Finished parts ending like they were offset some how...

Hello guys,

Having big problems with my 4 axes router here, using LinuxCNC version 2.7.14 through LPT port all my parts end up like they were offset some how, checked the information on LinuxCNC site adjusted stepgen parameters according to my Leadshine stepper drivers but nothing. My X axis is gantry type with two motors and two drivers using the 5th. axis for the second motor.
Looks like nobody had this problem any ideas before loosing my hair?
Thank you all!

Hi BigJohnT,
The parts are starting out in the correct place and moving as they are cut. looks like missing steps but did all that manual and forums say about that.
Going to replace the PC to see if the problem is hardware related.??

1st thing to do is to make sure your parts are in fact not moving on the machine. This is a common problem on vacuum clamped router tables. Especially if the vacuum supply is marginal. But even with very good vacuum small parts can still not be held well and move.

2nd if the parts are not moving, then the machine is loosing position for some reason. Check all the usual suspects, missed steps/stalls of the stepper motor, slipping couplers, belts jumping teeth...

There are lots of reasons for a step motor to stall or miss steps, mechanical problems causing binding (worn/damaged/maladjusted ways, worn/damaged/maladjusted lead or ball screws, lack of lubrication), worn out step motors, motor wiring issues, control wiring issues, control configuration issues (bad timing...)

Machine mechanics and electrical are ok, the machine used to run with Mach 3 and ethernet smooth stepper before with out this problem, the system was moved to LinuxCNC due that the ESS stop working.

Machine has home switches and making the test with a gauge the X axis end up 0.01 offset after 1000 moves back and forward. of course after this doing G53 X0 ends up wrong.

Could one of those blue Chinese BOB cause something like that?

Large latency spikes could cause missed steps by causing a velocity step larger than the motor can handle.

Step timings set too close to the driver's minimums (slow optos on a cheep breakout board can effect pulse lengths), too high acceleration, too high velocity, or electical interference can all cause missed steps.

The computer is an Intel core 2 duo with 1GB RAM not sure if this is a good configuration, I really do have a lot of distrust on the cheap BOB.
Already tried reducing acceleration and velocity at very low speeds but no luck with that also the problem keeps on the X axis can an opto be different than the rest?