A4000T mouse problem

[Castellen]

CIA or Paula?

Unlikely to be a CIA fault as this only reads the left mouse button/fire input, nothing to do with movement input despite some common misunderstanding.  You've already said the mouse buttons work OK.

Unlikely to be a fault of Paula as this device only reads in the four analogue paddle inputs (including right mouse button).  Again, you've said the buttons work.

Checking that the +5V supply is active is a good suggestion.  This is on pin 7 of both 9-way connectors.  You can use a bare metal part of the chassis or pin 8 of the mouse/joystick connector as a ground/0V reference.

In the A4000T this 5V supply is via polyswitch fuse F160 on the main board.  It's located on the edge of the board, between Paula and the keyboard connector.  These devices are essentially PTC thermistors in that they're near short circuit under normal conditions.  When fault current flows, they get hot and exhibit a high resistance, thus limiting current.  They revert to a normal low resistance state when the fault current is cleared and the device cools, hence they're often called self-resetable fuses.

If your main board is an A4000D (in a tower case), the 5V supply for the mouse/joystick ports is via polyswitch fuse F175, which is the large yellow disc shaped device behind the serial/Centronics ports.

If it's an A4000D that has had major RTC battery corrosion damage, it's often shift register U975 at fault that causes no mouse movement.

I wrote this article on mouse port repair which may help further.

[Castellen]

This is A4000T Escom board. I have 5V at pin 1,2,3,4,6 and I have 3.96V at pin 5 and 9, but I dont have 5V at pin 7. F160 seem to be OK, not hot and short circuit. I also have 5V at both side of F160.

That certainly looks like the problem then, the mouse needs the 5V supply to operate.  So just in case the wild joystick movement in the mouse port doesn't resolve it for you, the logical step is to trace the connections between pin 7 of the mouse/joystick port back to F160.  i.e. with the computer powered off, use an ohmmeter to check continuity from pin 7 back to F160.

There probably isn't much in between aside from those 40-way ribbon cables (does the Escom board have these the same as the AT version?) and perhaps a ferrite bead.  Don't rule out intermittent IDC connectors on the ribbon cables, though the 5V supply is fed in parallel from both cables so both would have to develop a problem for the 5V to dissappear, which would be uncommon.

The schematic of the ports board is here and the appropriate section of the main board is here.  Sorry about the lack of quality in those scans, my original paper copies looked to have been photocopied and reduced several times over, so that's as good as you'll get.

[Castellen]

When I check trace between pin7 and F160, it looks like short circuit. Also, when I check pin7 at the moment of power up I have about 1.6V, after that 1V, and after few seconds 0V.

You mentioned earlier that you had 5V at F160.  So if you have 5V at F160 and there is nothing at pin 7, then there is obviously no connection between the two.

So you need to check again; is there 5V on both sides of F160 while there is not 5V at pin 7?

If the supply looks OK at F160, then follow the tracks through to the ports board.  It probably goes through a ferrite bead or low value resistor on the way, so this would be the first area of suspicion.

[Castellen]

Definitely I have 5V at F160 and ferrite FB556 at both sides, but there's no connection between pin 7 and fb556. Stripe cables and I/O port module are OK. I can't see resistors between FB556 and PIN 7 (pin4 at 40 pins motherboard connector). Any suggestion?

Check continuity between FB556 and pin 4 of the mainboard connectors CN550 and CN560.  To find pin 4, look at the board so that the power connector P8 is in the top right hand corner.  Pin 4 of CN550/CN560 is the second pin in from the right side of the bottom row of each connector.

According to the schematic, the 5V route is F160 > FB556 > pin 4 of CN550 & CN560 > through both ribbon cables > ports module > pin 7 of joystick and mouse connectors

You may have to remove the main board from the chassis to inspect the tracks and vias on the bottom side of the board.  There may be an open circuit track if there has been an accidental short on the mouse or joystick connector, which is common if you plug a metal body connector not squarely in while the computer is on.  The problem with using polyswitch fuses for short circuit protection is that they are very slow acting, so you get high fault currents flowing for a relatively long time.  So you might find a bottom side track on the main board has fused open circuit simply because the polyswitch was too slow.

Shouldn't be too hard to find as you've mostly narrowed it down now.

[Castellen]

I have stable 5V at F160 and FB556, and I can't find any track between FB556 and Pins4. There's small sign of corrosion (possibly capacitors leaking before replacing) at bottom side of motherboard.

Check continuity between mouse connector pin 7 and what you believe is pin 4 of CN550/CN560 to verify you have the correct pin.  If that's OK, then the track between CN550 pin 4 and FB556 is open circuit.

It's a four layer board, so it may run on an internal layer.  In which case you'll need to solder a jumper between CN550 pin 4 and FB556.  You should be able to solder directly to the via following FB556 for a tidy job on the bottom side of the board.

[Castellen]

Hope this is safe (jumpering between Fb556 and pin 4).

That's why I said to verify what you think is pin 4 is actually pin 4.

Check continuity between mouse connector pin 7 and what you believe is pin 4 of CN550/CN560 to verify you have the correct pin.

If you connect 5V to the wrong place, the results are potentially disasterous.  However, you know that the +5V supply should be on pin 7 of the mouse port and that should be connected to CN550 pin 4.  Due to the absence of the supply voltage, you know there's a fault and have narrowed down where it is.

After you complete the repair and before you power on the computer, you should measure <2 Ohms between mouse port pin 7 and F160.

If you're not comfortable carrying out the repairs, I'm most happy to do it for you if you want to send the board to my service centre in New Zealand.

[Castellen]

I did it and everything works fine now! You are real wizard! Thank you very, very much!

Glad to hear you sorted it out, figured it wasn't going to be overly complex.

Sadly I'm far from a wizard, only an electronics engineer.