Thanks Jope, sliding door here as well... let's see if I can make this unit work again.

Hey there, I'm still having issues with my A1000 tank mouse. :-\

First of all, I've made some rough measurements/test with a micro-oscilloscope (DSOshell) and I've noticed there is only one square wave signal coming from each axis, not two.

Now, I understand the chip is a comparator. Is it safe to assume reference voltage should be the same across all four channels?

The schematics posted on the first page seems a bit different than the the actual unit I own. I haven't checked actual signal paths, but there are definitely differences regarding resistor values (unless I'm missing something?!)

Photos here:
https://imgur.com/a/nxFgAUl

Last but definitely not least, assuming there's something wrong with phototransistor and/or the IR emitter I can't figure out how to remove the plastic cartridge (?) that cover the former. Any suggestions?

1) The two or four IR emitters can be checked by plugging the disassembled mouse into your amiga, powering the amiga on, and using a smartphone camera to look for the normally invisible infrared light produced by each of the the IR emitters when they are powered.

If you cannot see any IR light via your smartphone camera confirm that there is a voltage drop across the affected IR emitter's electrical connections.

2) each axis needs to produce two "out of phase" square waves when that axis is rotated. The amiga uses these to determine the mouse's direction of travel on that axis.

These two "out of phase" square waves are produced by the interruption of the IR light detected by the two photodiodes when the slotted wheel rotates for that axis.

For the mouse to work correctly:
a) the axis wheel slots must be clear of any light obstructions,
b) each IR emitter needs to be powered, clean and working,
c) the photodiodes need to be properly biased electrically, be clean and be correctly aligned to see the corresponding IR emitter thru the wheel slots,
d) the voltage comparator IC needs to be powered (pin 12 to gnd, pin 3 to 5v), be properly biased, and have each of the four internal voltage comparator's input pins connected and biased correctly.

Hey @jlin_au thanks for your detailed answer!

I'll provide some extra feedback as soon as possible but I'd like to anticipate that unfortunately, due to my still limited electronics skills/knowledge, I do not know a lot about proper biasing. :-\

I do know about quadrature encoder signals though and how to measure voltages etc, will also try to draw a schematics that perhaps might help solving the issue.

Ok I think (hope) I have an interesting update.

First and foremost: via oscilloscope I can confirm I have all four quadrature signals H, V, HQ and VQ available at the mouse connector. Yet, there is absolutely zero mouse detected motion right now.

I also checked inside the mouse: the IR emitters do not show any light when captured by smartphone camera, but as I said quadrature signals are there and I can see about 1.11 voltage drop at each emitter's end (there are four, two per axis/wheel); quad comparator has roughly 4.8v power supply, but as I said right now I'm missing the required knowledge to check for proper bias.

Now, the one thing that seems different to an untrained eye (like mine) between this mouse and the other two, perfectly working ones is signal levels.

Quadrature signals on the non-working mouse seem to have a peak value at around 5 volts, sometimes a little bit less, while the two working units are always well over 5 volts.

Is this actually related to my issue? Anything more that I could (easily, hopefully) check? Perhaps an oscilloscope capture of the quadrature signals might prove useful to find a solution? I really don't know what to do anymore.

Problem solved, apparently.

Talked about this in a dedicated electronics forum, I got the suggestion to put a 470 ohm resistor in parallel with the 270 ones (smd marked with 271) and now both axis seem to work flawlessly.

I should get a scope very soon so that I can verify that quadrature signals are actually ok but I guess the mouse pointer moving fine is already a very good sign.