Alright, I thought I'd start a new thread for this since it doesn't really belong with the sidewinder. This one took me only about a days worth of work to get to a working point, where the sidewinder took me several weeks to figure everything out and find the software that would reverse the axis and everything. Also the buttons were much better positioned than the sidewinder where I could desolder and remove them much easier than the sidewinder.
This is the pcb that it started with. It's much longer and narrower than the sidewinder:
That's how it started out with everything in place, and all the wires removed.
Got everything desoldered here:
Everything finally resoldered back in place:
Here I started to make the arms for the ball to do a test, I messed up and got front back confused so left and right was wrong here, so I ultimately had to remove and reposition them. :
This was just before I plugged it in for the first time and tried to test it. I initially wanted to use this kensington turboball ball, but despite my earlier tests after plugging it in, the sensor failed to register the ball. I assume because the ball is translucent. This is when I still had the arm mountings for the ball reversed.
I then switched to a microsoft trackball explorer ball. But again this sensor seemed rather picky, and there were slight dead spots in the movement of the cursor. I tried to reposition the ball a number of times, but nothing seemed to help. I'm guessing the sensor didn't like the translucent nature of the tbe ball either.
Finally I went back to a logitech wireless trackman ball which is what I used in the sidewinder hack. The sensor seemed to really like this ball, and with it it has a realy smooth and detailed action.
So after that it's completely working although still borgified ass ugly looking. I'm forced to leave everything sort of in this position. I think it's much cleaner still than the sidewinder hack, and ultimately it would be much more positionable because the dpi switches are with the left and right mouse buttons instead of being on the back which was difficult to use and to design around.
The action and detail in practice is probably about 10-20% smoother I would say than the sidewinder hack at least. Probably due to the extra 2000 dpi.
I would say the lachesis is a much better fit for a gaming trackball than the sidewinder at least for the electronics. I also like the onboard memory and profile switching much more than the sidewinder. For the bad parts, I think the sensor is much more sensitive to what it will respond to, and I really had to get the ball to a close tolerance to the sensor in order for the action to be 100% smooth. Another negative is that the sensor is placed at the rear of a large pcb, so you need to have a large part sticking out in front of the ball which will make it look fairly awkward unless you come up with some sort of covering for it that will make it look ok.
The only functional part left to do is to come up with some sort of cap system for the secondary keys. Currently they're just raw switches and not very useful. They may also need to be repositioned.
All in all I think it's going to be hard, although somewhat more possible to make it look better than the sidewinder. I think it's pretty cool either way since it's probably the first 4000dpi trackball in existence.