This project uses the same AG1171 SLIC module; it's just not enough by itself and still requires a microcontroller (or at least non trivial logic) even before you get to processing the audio streams.
Lovely job, I like the "here's where I've got to so far" style. This gives people enough to work from and make a hand-wired version while he's waiting for boards to come back. Even if he runs out of time to do anything it's leaving a working design and firmware for folk to start from.
It's definitely too easy to put off publishing until it's "properly" done and then it never happens.
There's a lot of hardware out there that does 95% of what OP's does, though you'll pay a pretty penny for it (e.g. https://www.peplink.com/products/cellular-modems-adapters/po... ) though ther's been a flood of them from noname sellers on Aliexpress and eBay under the banner of "LTE Landline Converters"
For the 21700 brand if you want to do this without spending a lot on batteries, but also be sure that any misconfigured hardware can't drain the battery below a damage threshold, use a button top protected 21700 cell. Here's a good 'house brand' one that is popular with 21700 flashlights.
Mmmmmmmmmmhhh. Maybe better to damage
A battery than to have a drone fall out of the sky? Also most protected cells have modest current capabilities, and might not be enough for many drone applications.
By its nature, MUAVs tend to push batteries to their limits and lifecycles tend to be limited. A battery that does mission X must have a maximum weight and a minimum capacity, and a minimum current spec. If you increase the life cycles, you decrease the maximum performance (range, payload, etc).
When you get into bigger drones, more modest restrictions on mission capability give acceptable tradeoffs, but for MUAVs short cycle life is often seen as an acceptable cost.
Frequently , after a hundred (or often much less) cycles in a MUAV a battery is retired from stressful missions due to significant drops in capability. It’s still fine for other uses, but at 90% it’s no longer sufficient for frontier MUAV applications.
The best choice to keep drones from falling out of the sky is to make sure that your auto land is set to a reasonable battery level (calibration by height if you will be at altitude) and use cell cutoffs at or below the minimum safe voltage if at all. (They are usually set to a few tenths of a volt above the minimum safe, often as much as 500mV.) The final battery voltage should be managed manually and batteries tested if over-discharged.
At least, that’s what prevents the most hardware from falling out of the sky onto people’s heads, sometimes bursting into flame on impact. Usually that is seen as the least desirable outcomes of a battery stress situation.
It is interesting to note that Kobalt 24v power tool batteries don't have any internal over-discharge protection. I guess this can make them better-suited for flying telephone use than some alternatives.
Accepting 5V over the RJ-11 jack seems asking for trouble. Someone is going to plug it into a regular phone socket at some point, so it needs to be capable of surviving at least 60VAC. Even if a secondary pair is used, there are so many different wiring standards in use, it'll end up plugged into something live sooner or later.
It's been a very long time since I had a cable modem that didn't have a pair of 4P4C (aka RJ11 or POTS) jacks on the back. These exist to make available the VOIP service that I haven't wanted for decades.
Others using this same enormous, American ISP probably have similarly-equipped gear.
But I must admit that it has also been a very long time since I connected anything there. I don't know if they supply 48v talk battery or not when the modem has a data-only configuration.
Calling people and playing with phones. I have an internal Asterisk server for me and a few friends. The nicest thing is that when the phone rings, it's a person who I know, not a scam. (Unknown external numbers get direct to voicemail)
Also great for my friends' kids. They can call each other or their grandparents or whoever, without having to touch a cellphone that tries to distract them with YouTube or whatever.
My house uses mostly analog with the Adit 600 (6 lines, 3 using the 70s wiring in the wall, 3 dedicated to specific phones). Everyone else has cheap Polycom desk phones and uses SIP/RTP.
Luckily telephone lines are very current limited. You'd be hard pressed to draw more than 100mA even with a dead short at ring voltages. One zener diode might be all you need.
Also not the most unusual thing I've seen over a RJ11, which includes keyboards, RS-232, other random digital IO and full USB with data. But these are normally on items that are not phones.
It would... if you're willing to just let the thing draw up to 0.1W all the time, and fail catastrophically if anything goes wrong with the zener or the supply is more capable than you anticipated (i.e. you plugged into a PoE jack by accident). This is a very bad idea.
We have a similarly converted rotary phone at my local astronomy club's observatory, used to receive visitors. It's driven by an arduino if I remember correctly. The main issue is that the sound quality of its built-in speaker is quite poor, especially compared to modern cell phone speakers we're used to nowadays, with a narrow frequency window and frequent echoes.
Interesting. I remember voice clarity being exceptional on old handsets and being really subpar on gsm and later. Only now we FaceTime voice and discord the quality seems to be coming back.
