Trackball Mouse Hack - Horizontal Axis Troubleshooting

[stimpyjoy]

I purchased a used trackball at http://www.ccagshow.com/ and have followed various tutorials to wire it through a PS/2 mouse to connect it to my computer.

Tutorial Links for Reference:
http://www.techdose.com/projects/Arcade-Trackball-Mouse-Hack/363/page1.html
http://mamewah.mameworld.info/mousehack.htm
http://www.willcoxonline.com/mame/controlpanel/controlpanel.htm

The problem is that I am able to move the cursor vertically, but not horizontally. When I swap the wires so that the trackball Horizontal leads to the mouse Vertical and vice versa, I still only get vertical movement of the cursor.  This leads me to believe that the trackball is working and there is something wrong with the mouse wiring.

I have photographed and labeled all the wiring.  Can anybody tell me what I might have done wrong?

[NickG]

looks like your H2 trace is cut, check continuity to the IC.

[PL1]

Have you used a cell phone/digital camera to check if the IR LED part of the optical circuit is working? (2 LEDs per axis)

When you slowly turn the encoder wheel, does the voltage vary between high and low on both H1 and H2?




Scott

[stimpyjoy]

Thanks for the replies.  This is my absolute first foray into electronics, so please forgive my ignorance.

looks like your H2 trace is cut, check continuity to the IC.

Can you actually see that from the photo? I'll check continuity when I get home.

Have you used a cell phone/digital camera to check if the IR LED part of the optical circuit is working? (2 LEDs per axis)

I didn't realize there were two LEDs per axis, but that makes sense.  As you can see from the photos, I have removed the emitters and receivers from the mouse, so I assume you mean the trackball LEDs.  I thought I had eliminated trackball error as a possibility by proving that either trackball axis can drive the cursor vertically when wired that way.  Am I missing something? 

When you slowly turn the encoder wheel, does the voltage vary between high and low on both H1 and H2?

I'm afraid this is way over my head. Can you tell me how exactly I can use my multimeter to check this? I want to learn but I'm a total noob.

Thanks!

[PL1]

Have you used a cell phone/digital camera to check if the IR LED part of the optical circuit is working? (2 LEDs per axis)

I didn't realize there were two LEDs per axis, but that makes sense.  As you can see from the photos, I have removed the emitters and receivers from the mouse, so I assume you mean the trackball LEDs.  I thought I had eliminated trackball error as a possibility by proving that either trackball axis can drive the cursor vertically when wired that way.  Am I missing something? 

Sorry, mis-read that part of your post.   

The trackball part of the circuit does appear to be working fine.

The thing to test is whether the signals from the trackball are making it to the IC on the mouse PCB or if one of the traces are shorted/open.

When you slowly turn the encoder wheel, does the voltage vary between high and low on both H1 and H2?

I'm afraid this is way over my head. Can you tell me how exactly I can use my multimeter to check this? I want to learn but I'm a total noob.

Standard IC layout is that if the "notch" on the IC is up, pin 1 is the upper left pin -- in your photo the notch is down, so pin 1 is the lower just above the "R3" label on the PCB.

Set your multimeter to DC Volts. (5v range)

Connect the black lead to ground, touch the red lead to IC pin 10 or 11 for the X-axis and pin 12 or 13 for the Y-axis -- see pgs. 5/6/17/20 of the datasheet here for the full pinout/description/electrical characteristics/schematic.

As you slowly move the trackball, you will see the voltage increase/decrease as the teeth of the encoder wheel pass/block IR from the LED striking the photodiode sensor on the trackball PCB.

Looks like the highs should be around 3.2v-3.8v and the lows should be around 1.2v-1.9v for the TP8452AP IC. (pg 17 of the datasheet)

Compare the readings on the good axis to the bad axis.

If one of the lines is stuck high, stuck low, or just not changing voltage like the other lines are, that will tell you where to look.

If all 4 lines (X1, X, Y1, Y2) are good, the problem may be a damaged IC.   


Scott

[stimpyjoy]

Ok, so I'm home and I've done some tests. Here's what I found:

I thought I had eliminated trackball error as a possibility by proving that either trackball axis can drive the cursor vertically when wired that way.

I was wrong. The horizontal optical circuit on the trackball is not working. The LED is not visible through my cell phone.  The vertical one works.  I don't know why I thought I had it working, but it doesn't now. In fact, when I switched the wires, I got NO cursor movement.

looks like your H2 trace is cut, check continuity to the IC.

I still don't see a problem just by looking at it, but there was a problem with my blue "H2" wire soldered to the mouse board. No continuity. I replaced the wire and now I can get either X or Y cursor movement but only from the one side of the trackball. I initially discovered this when I held the wires directly to pins 10 & 11, 12 & 13 on the IC. 



So now it seems my problem is a busted optical circuit on the trackball. Any idea what my next step should be?

[PL1]

Viewable angle on the LEDs can be tricky-- compare the voltages on the two trackball PCBs to be more certain.

