightforward imo.
Really? I would have thought that'd be one of the most complicated things to replace.
For other modules they're often used across many different cars/engine/transmission/body type and regulatory markets. "Coding" a module tells it what accessories/options it's connected to, what else is in the car, what regulatory market it is in, and so on.
There's also "coding" for how the module behaves with certain features. For example, body control modules on US cars won't allow you to use the remote to roll up the windows, but by flipping one of the coding bits, you can enable it. There's also silly stuff like enabling / disabling gauge sweep on the dash.
(Explaining this mostly for others, you seem aware of at least some of this)
In Australia Automotive Electronician is a trade itself.
A good mechanic should be able to point you to their preferred auto electrical workshop when they suspect an issue is outside their own scope.
It's with the price just to see the looks on the faces of stranded motorists you help jump start in one or two minutes after they've been trying for an hour to jumpstart a minivan with a hatchback.
After a short and inconclusive session looking for leaks through the fuse box, and then many, many hours of crawling around under the truck with a voltmeter, I discovered a frayed and heavily corroded +12v lead associated with the tow harness rubbing against the chassis. Fixed it. Issue resolved.
Unfortunately, a majority of mechanics simply aren’t competent, and the remainder don’t want to touch jobs which are mysteries, as they’ll just end up with a customer refusing to pay.
Thankfully there are a lot of car enthusiast forums out there that have been plugging away on vBulletin for years. Someone on an Acura forum had figured out that the bluetooth module was always on and looking for a connection. I tried disconnecting it, and the problem was fixed.
South Main Auto
Watch Wes Work
musty1
https://www.youtube.com/@JimmyMakingitwork
lol @ Tavarish and Samcrac mentioned here, they are TV entertainers who are more concerned about drama. If you want to learn how to repair stuff, these guys are where its at ^
Probably worth taking it to an auto electrician, rather than a mechanic.
That is certainly correct. Troubleshooting electrical issues takes a different set of skills than most mechanics are trained to deal with. Most mechanics are so busy that throwing parts at a problem is the best way to deal with issues where the cause isn't readily apparent. It's not worth their time to really dig into a difficult issue because if they spend 8 hours of shop time and try to bill that to a customer, the customer is going to just balk anyway. Better to throw a part at it, bill 1.5 hrs, send it out the door, and let the next guy deal with it. And then move on to some more profitable muffler and brake repair jobs.
This is thankfully not ALL mechanics, but it certainly is a large percentage of them, IMO. (Yep, even dealerships.)
I guess mechanic's time is more valuable than parts.
this is like saying that computer programmers don't like digging into the proprietary and binary blobs from other vendors.
I am trying to say it's not fair to blame mechanics because even if they want to dig into the electronics, the car manufacturers take active measures to prevent 'random mechanics' from digging around.
I suppose there are many rational and reasonable way to justify why this is so; probably all having some form of "because safety" and "because IP of vendors".
The next time a mechanic tells you cars can't sit for a few days without a battery drain problem, ask how people who park at the airport get home after a two-week trip.
the real key to long storage (beyond a trickle charger) is clean gas.
right now, though, it only has to handle 2 weeks of sitting at most (in an attached garage).
It’s more common for routine maintenance to be charged by book time since it’s a standard procedure, less common for more complicated issues
Refinement: That's kind of how dealer mechanics work when they know that book time they are billing is more than the time they actually need for the task. For all the other situations, there is book time like "diagnosis, complexity high, diagnostic equipment involved" billed per 6min unit.
Having worked as a mechanic in my youth, and worked as a software engineer in SV, the kinds of troubleshooting skills and detail-oriented attention span involved in diagnosing these issues are far more common in the latter than the former industries.
My impression is it's rare for someone with such abilities to stay a mechanic, they can earn far more money in tech, with less exposure to hazards.
On the subject of mechanics failing to diagnose electrical issues, I have my own story too:
Decades ago, back in IL, a friend inherited a low-mileage minimalist Ford Escort hatchback, manual trans, crank windows, it was a great little econobox to inherit, on paper. He kept having the battery die on him. Not being a mechanic or even a hobbyist gearhead himself, he kept bringing it to shops. They replaced the alternator, the wiring harness, the battery, the starter, they just kept throwing parts at the car. This is what most "mechanics" do nowadays; a poorly informed process of elimination via new parts, on your dime.