How can that be if the speakers in phones ubiquitously were so bad?
My understanding is that at some point the audio of phone calls ended up being compressed multiple times on its way to the receiver as it went through different voip setups and such.
Prior to that, uncompressed or compressed once calls weren't so bad on bad speakers.
In contrast, I've been messing around with some old analog phones and VOIP stuff recently, and I've been blown away with how good the quality is over what I was expecting. The audio bandwidth is low yes (it's G.711u, so 8 KHz), but the actual quality is great, with no hyper-compression artifacts or over-applied background noise filters. And yes this is on the original 60's carbon microphones.
The phone calls I normally make on my cell still all sound like they are underwater.
Cool! But who remembers phone numbers anymore? Maybe a small screen that you can navigate by dialling 1 (down), 5 (up) and 0 (select) would be a cool add-on (after dialling 0 or 411 to reach the operator perhaps).
Fashion bag. Dock your cell-phone inside, bluetooth pair to the purse (w/ functional receiver + off-hook sensor) and treat that as a siri-enabled headset in a phone w/ fabric coiled cord.
"Ring ring! Hello? Hey Siri, what's the weather and call my ex so I can hang up on them!"
42 comments
[ 15.4 ms ] story [ 2655 ms ] threadhttps://youtu.be/nibV_idJjlg
https://www.ebay.com.au/itm/305557776577
It's definitely too easy to put off publishing until it's "properly" done and then it never happens.
https://liionwholesale.com/collections/batteries/products/pr...
By its nature, MUAVs tend to push batteries to their limits and lifecycles tend to be limited. A battery that does mission X must have a maximum weight and a minimum capacity, and a minimum current spec. If you increase the life cycles, you decrease the maximum performance (range, payload, etc).
When you get into bigger drones, more modest restrictions on mission capability give acceptable tradeoffs, but for MUAVs short cycle life is often seen as an acceptable cost.
Frequently , after a hundred (or often much less) cycles in a MUAV a battery is retired from stressful missions due to significant drops in capability. It’s still fine for other uses, but at 90% it’s no longer sufficient for frontier MUAV applications.
The best choice to keep drones from falling out of the sky is to make sure that your auto land is set to a reasonable battery level (calibration by height if you will be at altitude) and use cell cutoffs at or below the minimum safe voltage if at all. (They are usually set to a few tenths of a volt above the minimum safe, often as much as 500mV.) The final battery voltage should be managed manually and batteries tested if over-discharged.
At least, that’s what prevents the most hardware from falling out of the sky onto people’s heads, sometimes bursting into flame on impact. Usually that is seen as the least desirable outcomes of a battery stress situation.
I got this article conflated with another I was reading. My apologies to all.
Still, if you don’t want your phone to fall out of the sky, Put it on a drone with proper battery limit discipline.
;)
I got this article conflated with another I was reading. My apologies to all.
Others using this same enormous, American ISP probably have similarly-equipped gear.
But I must admit that it has also been a very long time since I connected anything there. I don't know if they supply 48v talk battery or not when the modem has a data-only configuration.
Also great for my friends' kids. They can call each other or their grandparents or whoever, without having to touch a cellphone that tries to distract them with YouTube or whatever.
My house uses mostly analog with the Adit 600 (6 lines, 3 using the 70s wiring in the wall, 3 dedicated to specific phones). Everyone else has cheap Polycom desk phones and uses SIP/RTP.
We still have a landline. It's part of the cable package, but it mostly just gets robocalls.
I am not normal.
Also not the most unusual thing I've seen over a RJ11, which includes keyboards, RS-232, other random digital IO and full USB with data. But these are normally on items that are not phones.
It would... if you're willing to just let the thing draw up to 0.1W all the time, and fail catastrophically if anything goes wrong with the zener or the supply is more capable than you anticipated (i.e. you plugged into a PoE jack by accident). This is a very bad idea.
The Secret Life of Machines TV show has a fun demo of how the speakers in old telephones produce sound https://youtu.be/MCgCSMq5Xpo?si=CT5aOGttBpPk9tEM&t=458
How can that be if the speakers in phones ubiquitously were so bad?
Prior to that, uncompressed or compressed once calls weren't so bad on bad speakers.
The phone calls I normally make on my cell still all sound like they are underwater.
Have a bot do the rest, exactly like a telephone operator.
"Ring ring! Hello? Hey Siri, what's the weather and call my ex so I can hang up on them!"
An uglier way to do it is with an off the shelf pulse to tone box and a router with VoIP support, like an N300.
Do you think it could use WiFi calling? I suspect that can't be done is it requires one of a limited set of carrier approved phones in each area.