Re: [CAD_CAM_EDM_DRO] Re: Loosing position and need help...

mmurray701 wrote:

>
>
>>Most likely there is a problem with the timing of the direction
>>
>>
>signal.
>
>
>>If the direction signal changes state at the wrong time relative
>>
>>
>to the
>
>
>>step pulses, it can move one encoder count in the wrong direction.
>>If this only happens occasionally, it will accumulate errors at a
>>
>>
>very slow
>
>
>>rate. I don't know if the step to direction timing is adjustable
>>
>>
>in
>
>
>>TurboCNC.
>>
>>Jon
>>
>>
>
>Jon, can you simplify this for me a little? What can I do to test if
>this is the problem and how do I fix it?
>
>

Not very easily! Do you have an oscilloscope? No? The it will be DARN
hard to be scientific about it. If there is an adjustment in the
software to
change the timing of the step to direction signals, then maybe you can try
changing that until the error goes away. If there is a polarity setting on
the step pulse, then try setting that to the opposite. If the direction
signal
is changing when the drive THINKS it is getting the step command, it can be
confused. If the program is changing direction when it thinks it is BETWEEN
step pulses, but the drive interprets the step signal oppositely, then
you have a
strong possibility of confusion.

>To everyone else, thanks for all the replies. The step rate isnt a
>problem from the computer. I'm actually running 340's to be honest
>because of that reason, to keep the step rate down. Just thought id
>mention using 320s since more people know about them and hoped to
>get more replies.
>
>

Ah, I think the 340 may be more sensitive to this phenomena! I know the
early
ones definitely were! If the direction signal changed during the time
the step
multiplier was still pumping out steps to the basic G320 drive underneath,
the direction of the last few steps could be in the wrong direction.
You might
set the multiplier for X1, or even remove it and see what happened.
Depending on the exact version of the 340 step multiplier, the requirements
for the timing of the direction change relative to the step pulses is
different, I
think. You might lift the cover, get the rev # off the board, and call
Gecko Drives
and verify what the drive actually requires. They can also tell you how
to remove
the step multiplier. I have heard a few people have needed to do this,
and it fixed
this exact problem.

>So where do I go from here? Try running it with a seperate power
>supply for the encoders seems to be the next step. Can this actually
>cause a slight problem like I'm experiance or would they more than
>likely not work at all? Anybody run into this problem before by not
>having a seperate power supply?
>
>

I have a lot of doubt this is your problem, especially now that I know you
have G340s.

Jon