I hadn't been in contact with this friend for years when I heard about this "cursed" low-mileage car sitting in his garage, full of new parts with invoices totaling well over $1k. He was car-less at the time because of this situation. I offered to fix it for him, but he didn't have any confidence left in the vehicle or my ability to fix it, he was understandably fatigued by the whole thing. So I offered something like $250 and took the car off his hands.
15 minutes with a voltmeter revealed a huge voltage drop across the negative battery terminal and the chassis. Followed the negative strap to where it attached to the chassis and the area was corroded (recall it's an IL car). Removed the rusty bolt, wire-brushed the unibody steel behind it, the bolt and cable lug, slapped dielectric grease on everything and reassembled.
The car charged the battery fine and ran like a champ. I ended up selling it back to him a year or two later for basically what I paid plus a few hundred for my trouble/towing etc. He wouldn't even take it back until I had driven it for years to prove it was fixed. It was that brutal an experience for him, dealing with "expert" "mechanics" bleeding him dry.
No, this is not normal. When the cars are manufactured, they often sit for more than 3 days just to get on a train for delivery. During transport they often sit for more than 3 days.
Mechanics don't like fishing expeditions for nonessential things. $100/h for 2+ hours diagnostic, followed by tedious parts to replace and pair up.
A new battery every couple of years and a trickle charger probably seems like better value to them.
Just need to hold a button when turning on the meter to disable the auto power of timer, then you run software on a laptop to log the data to csv
If the battery is over 3 years old, I would replace it and start anew simply because of the life/abuse it'd had so far.
And then I'd start at the obvious ones - radio, aftermarket radio, alarm, etc. You can literally put a multimeter along the fusebox to see what is pulling voltage after - but most likely it will be something along 12V connected switch.
Wiring isn't problematic, but if you have a hitch installed it may be wired to always be on.
If anyone takes a car to a mechanic and asks for the estimate and the estimate is (100$ an hour until I find the leak and it may take a whole day or two), then most people are gonna gag.
You need to find an expert in your car.
How much of your clients' time (and therefore money) would you spend chasing down a bug in software with an unknown root cause, which never appears as long as operations schedules a restart every couple of days? Software engineers are all too happy to dive into those rabbit holes, but that's because we don't usually have to look working stiffs in the eye and hand them the bill for our services.
That's likely not an incompetent or lazy mechanic. It's an honest one who doesn't believe they can solve your problem at a price you'd accept.
The first mechanic, on the other hand, I'm a lot more suspicious of. Did they bother testing the battery before replacing it?
Maybe the keyless entry computer stays awake to try to sync with a fob and then sleeps on success? If a user chooses not to use a fob at all, maybe it stays awake indefinitely trying to find a key to sync with.
If the software on the car was poorly written, it may not have had a good back-off on the fob check, or if the fob itself was malfunctioning it may have left the car in a weird state.
While I wouldn't expect a fob to drain a car, it certainly doesn't seem implausible, especially given what the fix is.
Apparently when these sorts of modules are dying they fail to turn off and drain the battery by constantly searching for the key fobs. Incidentally while the battery drain was going on the key fob batteries were dying extremely quickly as well, also suggesting a lot more communication than normal was occurring.
https://web.archive.org/web/20060213052358/http://www.campbe...
[1] https://www.amazon.com/dp/B0B55RFM1Z
1 - https://www.youtube.com/@SouthMainAuto
If the car had been operational for years and years, why would bad code crop its head now?
The wrong thinking turned out to be: Car is safe in garage, interior lights go off, and my hands are full now -- why bother going back to the door to shut it completely?
I tracked mine down using a different method which helps when your fuses don't all have test points (and no conversion table needed).
- Disconnect negative battery cable
- Switch your multimeter to 10A (don't forget to swap lead ports)
- Connect the multimeter in-line between the negative cable and negative battery post
- The readout on the multimeter should now tell you how much drain you have
- Start pulling fuses one by one until you see the amps drop to normal
nb: I am not affiliated with this chan, just a fan.
