Big Clive's videos are always informative and invariably entertaining. He goes in to so much detail, and often includes little circuit sketches. Perfect for casually learning about electronics.
Because he works as a rigger / repairer in the entertainment lighting industry, he has an unfortunate fixation with simple lighting devices. There are so many other interesting simple electronic devices he could explore. I used to subscribe, but got bored after too many Poundland LED lights.
His format is great, and he can certainly break it down well, but the content needs more variety.
mikeselectricstuff (https://www.youtube.com/channel/UCcs0ZkP_as4PpHDhFcmCHyA) is great if you want some variety of similar teardowns. He reversed engineered the Ipod Nano LCD amongst other things and managed to get his hands on some fully ceramic mil-spec CPU boards from the 80s(?) (Impressive hardware)
This, 1e6 times. CS/EE here and my father had an electrical automotive shop, and my grandfather and did a HeathKit electronics-by-mail course... and I still didn't have a solid grasp of practical electronics until the lives of Dave of EEVblog, AvE, Bigclive, ElectroBOOM, Strange Parts and working an internship at a GPS manufacturer.
There ought to be some engineering/trades cross-training required even if not going into that trade if one produces things that that trade will encounter.
Also, there ought to be an electronics-by-mail course subscription of not just basic electronics but moving towards advanced knowledge, theory, techniques, and their applications. People would buy it if it were consistently good and had support behind it.
I recommend Everyday Practical Electronics - great stuff and has been for decades, I think my father still has some copies of Everyday Electronics / Practical Electronics (I think the latter was more advanced; the two merged sometime between our electronics magazine buying childhoods) he bought as a teenager, I remember building a couple of projects from them (sound to light, and an electromagnet) before subscribing myself.
Pretty much what made me study electronics at school, and then EE/CS.
Issues don't come with parts, if that's what you were imagining, but they don't need to, that'd make them needlessly expensive and less timeless. One can get started with basic projects with a general 'component kit' and then buy more of the more frequently used and specialised components as required.
Undergraduate EE is more of a mathematical modelling track that happens to result in buildable circuits, which will probably work reliably and do what you expect because you've modelled them correctly. It's not so much about hands-on tinkering, which is why not many courses include metalwork or enclosure design.
It's also not about creative EE - clever solutions that use the minimum number of parts, as opposed to average solutions that do the job but aren't outstandingly elegant. IMO you either have a talent for that kind of design, or you don't.
Production Engineering - mass producing buildable circuits and enclosures for the minimum viable price - is a different discipline again. The distance from a breadboard or soldered prototype to a shippable mass-produced product is huge - which is why so many Kickstarter projects come unstuck.
Production Engineering was literally mentioned once on my EE course, and that was all we learned about it.
Uncle Bumblefuck (aka AvE) is the same but for all sorts of things, usually metalworking and mechanical engineering. He also does a really good Canadian stereotype. His speech style is often poetic.
I've learned a lot of the basics from them. Not with the goal of becoming an expert, but just enough that I can speak the language.
This video of his made laugh for about an hour straight. I must've rewatched it at least a hundred times by now.
[profanity warning]
https://www.youtube.com/watch?v=QLWKNTUHiEU
I got one about a year ago as a backup (thought why not), actually works rather well, small, but perfect for laptop bag, cheap enough to not stress and small enough to not take up much space. Not pondered scrapping one for parts, but certainly something I've become more mindful of since Maplins closed.
I own one too! Very happy with mine, I have also loaned it to people with small hands, RSI and other reasons to use it.
The scroll wheel is not to my liking but otherwise it is a very useful mouse.
I did actually try to buy another but they had a different design to the one shown here and not as good. Maybe stocks vary by outlet or it is seasonally available.
On the shoulders of giants. This single chip mouse has its origins in HP, they needed precise paper tracking for big format printers. HP spun it off to Agilent, who spun off Avago Technologies, current optical mouse sensor giant.
Ironic that this tech is actually pretty bad for precise tracking.
As the paper thickness varies, or the paper roughness varies, the distance between the paper and the lens goes up and down, effectively zooming the image in and out.
Since the focal length is so small, even a surface roughness of a few micrometers can have a 1% change in the measured distance travelled from the sensor.
In the case of an optical mouse, you can test this out. First turn off any "pointer acceleration" by your operating system, so the cursor position represents the actual movement of the physical mouse. Now move the mouse to the left, push down on the body of the mouse, while moving it back to the start point in the real world. Notice the cursor isn't at the screen start point?
That's because pushing down very slightly changed the optical distance, and the mouse thought it moved further than it really did.
It worked great for its intended application. You dont have to worry about paper thickness/distance in a commercial machine using high quality source materials. I dont even think it would play major role, measurement is done by correlation, moving closer/further only scales the picture.
Commodity mice are terrible as distance meters tho https://www.youtube.com/watch?v=CIRKRzw54Zs because you cant completely turn off acceleration/interpolation inherent to particular sensor package. It would require low level access to sensor parameters/options, something only available to clients doing xx thousand units per month volume from vendors like Broadcom.
> I dont even think it would play major role, measurement is done by correlation, moving closer/further only scales the picture.
Scaling the picture is the problem. What parent is trying to say is that the distance the physical paper has to move for the sensor to see 1 pixel of movement depends on the height of the sensor above the paper.
