Gantry CNC - Dual Y's - Missing Steps

For over a month I have been trying to figure out why I'm missing steps on my machine.

A few days ago, someone looked at photos of my controller and after a few minutes said, "I know what your problem is!" You should have inverted your stepper pins in StepConf

I was elated! Finally I had an answer. So, I built a new profile and made the appropriate .hal changes needed to invert the pins.

When I homed the machine, wow... I was sorry I did that. Everything went backwards. Fine, I changed the sign on the .ini homing Velocity. I homed again. This time I took out all of my home switches. I didn't know it was possible to bend the springs arms on the switches backwards

I finally replaced all my switches with Prox Sensors and am now having homing issues, but I will get that ironed out.
My question here is this:

How would I have known to invert the stepper pins on the Parallel board? What's the clue?
More, what did this person see that gave him the clue that I had this wrong?

I'm using DM542 stepper drivers with a standard parallel bob

I have searched, and searched and can't find any answer to this question. Just help please...

Thanks so much

Are you still missing steps under the same load conditions ?

What is your backlash situation ?

Looking at the manual there is a jumper that seems to set whether a step is on the rising or falling edge. Technically speaking there should be no reason why inverting your step pins would prevent missing steps.

Why don't you ask this "sage" why ? They were the one that gave you that advice. Surely they should be able to explain why.

A common cause of missing steps is the wrong timing values (too short), overloading the steppers or poor quality couplers between steppers & leadscrew. The cheap chinese spiral cut couplers will twist due to the nature of their design. Good quality (expensive) oldham couplers are quite reliable, after 5 years I've had no issues with the ones I use.
A leadscrew that is not rigidly axially constrained will cause issues as well, which could look like missing steps. Also the use of plain bearings (eg a bearing that is used for skateboard or motorcycle wheels), as opposed to angular contact bearings, will eventually allow axial movement of the lead screw.

Yes, there is no technical explanation why inverting the axis should solve the problem. Of course its a try, but result is that something wrongly connected.
I agree to BeagleBrainz.
Timing problem of ini (DIRSETUP DIRHOLD STEPLEN STEPSPACE STEP_SCALE MAX_VELOCITY MAX_ACCELERATION )
or unstable / unadjustet mechanic.
Note that steppers lost steps over 500rpm.

All of your answers have been very helpful

Thank you

Back to the shop

I will update this thread when I figure out what the problem was

tommylight wrote:

.....short of being on LSD.....

I know where my crystal ball went now! I'd like it back though

So, I've been wracking my brain to find out why I'm missing these steps on Y2. Today I did the Linuxcnc stepper test over and over again. Y2 ends up being over .5" Usually by about a 16th of an inch.

I got to looking at my grounding on the system and wonder if this might be the problem. I have a Cisco power supply powering the drivers. Where the 2 5vdc outputs are on the BOB, the ground next to it has been tied to the Cisco ground as well. Is that the right way to do this?

That 5vdc ground ties to the usb connector ground on the board, of course going to the computer ground.

On the DM542 there is only one ground connection and it is the high voltage for the steppers. There is no ground for the 5vdc, so I'm thinking they get tied together. Am I correct in my thinking?

The usual way is for the inputs to the steppers to be optocouplers, this is to provide isolation between the high voltage & noisy stepper PSU and the logic power supply.
By connecting the 5vdc & High voltage PSU grounds together you lose that isolation.

Electrically your logic (BoB) and stepper psu (high volatge) should not have a common ground.

Are your Y axis driven by belts or lead screws ?


My money is on a mechanical problem.
Do you have any photos of your Y axis that show some detail ?