Wico joysticks

[rkidd7952]

Hi,

I'm not sure the problem is a CIA.  I'm looking at the schematics (http://www.amigawiki.org/dnl/schematics/A4000_Rb.pdf), and the directional signals from both ports are handled by a shift register (U975), which is read by Lisa.  Only the fire button/left mouse is handled by a CIA.

Has the clock battery been replaced or removed?  (Edit: Yup, I see the earlier post now.)  U975 is right in that area, so it could be damaged by a leaking battery.

I'd take a look around U975 for dirt/corrosion to the chip or the PCB traces.  There's a column of pull up resistors between the clock battery and the option header.  R984 and R988 pull up the right line from the joystick, so if either of those is open or flakey, the shift register would see a spurious joystick-right indication.  IIRC, some of the pull-ups are on the underside of the motherboard, so it would be good to check both sides.

For reference, here's a picture of the motherboard.  The pull-ups are immediately above the clock battery.  Just to the right, there's a stack of three horizontally oriented  16 pin chips.  U975 is the top most.

http://amiga.resource.cx/photos/a4000,6

Robert

[rkidd7952]

I guess one thing I can do when I get home is see if the problem happens if I have nothing connected to the port2, it should do nothing if I have it disconnected?

Hi,

Did you try this test?  If it's a problem inside the Amiga, the symptom should appear regardless of whether a joystick is plugged in.  If the spurious movement stops when the joystick is disconnected, I'd look at the joystick as the culprit.

I saw in an earlier post that you also observed graphics glitches.  It would be good to check voltage on the 5V power rail.  The joystick grounds various lines on the connector to register movement.  When the joystick is released, the lines are pulled up to 5v by the Amiga.  If the supply rail is flakey, the line might get pulled to an intermediate value that could be interpreted as joystick movement.  A flakey supply could trigger other weirdness, such as graphic glitches.

Has the motherboard ever been recapped?

Robert

[rkidd7952]

Hi, thanks for the advice.
No, when I removed either of the two joysticks the problem is not there. The motherboard has never been recapped but visually it looks clean.

Thanks for that info.  Did the graphic glitches persist with the joystick disconnected?

With the joystick plugged in, you could try manipulating the cable (particularly around the strain relief where it enters the stick) to see if there might be a broken/marginal wire within the cable.  Kind of a long shot since the issue affects both joystick.

In my experience, bad joystick cables cause the stick to stop registering input rather than register spurious input.  To register spurious input, the broken wire would need to short to ground.  The cable on my Wico is shielded, but the shielding isn't grounded.

I would check voltages if you can.  From the joystick port's perspective, the only difference I see between no stick and a stick connected with no input is that the signal lines (directional + fire) in the cable get pulled up to 5v.  The cable will impose a small additional capacitance on these lines, perhaps enough to register an intermediate value if the power supply is flakey.

The joystick port supplies 5v on pin 7.  This normally isn't used on a joystick, and in fact my Wico has no contact in that position.  (If you look into the joystick connector with the row of 4 pins on the bottom, pin 7 is second from the right).  Shorting that to ground through a bad cable could pull down the 5v rail and cause flakey behavior.  Another long shot, but worth looking at your joystick cable to see if there is a contact there.

Robert

[rkidd7952]

On a side note, the power supply is a newer Enermax 400watt ATX that I re-wired to work with the A4000 so I'm highly doubting that I have issues with the 5v supply.

Hi,

The 5v supply from the PSU is probably good, but if the electrolytic capacitors on the motherboard are dry or failing, the 5v supply traces on the motherboard may be noisy or show low voltage.  If you have a multimeter, I'd check the voltage on either side of the R984 or R988 pull up resistors.  I can post further details if you want to continue troubleshooting this.

A leaking electrolytic doesn't always give a clear (externally visible) indication.  I recapped one of my PSUs last weekend and found several leaky electrolytics where the residue was contained to the area immediately below the cap.  I didn't realize those were leaky until I removed the capacitor.

Another experiment you could try to test the hypothesis that the peripheral wiring is related to the spurious signal is to look for evidence on the mouse port with the mouse plugged in.  The same directional signal lines are used for a mouse or a joystick, but they're interpreted differently for the two cases.  With a mouse, a spurious right signal would cause the mouse position counter to count up and down by one notch.  To check this, set the mouse speed to the max, then look for a left/right jitter in the position of the pointer.  The symptom would appear with 50% probability* for a given mouse position on your desk, so you should repeat the experiment (move mouse, look for jitter) a few times.

*The mouse signals motion by generating a square wave on the joystick directional pins.  Half the time, the mouse pulls the pin to ground.  The other half, the mouse lets the amiga pull the pin to 5v.  The symptom would only occur during this second half.

Robert