[MitchellT]

I purchased a used K2 5050x in October. It's a three axis gantry type router with rack and pinon drives. Two servos, the Y and slave A,
move the gantry.

The servos are controlled by G320 Gecko drivers, and have US Digital E5 encoders. I'm a new Mach 3 user, but think I've gotten the config right.

Often, but not always, when the Y axis is heading towards home, the slave A axis will move in the opposite direction. As a result, the gantry racks.

Trying to fix this, I have replaced the computer, and the parallel cable. But it still racks and faults with maddening inconsistency.

As the A axis is the longest cable, could this axis direction reversal be the result of noise from the HEDS based US digital encoder discussed in this forum?

I have ordered the suggested capacitors to de-noise the encoders, but since the axis reversal wasn’t mentioned as a symptom, might there be other culprits to investigate while waiting for Digikey to deliver the caps?

Thanks in advance to all you knowledgeable folks.

Mitchell

[Khalid]

Are you using shielded cables for the encoder? Do you have extended the encoder cables or any joint?

[MitchellT]

Khalid,
It's the original factory cable made up of ...2 twisted pairs for the encoder, running parallel to a second twisted pair motor power feed from the gecko. No shielding on either. I'll double check, but not aware of any joints.

[Khalid]

First Rule... Seperate the power cable from encoder cable...
Second Rule.. Always use twisted pair shielded cables for encoder...
I don't know wether this will help you however can you swap the cable Y-axis with Z or A axis and see if the same problem persists!!!

[ger21]

You might want to measure the voltage of the direction pin at the drive when moving in opposite directions. IT should be 0V in one direction, and 5V in the other.
Sometimes 3.3v ports can cause this.

[MitchellT]

Khalid,
I know if I was building it from scratch, I would follow the rules!...the weird thing is, K2 CNC has been building these machines for over a decade. When I called them, they said they never heard of this problem, and had no idea of what would cause it.
I was wondering if the encoder noise problem discussed here by Gecko president Mariss Freimanis http://www.cnczone.com/forums/gecko_...osition-2.html
might be the problem....although it doesn't say this noise could cause the problem I'm having.
Anyway, I have ordered the parts, hoping I don't have to rewire the thing....I will let you know.

[MitchellT]

Gerry,
I will check it out. How could I fix this? do I need to look for an even older computer?
Thanks,

[ger21]

You can use a buffered breakout board, or an external motion controller that sends 5V signals.

[Khalid]

Dont order the parts.. let us help you troubleshoot with existing hardware... Sometime loose wiring may cause such issues.. Can you show us the pictures inside control box.. I want to check how they they combine the ground wiring..

[Khalid]

Do you use Mach3 software.. If yes, can you check backlash enable for both A and Z axis? or only Aaxis or Zaxis..

[MitchellT]

Hi Khalid,
thank you for your kind willingness to help me. Attached are the photos you requested.
I have already ordered the inexpensive capacitors and resistors to see if they help...I am delaying on the more expensive breakout board or motion control boards pending more testing later this week.
Just to be clear, the A axis is a slave of the Y axis. The A axis is the furthest away and giving the problem.
Thanks again,
MitchellAttachment 215390Attachment 215392Attachment 215394Attachment 215396Attachment 215398

[bhurts]

Check to see if the slave axis home switch is stuck or misbehaving. Just a suggestion but if it is it could try to move away from home before the y ever makes it there.

Ben

[MitchellT]

Hi Ben,
I will double check tomorrow...it seems to be ok....Y homes first, then A. Both back off fine. The A axis always moves away from the home switch when its miss behaving.
Thanks for the hint,
Mitchell