Looking for Penny & Giles / Traxsys trackball help please

[hsr]

Hoping someone has now or had used a trackball like the one I have and can offer up some wiring help to get it interfaced and could be so kind as to help me sort through a problem I’m experiencing. 

While gathering parts for a new small cabinet project, I acquired a Penny & Giles trackball.  My plan was to interface this ball to the PC through an Opti-wiz.

Tracing the leads gets lost with all the other resistors and the IC on the board.

My apologies in advance if the terminology below is wrong.  The trackball has a RJ45 connector, so I had assumed that by taking an Ethernet cable, cutting and stripping it back, I’d have the wires to interface with.  However, what I am finding is not what I had hoped for.  Of the eight wires, only three (total) show any continuity when checked with a VOHM to the X and Y LEDs and the X and Y optic sensors.

Here is a photo of the very same trackball where you can see the LED X, QX, LED Y and QY solder points:



This photo shows the Cat5 connector:


I am only seeing continuity in the following:
LED X – white/green
X left – white/green
X center – black
X right (where it’s marked QX) – white/black

LED Y – black
Y left – white/green
Y center – black
Y right (again where it’s marked QY) – white/black
The wire descriptions are the colors of the Cat5.

Realizing the LED are diodes, I reversed the polarity of the VOHM, but no change in results.  The resistance readings vary from zero to above the 2k scale.

I checked an old PS2 mouse hack last night that powers a Tron spinner I have and the results are similar in resistance readings, but that doesn’t obviously help me figure out where/how to tie into this ball except to think I need to apply 5v to the LED and take signals from the outer two leads of the sensors back to the Wiz for the respective X and Y directions.  Okay, seems easy enough, but all the resistors have me concerned that maybe the LEDs and sensors don’t want to see 5v.

So again, I am hoping someone might be able to shed some light on how (if at all) to possibly attach this trackball to a Opti-wiz.

Thank you in advance,
Jeff

[lilshawn]

to check a diode, you need to use the DIODE CHECK function on your meter.

and no, the LED will blow up if you feed them un regulated 5 volts.

even though they may run on 5 volts, the current needs to be regulated down to ~15-30 milliamps. and more and the LED goes PFFT!

if you are up to some hacking, i'd search the IC you have there on the bottom corner of the board and see what it is. perhaps this trackball is a serial connection and needs only a custom cable kinda like the old cisco console cables...

[hsr]

to check a diode, you need to use the DIODE CHECK function on your meter.

if you are up to some hacking, i'd search the IC you have there on the bottom corner of the board and see what it is. perhaps this trackball is a serial connection and needs only a custom cable kinda like the old cisco console cables...

Duh on the diode check...  Should have known better on that one.

If it is serial, does that mean it won't output what an Opti-Wiz wants to see?

The IC is a 74hc14d and I found this http://www.nxp.com/documents/data_sheet/74HC_HCT14.pdf documentation for it.  Little above my head on the subject, though it does show me where the voltage input, data in and data out is.  Hopefully, I can trace those back to the eight pins of the RJ45.  But, how to tell which outputs are the X and Y since there are six?

[lilshawn]

Traxsys is still around. perusing their site it appears you may need to interface it with a card to convert the rotation pulses to mouse data:

http://www.traxsys.com/OEMTechnology/SwitchesAccessories/InterfaceCards/tabid/1356/Default.aspx

For trackballs, a range of interface cards that will convert phase quadrature outputs into mouse compatible data and allow easy connection to industry standard mouseports including PS2, USB and RS232.

you could still hack this to hook up to the optiwiz i'm not exactly sure how, as ive never used one. shouldn't be too hard.

[hsr]

Thanks again Shawn,

I was just looking on their site and they have cables shown too.  They use Molex on one end and USB or PS2 on the other end.  Both (four different cables total) only use 4 wires total as 5vdc+, 5vdc-, signal + and either signal - on the usb or clock on PS2. 

They have a picture of cable that is RJ45 to Molex and it has six wires, but no technical document for it.

That matches (sort of) what I'm seeing out of my unit.

Here is the pinout for the IC from that pdf I linked above.  According to that document, 7 is the vdc-, 14 is vdc+, the remaining even numbers are outputs and the remaining odd numbers are inputs.

Here is how they correspond to the RJ45 connector on the trackball.

