Losing steps on Z axis

[Arzo10]

Ok so, a couple of weeks ago i changed my stepper driver to a totally different one that was in it.

Yesterday I tried doing a halftoner tool path (loads of z movements)
Halfway through i notice hey this looks abit deep here. So i checked and i had lost like 18mm. So ok i zero and try again. Same thing happens.

Run a profile cut today on some small items and i lost 2mm

I have tried decreasing acceleration and velocity on the z axis in Mach 3
This didn't work.

Checked all set screws and bolts on z everything is tight.

Is this just totally the wrong driver or what is happening here it is driving me bonkers please help.

[Clive S]

Ok so, a couple of weeks ago i changed my stepper driver to a totally different one that was in it.

Yesterday I tried doing a halftoner tool path (loads of z movements)
Halfway through i notice hey this looks abit deep here. So i checked and i had lost like 18mm. So ok i zero and try again. Same thing happens.

Run a profile cut today on some small items and i lost 2mm

I have tried decreasing acceleration and velocity on the z axis in Mach 3
This didn't work.

Checked all set screws and bolts on z everything is tight.

Is this just totally the wrong driver or what is happening here it is driving me bonkers please help.

I think you need to give more info ie links to drives and motors etc pics would help as well

[Neale]

Had a similar problem when I set up my new router. Checked all mechanical components, made sure no slippage anywhere. Still lost z position. Not sure how your drivers are connected but I would double-check polarity of step signals between BOB and driver. When you changed the driver, did you also have to make any changes in Mach3 (or other motion control software) - like step signal polarity or z direction? That was my problem - driver is triggering on wrong edge of step pulse and the end result is that you lose one step every time you change direction. Doesn't really show up on a simple xy profile cut but once you get a lot of z movement, that's when you see the effects. If that is the problem, fix is very easy.

So, how's the driver connected, and did you change anything when you swapped drivers? What were old and new drivers?

[m_c]

As Neale says, it does sound like a step/direction timing issue.

What make/model are the old and new drivers?
What controller are you using?
If you're not sure, post up some pictures.

[Arzo10]

Had a similar problem when I set up my new router. Checked all mechanical components, made sure no slippage anywhere. Still lost z position. Not sure how your drivers are connected but I would double-check polarity of step signals between BOB and driver. When you changed the driver, did you also have to make any changes in Mach3 (or other motion control software) - like step signal polarity or z direction? That was my problem - driver is triggering on wrong edge of step pulse and the end result is that you lose one step every time you change direction. Doesn't really show up on a simple xy profile cut but once you get a lot of z movement, that's when you see the effects. If that is the problem, fix is very easy.

So, how's the driver connected, and did you change anything when you swapped drivers? What were old and new drivers?

Hello, yes i had to change the direction​ when i put the driver in the previous was a leadshine DM856 and the new one i put in was one from cnc4you
Cwd556. Now what i will say is I did have to change the steps per unit but nothing else was changed. From changing the driver what i did notice was the sound of the z going up and down was considerably louder?

[Neale]

I understand that you are using a CSMIO motion controller. In all other respects, this is a good thing - I use one myself! However, assuming that it uses differential signalling to the driver (which is also a good thing) then it is very easy to get, literally, your wires crossed. Quick way to check is to go into Mach3 -> ports and pins -> motor output and click the Step Active Low box (changes from red cross to green tick). This is exactly equivalent to swapping the wires between CSMIO and the driver. Just do this for Z, if the other two axes are OK. Then do another test.

This is my red-face moment - despite working very carefully, when I wired my control box I managed to get these exact same connections swapped myself. However, as I say, if this is the problem (no guarantees but it sounds very familiar) then this will fix it. Then you can go back and recheck your "steps per" settings as these might have been affected by the problem. If this is not the problem, then nothing bad will happen and you can put the Mach3 setting back as it was and look further for the problem. However, this is so easy to do and check that it's worth taking a few minutes to try it first.

[Arzo10]

I understand that you are using a CSMIO motion controller. In all other respects, this is a good thing - I use one myself! However, assuming that it uses differential signalling to the driver (which is also a good thing) then it is very easy to get, literally, your wires crossed. Quick way to check is to go into Mach3 -> ports and pins -> motor output and click the Step Active Low box (changes from red cross to green tick). This is exactly equivalent to swapping the wires between CSMIO and the driver. Just do this for Z, if the other two axes are OK. Then do another test.

This is my red-face moment - despite working very carefully, when I wired my control box I managed to get these exact same connections swapped myself. However, as I say, if this is the problem (no guarantees but it sounds very familiar) then this will fix it. Then you can go back and recheck your "steps per" settings as these might have been affected by the problem. If this is not the problem, then nothing bad will happen and you can put the Mach3 setting back as it was and look further for the problem. However, this is so easy to do and check that it's worth taking a few minutes to try it first.

This seems to have fixed the problem :-) thanks alot

[A_Camera]

Hello, yes i had to change the direction​ when i put the driver in the previous was a leadshine DM856 and the new one i put in was one from cnc4you
Cwd556. Now what i will say is I did have to change the steps per unit but nothing else was changed. From changing the driver what i did notice was the sound of the z going up and down was considerably louder?

I don't know the DM856 or the Cwd556, but if you are using CSMIO then I think one problem can be the pulsing frequency. If I am not wrong, CSMIO is capable of running much faster than 200kHz and most drivers can only handle maximum 200kHz pulsing. Perhaps that was not an issue with the DM856, but may be an issue with the Cwd556. I know I can run my DQ542MA at 400kHz, even though they are also only for 200kHz, but since I occasionally experienced some issues when I ran them outside the specs, I reduced back to 200kHz and my problems were gone. So, have a look at your pulsing frequency and reduce to 200kHz if set higher. Also check the specs of Cwd556 regarding this, and keep your pulsing at or below that level.

[Neale]

I don't know the DM856 or the Cwd556, but if you are using CSMIO then I think one problem can be the pulsing frequency. If I am not wrong, CSMIO is capable of running much faster than 200kHz and most drivers can only handle maximum 200kHz pulsing. Perhaps that was not an issue with the DM856, but may be an issue with the Cwd556. I know I can run my DQ542MA at 400kHz, even though they are also only for 200kHz, but since I occasionally experienced some issues when I ran them outside the specs, I reduced back to 200kHz and my problems were gone. So, have a look at your pulsing frequency and reduce to 200kHz if set higher. Also check the specs of Cwd556 regarding this, and keep your pulsing at or below that level.

Although it is capable of high pulse rates, there isn't any control on the CSMIO to limit max pulse speed. However, this doesn't matter. Pulse rate is decided by how fast the motor turns, and how many steps per rev. So, 5000mm/min axis with 5mm lead ballscrew means motor turns at 1000RPM (assuming direct or 1:1 belt drive). 200 steps per rev with 8x microstep setting (pretty typical) gives 3200 microsteps/rev, that is 3200 pulses/rev. So, that's 320000 pulses/min, roughly 5300 pulses/sec. Well within the capabilities of any of the usual drivers. It is possible for pulse rates and wait time between dir and step pulses to be a problem in some systems (e.g. where you can change pulse length) but these are fixed in the CSMIO firmware and seem to work well. I can't think of an obvious reason why you would want to pulse a stepper much faster than this; usually the reason for needing faster hardware is to cope with encoder pulse output from servo systems rather than open-loop steppers.