If you want to know how many millimeters the paper moved based on sensor input, this is important (especially if you want precision on the order of dot size in high-dpi printing).
> you cant completely turn off acceleration/interpolation inherent to particular sensor package.
You can. All the avago (and clone) sensors I have seen output in units of 1/16th of a pixel, or are adjustable via config registers - and even then, it's still linear. It's your operating system which does acceleration, and thats the reason the mouse feels 'different' on windows/mac/linux.
He earned my respect from a video on a USB charger with an electrocution risk and why you should never buy cheap USB chargers. https://youtu.be/3Hdn0MuCK_0
30 comments
[ 2.8 ms ] story [ 66.4 ms ] threadWithout a doubt the funniest is the teardown of South Park's Nosulous Rift device - https://www.youtube.com/watch?v=Ucymk0q70aQ
https://m.youtube.com/watch?v=n2ZZbuOeNmw
https://duckduckgo.com/?q=presto+hotdogger
His format is great, and he can certainly break it down well, but the content needs more variety.
EEVBlog has good variety but I don't think he's heard the phrase "brevity is the soul of wit".
[1] https://www.youtube.com/user/KerryWongBlog
I can't thank him enough-- I should probably subscribe to his patreon actually, to thank him.
There ought to be some engineering/trades cross-training required even if not going into that trade if one produces things that that trade will encounter.
Also, there ought to be an electronics-by-mail course subscription of not just basic electronics but moving towards advanced knowledge, theory, techniques, and their applications. People would buy it if it were consistently good and had support behind it.
I recommend Everyday Practical Electronics - great stuff and has been for decades, I think my father still has some copies of Everyday Electronics / Practical Electronics (I think the latter was more advanced; the two merged sometime between our electronics magazine buying childhoods) he bought as a teenager, I remember building a couple of projects from them (sound to light, and an electromagnet) before subscribing myself.
Pretty much what made me study electronics at school, and then EE/CS.
Issues don't come with parts, if that's what you were imagining, but they don't need to, that'd make them needlessly expensive and less timeless. One can get started with basic projects with a general 'component kit' and then buy more of the more frequently used and specialised components as required.
https://www.americanradiohistory.com/Practical_Electronics.h...
https://www.americanradiohistory.com/Everyday_Electronics.ht...
Undergraduate EE is more of a mathematical modelling track that happens to result in buildable circuits, which will probably work reliably and do what you expect because you've modelled them correctly. It's not so much about hands-on tinkering, which is why not many courses include metalwork or enclosure design.
It's also not about creative EE - clever solutions that use the minimum number of parts, as opposed to average solutions that do the job but aren't outstandingly elegant. IMO you either have a talent for that kind of design, or you don't.
Production Engineering - mass producing buildable circuits and enclosures for the minimum viable price - is a different discipline again. The distance from a breadboard or soldered prototype to a shippable mass-produced product is huge - which is why so many Kickstarter projects come unstuck.
Production Engineering was literally mentioned once on my EE course, and that was all we learned about it.
0: https://www.youtube.com/watch?v=NWGCS1U-xa8
I've learned a lot of the basics from them. Not with the goal of becoming an expert, but just enough that I can speak the language.
In swearing? I can recommend BBC TV series "The Thick of It" - which actually had a swearing consultant employed on the scripts.
https://youtu.be/nB1DlBpyS9w?t=111
The scroll wheel is not to my liking but otherwise it is a very useful mouse.
I did actually try to buy another but they had a different design to the one shown here and not as good. Maybe stocks vary by outlet or it is seasonally available.
Interview with Optical mouse inventor: Oral History of Gary Gordon https://www.youtube.com/watch?v=TxxoWhCzIeU
As the paper thickness varies, or the paper roughness varies, the distance between the paper and the lens goes up and down, effectively zooming the image in and out.
Since the focal length is so small, even a surface roughness of a few micrometers can have a 1% change in the measured distance travelled from the sensor.
In the case of an optical mouse, you can test this out. First turn off any "pointer acceleration" by your operating system, so the cursor position represents the actual movement of the physical mouse. Now move the mouse to the left, push down on the body of the mouse, while moving it back to the start point in the real world. Notice the cursor isn't at the screen start point?
That's because pushing down very slightly changed the optical distance, and the mouse thought it moved further than it really did.
Commodity mice are terrible as distance meters tho https://www.youtube.com/watch?v=CIRKRzw54Zs because you cant completely turn off acceleration/interpolation inherent to particular sensor package. It would require low level access to sensor parameters/options, something only available to clients doing xx thousand units per month volume from vendors like Broadcom.
Scaling the picture is the problem. What parent is trying to say is that the distance the physical paper has to move for the sensor to see 1 pixel of movement depends on the height of the sensor above the paper.
If you want to know how many millimeters the paper moved based on sensor input, this is important (especially if you want precision on the order of dot size in high-dpi printing).
You can. All the avago (and clone) sensors I have seen output in units of 1/16th of a pixel, or are adjustable via config registers - and even then, it's still linear. It's your operating system which does acceleration, and thats the reason the mouse feels 'different' on windows/mac/linux.
He earned my respect from a video on a USB charger with an electrocution risk and why you should never buy cheap USB chargers. https://youtu.be/3Hdn0MuCK_0