Post your worthless and/or pointless repairs

[SistMaticZ]

This was a worthless repair but for the sake of my early training days things got fixed.
I remember having four IGBTs exploded because of a stupide isolation problem with heatsinks.

After this i learned to always use a light bulb as a fuse in case i do some mistake.

I need to be more professional.

[lti]

I fixed a battery-backed nightlight that drained the batteries when unplugged. This one had three AA batteries (not rechargeable), and it was sold as being usable as a flashlight when unplugged. The battery/AC switching circuit was the problem. There was a transistor (Q1 in the schematic attached) that was biased on when there was no AC power, even when the power switch was off. I just connected R3 to the LED cathode instead of the battery negative.

[eccerr0r]

That is an upgrade, not a repair!

[sam_sam_sam]

I have done some repair that I sometimes wonder if it was worthwhile to do it

On another note
Every so often I have tried to mod something and it went very wrong ( but I have learned some very important thing not to do the next time around

[eccerr0r]

After a few years of not knowing what to do, I fixed my electric carving knife - the kind with two reciprocating blades.

One of the two (nylon?) plastic "connecting rods" broke. I had tried "plastic welding" with a solder iron, but it does not hold.

However recently new avenues opened and I hacked up a 3D model of the connecting rod and fed it to a 3D printer. The new PLA connecting rod fixes the problem and another broken device is saved from the landfill!

(took me a few tries to get the 3D model just right, alas, even the third and final rev wasn't perfect. Not sure why it printed differently than seen in the slicer, but neverthess it was close enough, a bit of ripping with needle nose pliers and swiping the metal parts from the old unit resulted in a fully functional connecting rod. 6 grams of PLA needs to get sent to the trash can as prototype waste, and if it breaks again I can easily make another one! And quite possibly if the other one breaks, hopefully it's the mirror image of the one I created!)

Ugh more thoughts - I really should put some lubricant on the "crankshaft bearing"...what is safe with both PLA (the new one) and nylon (the old one)?

And more thoughts - was it worth it to fix this thing? Should have just bought a new electric knife "FRU=whole unit"?

[sam_sam_sam]

Originally Posted by eccerr0r

However recently new avenues opened and I hacked up a 3D model of the connecting rod and fed it to a 3D printer. The new PLA connecting rod fixes the problem and another broken device is saved from the landfill!

(took me a few tries to get the 3D model just right, alas, even the third and final rev wasn't perfect. Not sure why it printed differently than seen in the slicer, but neverthess it was close enough, a bit of ripping with needle nose pliers and swiping the metal parts from the old unit resulted in a fully functional connecting rod. 6 grams of PLA needs to get sent to the trash can as prototype waste, and if it breaks again I can easily make another one! And quite possibly if the other one breaks, hopefully it's the mirror image of the one I created!)

Ugh more thoughts - I really should put some lubricant on the "crankshaft bearing"...what is safe with both PLA (the new one) and nylon (the old one)?

And more thoughts - was it worth it to fix this thing? Should have just bought a new electric knife "FRU=whole unit"?

I really like the not worth repairing something part of your story

I have done this my self making 3D parts to either fix something or to modify something and it a lot of trial and error to make it fit

On your question about what type of oil to use because PLA is supposedly made from corn then I would think corn oil should work just do not use a lot of it or could gum it up

Sometimes I do things just to see if can be and how difficult is it to do

[eccerr0r]

Yeah well this is the pointless and worthless repairs thread after all!

At least I don't need to buy another electric knife anymore and finally can say I made something "worthwhile" that was truly "made" and not merely downloaded and printed...

[momaka]

Originally Posted by eccerr0r

And more thoughts - was it worth it to fix this thing? Should have just bought a new electric knife "FRU=whole unit"?

Maybe. Or maybe not.
If your 3D part lasts longer than the original part, I'd say you did well, even if the time you spend doing all of that comes out to more than buying a new carving knife. But after all, what's the point of buying new stuff all the time if they'll just keep breaking. Might as well fix it once and do it right. Right?

I like your story, BTW. Would have been cool if you had some pictures too, of course.

Originally Posted by sam_sam_sam

Sometimes I do things just to see if can be and how difficult is it to do

Same here... which is why I like to take on so many "worthless" repairs.
After all, they aren't really worthless if you learn from them, and even better if you fix/restore whatever it is that needed fixing/restoring.

[lti]

I have an ongoing pointless repair/mod (something that never worked properly, just like the last one). It's one of those "USB 3.0" HDMI capture devices that actually uses USB 2.0. The hardware is capable of USB 3.0, and someone found one that actually uses USB 3.0 and dumped the firmware.

https://github.com/ArsenioDev/USB-Ca...Card-Reversing

Those dumps are the wrong size, but it looks like the chip is mostly empty. I want to see if that firmware also fixes the audio problem I'm having. The captured audio skips ahead by a small period of time (a few milliseconds). It just sounds like a click or pop.