Least surprising finding
Hard for mechanics to diagnose because they had no idea what that device was and also one of their first steps was to unplug it, put it in a cupholder, and plug their own OBD-II reader in.
Now the Tacoma is parked again most of the time because I got a Tesla and I'm noticing a voltage drop. I need to do more research, I don't know if this is normal for being parked or if there's a drain.
Another story I have is my computer geek friend finding an issue with my Subaru Legacy, when no other mechanic could (I tried 3). The issue was that the car was old, first gen OBD, he downloaded some legacy asm ECU reader and made a diagnostic cable from spare parts, the issue was that the car would not start when cold, or be extremely hard to start. The culprit was damaged wire that tells the ECU that the engine is in cranking condition, therefore needs a different fuelling mode, without seeing this mode there was not enough fuel to start when cranking. Eventually this was found by hooking up an oscillograph to see the injector impulse length (duty cycle), as the old gen. diagnostic didn't show any errors, he compared a working car injectors to mine when cranking cold and found this. 10 years later I'm still amazed by his skill and dedication, I'd have scrapped the car otherwise. The car was sold eventually to another person who restored it (rust repair mostly) and his wife still drives it to this day, a 1994 Subaru Legacy.
https://news.ycombinator.com/item?id=35516985
https://www.youtube.com/@SouthMainAuto
Eric regularly takes the viewer through the entire process including sometimes measuring electronic components with diagnostic tools that include oscilloscope functions.
Ivan has done some parasitic drain diagnosis in the last few videos, including a new custom made logger that he had sent to him for tracking drain over long duration (like overnight or a couple days).
If you like South Main Auto and you havent' seen Pone Hollow Auto Diagnistics, I think you'll like it.
https://www.youtube.com/c/PineHollowAutoDiagnostics
https://www.amazon.com/gp/product/B07Y5TKMZ4
It is powerful enough to actually charge the battery instead of just maintain it, and it has a charge controller so it won't overcharge the battery. It did fail me once when we had a couple months of rain and overcast weather, but kept the car going otherwise. Eventually replaced the battery with an AGM battery and that plus the solar charger has worked very well.
Here's one such: https://www.youtube.com/watch?v=qDqlG5bRq8k
It turns out that the BECM (body ECU) wakes up from sleep if it hears anything coming in from the RF keyfob receiver, on 433.92MHz, a standard ISM frequency. If it's not a valid fob code (and there is a surprisingly complex rolling code sequence) it'll go back to sleep after 20 minutes or so.
The oil tank gauge for my central heating had a remote sensor that... yup. Transmitted on 433.920MHz, of course.
The aerial for the keyfob is on the rear right corner of the vehicle, the oil tank for the heating was on the left-hand side of the driveway right at the back of the house, so when I backed into my driveway it was right beside the receiver.
I managed to do just fine with popping the lid off the oil tank once a month and looking inside to see how much fuel was left, having taken the battery out of the tank sender.
But I also get why you might just leave it disconnected - if it's pulling a few more amps with no clear cause, what is going to go wrong with it next, and what vital system will it take with it?
Of course, when the array controlled circuit would be vital for entering or starting, this would not be an option.
The real reason and the part I skipped was the leakage current was getting worse over time. By the time I replaced the MICU it was close to 2A draw. It was just a matter of time before it died.
I bet the old popular theft of car radio is a root cause; cars needed to stay running to keep the alarm active and power the radio continuously so it wouldn't require the unlock code every time you start the vehicle.
It's actually pretty impressive how much power even a 200mA draw at 12V is over a few days; it's about 58 Watt-hours or 3.5 iPhone batteries per day.
The design of most cars can't tolerate losing 12V DC power without problems. Clearing superfluous codes after a battery swap is SOP. Cars are build for two simultaneous customers: regulatory regimes and work-a-day 13.5K miles a year workers that start (and thus charge) their cars at least every 48-72 hours. Other use cases aren't really considered.
Some are better than others though. I have a 21 year old Nissan with a factory alarm and key fobs that sits for a month or more and starts fine every time. Heating helps a lot.
Some cars computers block the start by software if the battery is "slightly" low (even if would not be too low to start).