1. On the horizontal trackball PCB,  connect the black lead of your multimeter to a ground like the frame of your computer and use the red lead to check the voltages at the following points:

-- +5v should be 5v and Ground should be 0v (never hurts to double check things like this, even though they're probably not bad)

-- H1 and H2 should change between high and low as the encoder wheel turns.

2. If H1 and H2 are good, skip to step 4.

3. If H1 or H2 is bad, disconnect it somewhere between the trackball PCB and the mouse PCB and test it again.

In troubleshooting, this is sometimes referred to as "divide and conquer" or "splitting the circuit".

If the line tests good after you split the circuit, the problem is somewhere between the split point and the IC.

4. Check data line voltages at IC pins 10 and 11 -- should be the same results as step 2.

5. If one data line voltage is bad, there is a problem somewhere along the trace for that line -- clean well around all the areas you've soldered with a Q-tip/isopropyl alcohol and examine the PCB closely with a magnifying glass looking for shorts/opens.

6. If it still doesn't work, unplug the trackball (no power going to it) and use your multimeter in Ohms/continuity mode to check that you don't have:
-- an open trace
-- a short between two traces


Scott

[NickG]

Thanks for the replies.  This is my absolute first foray into electronics, so please forgive my ignorance.

looks like your H2 trace is cut, check continuity to the IC.

Can you actually see that from the photo? I'll check continuity when I get home.

Yes, in the bottom left photo it appeared to be lifted/cut near the thru-hole where you desoldered the LED.

[PL1]

in the bottom left photo it appeared to be lifted/cut near the thru-hole where you desoldered the LED.

Agreed.

Also, be sure to clean all the solder splatter off the whole PCB using isopropyl alcohol and a Q-tip, an old toothbrush, or an acid brush.




Scott

[RandyT]

If you can get X or Y output, but only from one side (axis) of the trackball, it's pretty much a confirmation of a dead optical PCB.  The optical PCBs, specifically the opto-sensors, are for the most part a consumable item.  The IR LEDs in these things have a much shorter life than a normal LED, and will dim over time.  Usually, the first indicator is erratic behavior, and then total failure isn't far away.   That's not to say the IC, or some other part on the board isn't the culprit, but the IR LED's are usually the weak point.

Replace the PCB (might as well do both of them, but that's up to you) and it will very likely solve the problem.

[stimpyjoy]

Viewable angle on the LEDs can be tricky-- compare the voltages on the two trackball PCBs to be more certain.

1. On the horizontal trackball PCB,  connect the black lead of your multimeter to a ground like the frame of your computer and use the red lead to check the voltages at the following points:

-- +5v should be 5v and Ground should be 0v (never hurts to double check things like this, even though they're probably not bad)

-- H1 and H2 should change between high and low as the encoder wheel turns.

2. If H1 and H2 are good, skip to step 4.

3. If H1 or H2 is bad, disconnect it somewhere between the trackball PCB and the mouse PCB and test it again.

In troubleshooting, this is sometimes referred to as "divide and conquer" or "splitting the circuit".

If the line tests good after you split the circuit, the problem is somewhere between the split point and the IC.

4. Check data line voltages at IC pins 10 and 11 -- should be the same results as step 2.

5. If one data line voltage is bad, there is a problem somewhere along the trace for that line -- clean well around all the areas you've soldered with a Q-tip/isopropyl alcohol and examine the PCB closely with a magnifying glass looking for shorts/opens.

6. If it still doesn't work, unplug the trackball (no power going to it) and use your multimeter in Ohms/continuity mode to check that you don't have:
-- an open trace
-- a short between two traces

Scott

I clearly see two purple lights through my camera on the working side, and nothing on the other board.  Still, I'm going to try these steps when I get the chance. I'm not really in a hurry to finish the project, but I'm definitely enjoying the learning process.

Thanks for the replies.  This is my absolute first foray into electronics, so please forgive my ignorance.

looks like your H2 trace is cut, check continuity to the IC.

Can you actually see that from the photo? I'll check continuity when I get home.

Yes, in the bottom left photo it appeared to be lifted/cut near the thru-hole where you desoldered the LED.

Wow. Impressive. Having desoldered and resoldered the h2 wire last night, I may have inadvertently reconnected the trace.  I'll have to look closely tonight. Also, Scott, thanks for pointing it out in the photo.

Replace the PCB (might as well do both of them, but that's up to you) and it will very likely solve the problem.

I figured this would be the case. I found what looks to be a compatible board from Bob Roberts here: http://www.therealbobroberts.net/tboptic.html.

Are there any other resources I should know about as an alternative?

[PL1]

Yes, the Happ A052-1011-00 (red board) is an improved replacement for the A052-1010-00. (green board)

TheRealBobRoberts carries the red boards (Atari TB optic board A052-1010/1), as do Arcadeshop (Atari/Happ Trackball Optic Board), 8-Line Superstore and Twisted Quarter.

Divemaster127 may also have some in stock.

Drop him a PM if you'd like to get them from his ArcadeEmulator.net shop -- they aren't listed on the site, but his prices are good enough compared to Happ's to make it worth asking.   


Scott