The design of this thing is goofy. It converts the audio from digital to analog and then back to digital just so they can mix in a microphone input (and forgot DC blocking caps), and the HDMI output is split from the output of the video scaling chip. The output is always scaled to 1080p, but it also means that the output has HDCP stripped.

So far, I have improved stability by replacing the diodes in the power input stage (a simple diode-OR circuit) with diodes that have a lower forward voltage.

[eccerr0r]

Dang, no pictures, it looks pretty much the same as the original except I covered up one of the fairly large holes that was used to ensure it can be injection molded. With that big hole covered up with plastic, it immensely adds to the strength of the unit and suspect this is many times stronger than the original. Now I'm pretty certain the other one will break at some point, and then I'll get the mirror image printed and pictures will be available then.

Anyway onto another 3D printer project. I picked up a radio controlled toy car from the trash some many years ago, and turns out... it's broken. Steering gearbox is destroyed. So... 3d print gears?

Not so lucky this time. While I think I got the gears so that the steering motor can operate the gearbox, the return spring can't reset the motor back to straight/resting position as the gears stick. Once again I need to put some sort of PolyLactic Acid compatible grease which I have no clue what I can use as lubricant, and my gears are still probably slightly off and don't mesh properly due to size (along with printing errors).

Not sure if this too is a very valuable repair, though fortunately it's one of the medium sized cars and not one of those matchbox sized cars. It does require a ridiculous number of AA batteries of which I'll have to find some other way to power it for testing.

[lti]

The capture device firmware update fixed the audio. Unfortunately, it still won't connect at USB 3.0 (Windows says that it's USB 3.0 capable now, so it's probably a bad cable), and the compression quality is even worse (a good USB 3.0 cable should allow me to disable MJPEG compression).

I also have a cell phone picture of what the board looks like now. I added the extra caps on the input of the WM8960G (between it and the CS4344 DAC next to it) and screwed the new diode (replacing D5 and D6) onto the heatsink.

By the way, the firmware I had was identical to the 4MB file on that Github link. It's a 512kB Flash chip, so there are eight identical copies of the data in that dump. The 64kB file was the one that worked, and the rest of the chip is blank.

[momaka]

Originally Posted by eccerr0r

Now I'm pretty certain the other one will break at some point, and then I'll get the mirror image printed and pictures will be available then.

Heh, you never know. Sometimes, those quick "temporary" repairs end up being not-so-temporary.

Originally Posted by eccerr0r

Anyway onto another 3D printer project. I picked up a radio controlled toy car from the trash some many years ago, and turns out... it's broken. Steering gearbox is destroyed.
...
Not sure if this too is a very valuable repair, though fortunately it's one of the medium sized cars and not one of those matchbox sized cars. It does require a ridiculous number of AA batteries of which I'll have to find some other way to power it for testing.

Yeah, definitely not worth the repair, but still could be fun to do on a rainy day anyways.

Actually, cheap toy RC cars played a major part in what got me interested in electronics. Thus, all that time I spent tinkering with them when I was pre-teen (and even well into teen age) has made me learn a lot about motors, torque, and gear ratios. I remember we were introduced to levers and gear ratios in an engineering elective I took in high-school. Most people had no idea what gear ratio to use when we had motorized projects, whereas all of my projects always worked on the first try. So anyways, I guess those "worthless" repairs I did on my early RC cars weren't so worthless in terms of knowledge gained.

You can use a standard 5V power adapter if the car uses 6 AAs or less. If it's 8 or more, you can use a 12V adapter. Actually, a 5V adapter should work even if the RC uses 8+ batteries... the wheels just won't have a lot of torque and may spin a lot slower. Functions may also "jerk" On and Off if the voltage starts dropping too much. With that said, check what motor the RC has to get an idea for the current. Typically, a Mabuchi equivalent RS-260 or RS-280 motor (about 27 or 30 mm long, respectively, and 24 mm dia.) should run OK on a 3 Amp power adapter. However, if the motor is RS-360 or RS-380 size, you'll probably want at least a 5 Amp adapter for testing with the wheels off the ground (will need more if you want to see the RC drive on a surface - at least from stand-still / stalled position.)

And while on the topic of toy RC cars, I also have picked up quite a few myself from the trash many years ago. Fixed a few of them and may end up giving to some of my nephews if they are interested.

Anyways, I might post about them (with pictures) another time. One of them was quite garbage, but does have potential to be an OK RC car and I added some cool LED lights to it too, among other fix-ups. I also still have my RCs from back before I was even teenager. Now some of those have quite a lot of miles on them - literally. The most driven one, I think, has a few thousand easily (I used to run with it almost every day after school for 1-2 hours for at least 2 years.) The plastic bushings on the wheels are living proof - over 2 mm of play sideways and 1.5 mm of "wobble" wear.

