Oscillating stepper motors

Hi, i'm having a wierd problem with my z axis motors.
First of all my configuration is the following:
i'm running the latest version of linuxcnc (2.7.11), and i'm controlling the machine with a simple parrallel port BOB.
I have 4 nema 23 motors ( X Y Z Z ) with the 2 Z motors sharing the outputs on the breakout board (i know, it's not ideal, but for a 3d printer prototype, even if well built, it's enough for now), powered by some chinese drivers (something like this www.ebay.it/itm/TB6600-Single-Axis-4A-St...6:g:D2MAAOSw-olZ6LX1)

The problem is the following: considering the fact that i need perfect allignment of the z axis gantry i disabled microstepping on those 2 motors, but now i noticed a strange problem.
(apart from being quite stuttery, but that's normal without microstepping) when running at ~ 220mm/min those 2 motors moves back and forth like crazy, they simply vibrate and rotate in both directions randomly.
It's enough to increase or reduce the speed of 30-50 mm/min and they start working as they should.
I tryed changing the steps settings in the hal file (i found a thread in this forum with some suggested values) but i couldn't see any difference,

Do you know how i can fix this?

Thank you for your attention.

If you disable microstepping, its very likely that you will have resonance issues
( and crazy back/for motions at certain speeds sounds likely to be caused by resonance )

I would try 1/4 or 1/8 stepping and see if the problem goes away

The problem actually disappears already with 1/4 microstepping.
Do you think that using that setup could create problems with a gantry configuration?
I read somewhere that microstepping could lead to misalignment as the motor can skip to the following / previous step.

Ok, that's not too bad.
My only concern was that the errors could add up, but thinking better at how stepper works that should not happen (unless they loose steps for eccessive load).
Even if one of the motors jumps at the following step instead of making partial steps, the error should be canceled when the other motor also reaches the end of the full step, shouldn't It?

When a step motor losses steps, it loose one full step or more, and it will not self correct. This should not be confused with position inaccuracy of a micro-stepped step motor, Micro-steps will self correct up to the point where the step motor is pushed over the "hill" into the next full step "valley".

Ok i got it now.
The first time i read about microstepping errors i didn't think at how the motor actually works and i misunderstood it, thinking that it could add up.
I will enable 1/8 microstepping on the z too then.