Run a stepper from 2 DM542 drive

[gabi68]

Woa. The forum system didn't send any emails regarding a new post on my thread. Thank you everybody for your opinions. Regarding the accuracy, a step or two will not harm (this is a fiber laser not milling machine) the focus is forgiving, but a disabled driver will cause the Z axis to fall, not good.
What is the best solution for this situation? Anything already made it out there?

Thank you
Gabi

[dazp1976]

Woa. The forum system didn't send any emails regarding a new post on my thread. Thank you everybody for your opinions. Regarding the accuracy, a step or two will not harm (this is a fiber laser not milling machine) the focus is forgiving, but a disabled driver will cause the Z axis to fall, not good.
What is the best solution for this situation? Anything already made it out there?
Thank you
Gabi

A motor with a brake then?.
Engage / disengage brake as you switch?.
I'm prob clueless and talking crap though.

[A_Camera]

Can you be more specific? A diagram will be awesome. Thank you

I am sorry but I have no diagram. Anyway, it is a simple circuit with one switch and a relay which flips the step, dir, en and gnd signals, all four at the flip of the switch. The circuit is really simple.

I made a simple sketch and scanned it for clarity.



UCCNC generates the pulses if I flip the switch and select it, or the Arduino box if I select the lathe mode. The servo is a servo with integrated driver, but it could be a DM542 also, it makes no difference as long as the step, dir, en and the gnd can be used to control it. The relay in my case is one I had in my drawer, if you can't find one which switches all the four signals at once, you can use two, or even four separate relays as long as you switch all at the same time. Just connect the coils in parallel and you'll be fine. Of course, as I said before, it assumes that both controlling devices work independently and you don't need to rely on exact positions between each devices. In my case, I don't care where the motor stops, the lathe accelerates and decelerates to a speed I want it to, and if I use UCCNC then I have to zero the position for UCCNC, so the position is lost every time I flip the switch, but for me that's perfectly fine.

I don't know if you have seen this before, but I made a video about the control box. This is an early version, since then I swapped the stepper to a servo and built my lathe, which now can spin considerably faster.

[gabi68]

What kind of relay? A link, please?

[A_Camera]

What kind of relay? A link, please?

I have no link. As I said, I had it in my drawer. You have to Google.

[Doddy]

Ignoring the behaviour of the energised coils causing back-emf when disconnected (Jazz's reply) - there's another issue at play here - the stepper drivers would have their own instance of the X/Y coil phasing - swapping between controllers would introduce a discrete change in motor armature position - which I would expect is not desired. That and blowing up the drivers.

As previously explained - use a control system that allows the step/dir to be multiplexed. If using open-collector drives it's possible to wire-or the cathodes on the driver step/dir inputs.

[BeagleBrainz]

If you disable the drive before switching the switching the STEP & DIR it may help with not generating any spurious steps. Just don't forget to re-enable the drive before kicking off again.

[Doddy]

If you disable the drive before switching the switching the STEP & DIR it may help with not generating any spurious steps. Just don't forget to re-enable the drive before kicking off again.

Not intended to drag this conversation onwards, but for clarity on the behaviour of a stepper driver:

A stepper driver can be described as a finite-state machine. It will present a series of currents to each of the two (e.g.) coils according to the number of micro-steps selected. The following is an extract from the data sheet for the TB6600 - though the principle applies to any stepper driver.



You should be able to see that for a 1/8 micro-step, there's 32 discrete states for the A/B coil currents. Whether you move CW or CCW, you're just stepping either forwards or backwards in the state-machine ("phase", in my original post).

My point?, if using two stepper drivers, each will maintain it's own internal state as its operated. If you swap the outputs from one driver to another - unless there's the random fluke of each driver having the same state, there WILL be a discrete change in the coil currents appropriate to the different phase in the state machine that WILL result in a discrete change in shaft position. There's no mechanism available to synchronise the drivers. Disabling the drivers before switching the STEP/DIR will have no impact on this.

A_Camera's point of switching grounds is valid - or my lazy assumption that you share a common ground.

[BeagleBrainz]

Not intended to drag this conversation onwards, but for clarity on the behaviour of a stepper driver:

A stepper driver can be described as a finite-state machine. It will present a series of currents to each of the two (e.g.) coils according to the number of micro-steps selected. The following is an extract from the data sheet for the TB6600 - though the principle applies to any stepper driver.



You should be able to see that for a 1/8 micro-step, there's 32 discrete states for the A/B coil currents. Whether you move CW or CCW, you're just stepping either forwards or backwards in the state-machine ("phase", in my original post).

My point?, if using two stepper drivers, each will maintain it's own internal state as its operated. If you swap the outputs from one driver to another - unless there's the random fluke of each driver having the same state, there WILL be a discrete change in the coil currents appropriate to the different phase in the state machine that WILL result in a discrete change in shaft position. There's no mechanism available to synchronise the drivers. Disabling the drivers before switching the STEP/DIR will have no impact on this.

A_Camera's point of switching grounds is valid - or my lazy assumption that you share a common ground.

Very good points.

As long as the outputs driving the STEP & DIR are at a logic LOW (assuming common) is ground there should be an issue, but if the outputs are at a logic HI and using a mechanical switch to change over disabling the drive would prevent spurious steps, due to bounce.

All in all a better mechanical design would help.

Actually it would be worthwhile knowing the circuitry that drives the STEP & DIRECTION, before really delving in too deep. Considering that the design requires switching 2 different sources for driving the stepper who knows how the electrical side has been done.

[gabi68]

The fiber laser has a board called DLC2-M4-2D which is a 2 board connected with pins. The top board is a controller for X,Y,Z,A axis (can drive a Z axis, a rotary axis, an XY table and all the combination between). The bottom board is dedicated to run the laser head itself (2 small servo motors connected to 2 mirrors to steer the beam and stop it and start it depending the user design).
The top board is a step and dir controller and is connected to a stepper motor which operates the Z axis
I also have a device (with his own stepper driver) which has a measuring sensor for the focal distance (different lens have different focal distance) and this device needs to use the same stepper motor on the z axis.
This was my question: how can I change the drivers (safely) to use the same motor when I need to use one or the other device?

Thank you
Gabi