ksiumaxx wrote on 2025-06-18, 12:57:

Sorry for not responding, somehow I missed your message.
Nope, pin 2 is not connected to anything. Should it be for the trackball to work as a serial mouse? When I looked up the model on the internet I also only saw pictures of one with the 25 pin serial connector; none with the 9 pin. I assumed it was a PC mouse because I knew that Highscreen made IBM compatible PCs. Could it be that the trackball I own is an Amiga/C64 type mouse?

Sorry, I also missed your reply. It's odd, I think it probably is a serial mouse, with all those transistors for doing level shifting and weird split power supplies for the chip and LEDs. Are there any labels on the bottom of the case, if so, could you post a picture? But I'm fairly sure that a serial mouse would have to have something connect to pin 2 on the computer's serial port as that's the receive data pin. I'm sure someone can correct me if I've got that wrong.

Maybe someone tried swapping the connector or cable in the past so they could plug it in to a 9 pin port, but got it wrong? I'm slightly worried about the burnt resistor, I didn't think a serial port could supply enough power to do that, so maybe the chip or LEDs have also been damaged.

I hope you don't mind, but can we just double check which pin you're measuring as pin 2? If there's a label that suggests it is a serial device then we could have a go at guessing or working out what the pinout should be and rewiring it. I've managed that once before when working out an adapter for a dual serial / ps/2 mouse.

I'm terribly sorry for the mistake I made in my previous posts. The wiring I gave is incorrect and not matching the actual one in the trackball. I checked once again, and this is how it's connected (pins on the DB9 connector are numbered; no idea if correctly):
PIN2 - brown wire in the trackball
PIN3 - orange wire
PIN4 - green wire
PIN5 - blue wire
PIN 7 - white wire

I have absolutely no idea how I missed this mistake. I must've rechecked the wiring like 10 times already.

Ok, I think that makes more sense. So we've got:

1RXD 2 - Brown  (Tx)
2TXD 3 - Orange (-ve)
3DTR 4 - Green  (+ve)
4GND 5 - Blue   (GND)
5RTS 7 - White  (MS?)

Which I think is what I was guessing before. I've had a messy go at sketching out the schematic so it's easier to see what does what. You can see that the 390 ohm resistor is part of a current limiting circuit for the photodiode LEDs. If there's a problem with them then the ball won't work but the switches should still work. I think something to check is to plug the trackball in, turn on and boot in to DOS, then use your multimeter to measure voltages with the black probe on the blue wire and the red probe on each of the other 4 wires. Then load ctmouse, see if anything is detected, then re-measure the voltages. That should show if the correct voltages are getting to the mouse. If that all looks ok then it might be worth having a look at what the inputs to the chip are doing. It's always possible that the chip or the crystal oscillator is dead, but I'm not sure how to check those.

So I think I fried the PC I was testing the trackball with in the process... After checking the voltages, I plugged in a working serial mouse to check if it would recognise that, and the PC just stopped outputting video or giving any signs of life.

The voltages are as follows:
blue-white: 4.55V
blue-brown: -4.55V
blue-orange: -3.90V
blue-green: 3.80V
Ctmouse does not detect any mouse device with the trackball plugged in. I checked the voltages after running ctmouse, and they were all off by +-0.05. Same thing after a reboot. They were also slightly different from the first test.

Sorry to hear about the PC, I'm pretty sure that the serial port can't drive enough current to risk anything, so I don't know what might have caused that. Static shock? Or maybe a cracked solder joint/PCB trace near the port? Have you tried pushing/pulling the PCB in the area when trying to boot? Have you checked the PSU is giving correct voltages?

I'm guessing you'll want to get the PC working again first, but I think all those voltages look about right. One odd thing is that you'd previously measured the voltage on the supply to the chip as 6V, but now the orange-blue is only 3.9V. Orange looks like it goes to pin 9 on the chip and blue to pin 18. When you want to work on this again then I think the components that need to work to get the mouse detected by ctmouse are:
Q1, Q2, D1, R1, R2 - These do the level shifting from the data out pin on the chip (pin 8, 0 to -4?) to serial levels (-4v? to +4v?) on the Brown wire.
D2, R3 - I think the white wire signals to the mouse to identify itself, connecting to chip pin 6 through the 3Meg resistor R3. D2 allows white to pull the line -ve, but also allows it to provide current to the +ve supply once the signaling has been done.
U1, XT1 - chip and crystal

White should blip from +ve to -ve back to +ve when the driver loads, to signal to the mouse to identify itself. The mouse should then respond with a couple of bytes on brown.

It might be best to remove and check each component as some can interfere with each other (e.g. D1 is in parallel but reversed with the Base-Emitter diode of Q2, so will show as conducting in both directions). If everything checks out ok then it might mean finding a replacement chip.

High value resistors can be a problem to measure and can fail open-circuit. My multimeter only reads up to 2M so would just show R3 as open circuit even if was working or broken. I had a problem trying to fix an old electronic instrument once and it took me a while to realise that a 5M resistor, needed to get an oscillator started, had failed open.

Good luck with the PC.