[eccerr0r]

The silly thing about this RC car is that though it's crap, it's 2.4GHz and has a headlight... an infrared lamp...

Well I had enough AA batteries, was about to scrounge 8 "dead" NiMH AA cell which is enough to power the car, and I used my bench PSU to power the transceiver*. Steering still not quite right, sticks once in a while, but I can drive the car now without going just forwards and backwards. Probably have to go and redesign the gears again slightly, the annoying thing is that the gear ratio is for the motor to generate enough torque to actuate steering tie, but a spring returns the steering back to the center. That spring has to fight the gearing, "bearings", and the motor spinning and if any of them are not spinning just right, it won't return, which leads me back to the plastic gear lubrication problem.

Seems that people recommend lithium grease which appears to be fat based instead of long chain hydrocarbons. Sort of makes sense. Now the question is where to get a small amount without getting ripped off. Hazard Fraught Tools seems to have an 14 oz cartridge (used in full size automotive grease guns) for $4.49, and Wally World has a small 10 oz spray can for $11.69. I do wonder what the heck is up with this...

Plus I thought the normal grease I use for my cars were lithium, except all the grease I've used for my cars were pretty much universally black...


*transmitter and video receiver... it's got FPV sorta, figures, 2.4GHz had to be used for something.


---

oh I might have lucked out, I wonder if this stuff will work:

https://www.johnsens.com/all/4603-silicone-spray/

happened to be in my garage, but forgotten about it. Might have to grab a piece of waste polylactic acid and soak it overnight to see what happens...

---

Oh crap, no. The solvent/aerosol in this thing is NOT compatible with plastic. Perhaps if I spray some out and let it dry, and use the residue... Or in the case of the knife if I open it again, spray on metal, let it dry, and then put the plastic in it. Maybe this will work.

---

Mmm... grease... or not... ? https://www.webstaurantstore.com/nob...8LUBRIGRS.html

[eccerr0r]

Have to mention that food safe does not neccesarily mean plastic safe... I think that's a bit of confusion in the "interwebs." Food safe means if it contaminates food, nobody will get sick. Plastic safe means it won't dissolve and destroy plastic.

These don't have to both be true, unfortunately, so I'll have to ... ideally find one that works for both...

[lti]

I finally got a USB cable for that video capture box. It still used USB 2.0 only, and then it died completely ("device not recognized" error).

Also, the sample frames I posted earlier don't show how bad the video quality really is with the "USB 3.0 capable" firmware when it falls back to USB 2.0.

If this worked, I would have attached the firmware dumps.

[eccerr0r]

Uhoh. I think I rootcaused a failure on a DDR2 DIMM,,, which means I need to desolder and replace a ogawdtootiny resistor pack... which I don't have spares of...

I was thinking about soldering a ⅛W resistor in parallel to test my theory but this is way too big... not to mention signal integrity characteristics *sigh*

Not even a 4G DDR2 module...thus...worthless repair?

[momaka]

Originally Posted by eccerr0r

Not even a 4G DDR2 module...thus...worthless repair?

If you have the time and don't feel like doing anything else... maybe not.
I have a ton of little projects like that piling up already. Luckily, most don't take much space. Keeps me from never having a dull moment where I can say I'm bored and have nothing to do.

Originally Posted by eccerr0r

Uhoh. I think I rootcaused a failure on a DDR2 DIMM,,, which means I need to desolder and replace a ogawdtootiny resistor pack... which I don't have spares of...

No other defective RAM modules (or perhaps a dead GPU) to get one from? Even some motherboards may have such resistors around the RAM slots.

Originally Posted by eccerr0r

I was thinking about soldering a ⅛W resistor in parallel to test my theory but this is way too big... not to mention signal integrity characteristics *sigh*

I wouldn't worry about the signal integrity from slightly elongating the trace with an external resistor. Most likely, the module will still work just fine... provided you can solder that resistor on there. Now that (soldering a 1/8W through-hole resistor across the pads of the tiniest SMDs)... good luck! The pads on those SMD components are usually way too tiny to solder any through-hole components to them. Perhaps if you tried with some hair-thin wire and then solder that to the 1/8W resistor... but holy moly, that would be super ghetto!

[eccerr0r]

True I'm debating whether or not I want to sacrifice this 512M DIMM in case it too could be repaired, though technically sacrificing it for a module 4x larger is an obvious choice...

Would be nice to find something that truly has no value.

[eccerr0r]

FSCK there's no way I can solder wires around one resistor of an 0805 4x resistor pack... or anyone think this is even possible without shorting out the adjacent pins? LOL...

[stj]

0805? easy.
0603 is where i draw the line.