Because 7 and 10 are connected elsewhere on the board, I can understand the 6 wire cable shown on the Traxsys site.  Trying to figure out which pair of 2, 4, 8 and 10 combinations are the X output and which are the Y output is where I'm at now.

[lilshawn]

the chip is performing a preliminary cleaning of the rotation signal from the x and y sensors and transforming it to a straight digital signal... a kind of digital signal conditioning if you will.

you will just need to follow or "buzz out" with a continuity meter... the signal from the X receivers to the chip. the output side of the chip is going to be the x direction. once youve found where the signal goes in, it's easy enough to look on the datasheet to find out where it comes out. then buzz the output pin of the chip and find the appropriate pin in the connector.

[hsr]

Thank you again.

Yes, looking at the data sheet, I can see which in is related to which out.  At least I think it makes sense.

Which leg of the signal receiver should I be trying to trace?  There are three.  On the mouse hack I did many, many years ago, I used the outer two of the three as the signal+ and signal- from the sensor board (they were replacements bought on BYOAC in this thread for a spinner).

[lilshawn]

one of the legs is going to be a power feed, then the other 2 are going to be the signal (the slots on the encoder wheel are  cut in such a way in that one of the signals gets turned on slight before the other.) it's basically a transistor. It just uses light to turn on the transistor instead of a 2nd voltage.

you would hook one to xa and the other xb and then ya and yb. ground and 5 volts come from the board to power everything. when you move the trackball the signal oscillates from xa to xb over and over. this gets translated to movement.

that said, you will not want to take signals from those test/solder pads. you need the chip to do it's magic to the raw signal. find where the x signal goes into the chip and then comes out. it should be easy enough to find what pin in the connector it is. once you've figured that all out you can hook it up and test it. if you end up with a mouse cursor that is opposite (up goes down) then just reverse the a/b signal wires and it'll go the correct direction.

[hsr]

Ok, the power feed was obvious.

The other two leads of both X and Y are both connected to pins 1 and 7 on the IC.

The X direction at 1.6k ohm at pin 1 and again both to the number 7 pin on the IC which is the vdc- at 1.6k ohm also. 
The X direction at 4.7k ohm at pin 1 and again both to the number 7 pin on the IC which is the vdc- at 4.7k ohm also.

I'm lost (again)...

Pin 1's output is pin 2.

[ark_ader]

I have one of those (I took off the plastic shield) nice trackball but those buttons get in the way...

[lilshawn]

y looks like it goes into pins 3 and 5 on the IC... this means the outputs for those 2 channels are 4 and 6. you'll have to find out which pins on the connector are connected to 4 and 6 on the chip. this will be your ya and yb

I can't see from the photos where the x goes to. it disappears under the chip.

the reciever channels ya yb look like they run through r3/r4 then c3/c4 then Q3/Q4 finally r8/r9 to the chip.

the same scheme is followed through the x channels. test for continuity from R6/R7 to which pins on the IC.

[hsr]

y looks like it goes into pins 3 and 5 on the IC... this means the outputs for those 2 channels are 4 and 6. you'll have to find out which pins on the connector are connected to 4 and 6 on the chip. this will be your ya and yb

I can't see from the photos where the x goes to. it disappears under the chip.

the reciever channels ya yb look like they run through r3/r4 then c3/c4 then Q3/Q4 finally r8/r9 to the chip.

the same scheme is followed through the x channels. test for continuity from R6/R7 to which pins on the IC.

Shawn,

Thank you so much for bearing with me!

Following what you said above, I went back and checked again.  In the diagram I drew above, I had #2 (leg on the IC) listed as the upper right pin in the RJ45.  That was a mistake, it's actually #6 (leg on the IC).  6 (output) corresponds to 5 (input) and 4 corresponds to 3.  Those are the Y signals.  Tracing the leads on the printed circuit off the X (the way I think you did it), I find that legs 8 and 10 from the IC correspond to 9 and 11 (inputs on the IC) and would be the X signals.

So these are my six connections (the other two being vdc+ and vdc-).

As you said before, once I wire this into an Opti-wiz, if the direction is reversed, I should just be able to swap the outputs from their respective axis (Xa and Xb to Xb and Xa), is that what you meant?

I really can't thank you enough!

[lilshawn]

if the direction is reversed, I should just be able to swap the outputs from their respective axis (Xa and Xb to Xb and Xa), is that what you meant?

yes.