This is slightly off-topic, but I suspect the title attracts the right people and nobody else has been able to answer:
I came across an old PS/2 keyboard I'd really like to get working. I can best describe its symptoms as "it appears to be stuck in a boot loop." The LEDs go on for what could be the time it takes to run the self-test, then briefly go dark before it repeats. It does this indefinitely, as far as I can tell.
At first I thought it was an old capacitor across the power lines ( I assume as a low pass filter for unstable PS/2 power supplies) that had started leaking, causing it to brown out after some duration. However, replacing the cap did not appears to have changed nothing. I don't know what to do next and I have very little equipment to play with. Anyone have any ideas on where to turn to?
the cap across the power lines also provides a reservoir of power closer to the little microcontroller and other bits inside of the keyboard. an old keyboard might actually expect to be able to draw more power than a modern ps/2 port cares to provide, which would cause the microcontroller to reset at some point during its initialization process.
you could try to check that by putting a fast voltmeter across the power lines at the keyboard end and checking for drops.
you could also just try to fix it by replacing the cap with a higher capacitance one. maybe double or triple whatever capacitance is already in there. (can't go too high or the motherboard might think you've shorted the power line.)
might also check the resistance between power and ground in the keyboard (when it is disconnected, of course) and make sure that it hasn't developed any shorts or otherwise inappropriately low resistance between the power rails.
seeing how it behaves on different motherboards or some usb<->ps/2 converters might be informative, too.
in the unlikely event you're in the vicinity of boulder colorado, i'd be happy to look at it.
This seems somewhat likely and very interesting. No internal shorts, I have tested that thoroughly.
> you could also just try to fix it by replacing the cap with a higher capacitance one. maybe double or triple whatever capacitance is already in there.
Would, for testing purposes, two or three of the same capacitance in series do? I think the next step up I have on hand is like ten times the capacitance...
I'll see if I can find another motherboard with a PS/2 port. Unfortunately I'm located in Stockholm, Sweden, so no luck with Boulder. :(
> Would, for testing purposes, two or three of the same capacitance in series do?
nope; put two of three of what you've got in there already in, in parallel, that's how you sum capacitance. (internally they're huge plates held apart from each just the tiniest little bit, with capacitance =~ surface area. more caps in parallel makes more surface area, so more capacitance).
if we were using ideal capacitors, nicwilson's suggestion would be be fine. non-ideal caps have ESR, though, which will burn up some of the power we want to feed into the microcontroller if/when it tries to draw too much. paralleling caps lowers the effective ESR of the lump of capacitors capacitor.
> I'll see if I can find another motherboard with a PS/2 port.
there are also actual ps/2 to usb converters available as well. (not the little dongles that came with keyboards; those were just for keyboards that knew both protocols and needed physical conversion.)
something like
https://www.amazon.com/TRENDnet-Converter-Universal-10-4-10-... is probably just smart enough to perform the conversion, but not smart enough to limit the power going to the keyboard, so i bet you can pull 500mA through it, a fair bit more than the 275mA guaranteed to ps/2 keyboards.
Caps in series lower the overall capacitance (like resistors in parallel) , but raise the energy stored. Try putting 3 or 4 of the 10x caps in series (for 3.3x and 2.5x capacitance respectively).
I came across an old PS/2 keyboard I'd really like to get working. I can best describe its symptoms as "it appears to be stuck in a boot loop." The LEDs go on for what could be the time it takes to run the self-test, then briefly go dark before it repeats. It does this indefinitely, as far as I can tell.
At first I thought it was an old capacitor across the power lines ( I assume as a low pass filter for unstable PS/2 power supplies) that had started leaking, causing it to brown out after some duration. However, replacing the cap did not appears to have changed nothing. I don't know what to do next and I have very little equipment to play with. Anyone have any ideas on where to turn to?