[Smertrios]

I bought a servo system on ebay that uses gecko drives and at first seemed to be a great system... use more than 1 axis at a time and its "twitchy" and sometimes its twitchy with just 1 axis running but the twitchyness is a different motor than the one that should be moving.

This is what I believe is happening. Current is flowing through 1 wire in the cable which of course produces a magnetic field. The magnetic field induces a current in nearby wires (IE causing "twitchyness"). I already asked for a refund on that system as there is nothing I can do about this. Just posting here in the hopes there may be an easy fix before I ship it back.

[fkaCarel]

Current is flowing through 2 wires. It could also be a power supply voltage drop, causing the servo to lose torque, losing position, correcting this, taking current from the other motor, stop using current, the other motor using now more power and the cycle goes over and over again. The behaviour of a servo system is not the same as a stepper system.

Carel

[Smertrios]

Does that explain why when I make the X motor turn the Y which shouldn't be getting any signal to move would start twitching? I believe its the induced currents on the Y motors dir and step wires by the X motors dir and step signa that is causing it... either way this system is not working right.

[fkaCarel]

An idle motor looks like an idle motor. In fact it's a juggler. It's a process that can be disturbed. Just retyping: if you want calculate 20000 times a second:
Motor command = ((Gain / 4) * Position now) - ((Zero filter / 256)*(Gain / 4) * Position Last Time) + ((Pole / 256) * Motor Command Last Time) it would be fine if all the other conditions were a little stable isn't it?

Carel

[Smertrios]

when I say twitching I'm not talking about a condition like oscillating where it is moving back and fourth from the position it should be in I mean its actually "twitching" in 1 direction actually changing the position it should be at. IE its turning but not smoothy its "twitching" and should not be. I guess you would have to see it to understand whats happening it must not be common... by chance are the cables on your system shielded? These are not

[fkaCarel]

I know what twitching means. It's what my eye does everytime I see a post from you. But now serious: assembling components of various components makes you a system integrator. You have to decide where the error is, or that you put the error in by mismatching components, or worse put in an error over an error. I can't see nor measure it from here. About shielding: the A & B signals from the encoders are signals going one way (encoder ---> driver) and their phases are apart, there is always 1 signal (out of 2) going up or down. I have seen that if they are not at least separated from each other, they drive each other. You can buy round shielded cable, where all the wires are shielded. If all you have is flatcable you can make a shielded cable doing this:

Vcc-Gnd-A-Gnd-B-Gnd, it works.

Carel

[StealthDumpKits]

I guess nobody uses scopes anymore.

[fkaCarel]

Scope? Way too fancy. There is also a chance of breaking the law of permanency of problems by using it. It's beaten by world wide guessing games.

Carel