Driver help

[johngoodrich]

hi all, im after a little help. i recently upgraded to uccnc and two motors on the x axis. i cannot get movement from them. my y and z axis work perfectly. i am using a uc300 eth, with a stv2 breakout board. i have two matching drivers:
https://www.omc-stepperonline.com/digital-stepper-driver/y-series-digital-stepper-driver-2-4-7-2a-ac18v-80v-dc-36v-110v-for-nema-34-motor.html
29658
everything seems to be right, but i get no movement. i am getting the correct voltage to the drivers, and the dro's move correctly. all the settings seem correct and the motors lock when powered on but still no movement.
I'm at the point of pulling my hair out now, please help
29663

29659296602966129662

[Voicecoil]

Do you have a diagram of how you've wired these drivers? - it might help find your problem. I also notice on the first screen that there's no enable pin listed, could this be an issue?

[johngoodrich]

yep im not using enable. maybe i should, im not using it on my y and z and they work. i have x step pin 2 to step- on the driver x dir pin 3 to dir- on driver and +5v to step+ and dir+. motors are 8wire 4nm nema23 from cnc4pc which are wired bipolar parallel

[Voicecoil]

Looking at the manual it seems that the MF input on the driver is more of a disable than enable function, so maybe that's not an issue. On your screenshots I see that the B axis DIR is ticked as low active, whereas X axis isn't and neither of the STEP outputs are set as low active either - have you tried setting these?

[johngoodrich]

Had a play around this afternoon and I have my x and b axis working. I set both to active low step and all of a sudden they work. Unfortunately my y axis now doesn't, even though it did yesterday.
No settings have changed, I checked all the wires are still connected, but I just get a faint click when I try to jog the y.

[Muzzer]

Have you checked the minimum pulse width that is outputted to the drive? Too narrow / short and the drive's input circuit may not be fast enough to see it. Another related parameter might be the maximum pulse frequency - similar issue if you have this set too high, may not be "seen" by the drive.

[johngoodrich]

Yep checked the pulse width, all seems ok. I am thinking its the bob as I can the z axis wiring in to the y drive and it works, but when I plug the y wiring in to the z drive it doesn't work. Can anyone recommend a good breakout board to use with the uc300 etc. I would like the ub1 but can't justify the cost of it at the moment

[Doddy]

John, I've used the generic Chinese BoBs with opto-inputs - got two stacked on a UC300 without any fuss at all. Your experience might vary.

[johngoodrich]

John, I've used the generic Chinese BoBs with opto-inputs - got two stacked on a UC300 without any fuss at all. Your experience might vary.

Have you got a link to ones you have please?

[Doddy]

eBay item code 373449654531 - I bought a batch of 5 of those recently - all a generic design (HCT245 output drivers, Opto inputs driving a HCT14 Schmitt trigger input) - just generally a cheap workable design.

The only thing is you'd need a DB25-26W IDC header cable

[johngoodrich]

Thats exactly what I have now. Seems to be ok except I think the y axis has blown. Im going to try using the b axis output this afternoon, and If that works ill buy another couple

[Doddy]

It's certainly possible to blow a single output - though a little unusual. Definitely worth substituting B - yup. If you still have a problem then think of the signal chain (for example, you could use another channel on the UC300 - just remember not all are the same, but there's at least 2 with the 11-outputs/5 input configuration.

[johngoodrich]

Yep I tried changing from port 2 to 3 but no difference

[Doddy]

I've just read the thread through - yeah, try changing the outputs still, but keep in mind the stepper drivers. In a pinch, I'd try disconnecting the step/pulse input to the driver and just flashing 5v across it (make/break with a bit of stripped wire) - all correct polarity of course. Quick way to check that the driver hasn't been borked along the way. Zero-cost, removes a component from the diagnostics.

[johngoodrich]

I used z wiring on the y driver and it works so definately the bob

[johngoodrich]

Well I've sorted the issue. Turns out the new cat 5 cable i used to wire the drivers had a break it it somewhere. Used a new bit of cable and success. Doddy, do you use the screw terminals or do you have connectors to plug in to the board?

[Doddy]

I tend to use the screw terminals (with ferrules), bit daft since I have the connectors (I think). But screw-terminals allow you to maintain things with a banana, at a push :)

[johngoodrich]

Do you know what type of connectors they are?

[Doddy]

Do you know what type of connectors they are?

Well... I have JST-XH's in my purchase history.... and something visually similar in the shed that mates with the board... Yeah, I'd go with XHs.

[Kitwn]

Well I've sorted the issue. Turns out the new cat 5 cable i used to wire the drivers had a break it it somewhere. Used a new bit of cable and success. Doddy, do you use the screw terminals or do you have connectors to plug in to the board?

That rings a bell, I'm sure someone else had a similar problem a while back. Was it monofilament cable?

I've got one of these which turns a laptop into a two channel oscilloscope and has proved invaluable for tracing pulsed signals and would have found your broken wire very quickly. No doubt there are those who will say it's too complex a bit of gear for some people to use but in trying to fault-find on a CNC router you are already dealing with a complex bit of gear and need the right tools to show you what's going on. It's part of my standard electronics tool kit along with my multimeter and soldering iron. Is it expensive? Depends on how much money you would have wasted on replacing BOBs, drivers and other components that were never faulty.

Kit

https://au.element14.com/pico-technology/picoscope-2204a/oscilloscope-pc-10mhz-with-awg/dp/236542501

[Doddy]

No doubt there are those who will say it's too complex a bit of gear for some people to use

I agree 100% with the sentiment of your post, but you hit the nail on the head here. Not that, as a tool, it's complex - yes they can be (and heaven forbid you ever try a LeCroy which - despite warnings - I bought one through the company and hit myself every sodding day for the reversed timebase/Y-deflection controls (everyone warns you about it - but you think you'll get used... you don't!) The problem is that a lot of people know what signalling is expected, and a scope would work for them - but even without a scope knowing what to expect allows a whole load of diagnostics with nothing more than a meter, LEDs, bits of wire and a cup of coffee with a hobnob.

For those that don't understand the signalling, a scope would just be a gateway to the unknown.

Perhaps we need a sticky for the typical signalling across a CNC system and easy approaches to test each node. But that assumes a convention to the controller design. There's no easy win to this.

It's frustrating as hell to try to help someone a couple of hundred miles away over a few days or weeks when you realise that in the long run it'd be quicker to just jump in the car and fix the problem yourself. No disrespect to anyone intended - we all have our own areas of interest/experience.

[Kitwn]

I agree 100% with the sentiment of your post, but you hit the nail on the head here. Not that, as a tool, it's complex - yes they can be (and heaven forbid you ever try a LeCroy which - despite warnings - I bought one through the company and hit myself every sodding day for the reversed timebase/Y-deflection controls (everyone warns you about it - but you think you'll get used... you don't!) The problem is that a lot of people know what signalling is expected, and a scope would work for them - but even without a scope knowing what to expect allows a whole load of diagnostics with nothing more than a meter, LEDs, bits of wire and a cup of coffee with a hobnob.

For those that don't understand the signalling, a scope would just be a gateway to the unknown.

Perhaps we need a sticky for the typical signalling across a CNC system and easy approaches to test each node. But that assumes a convention to the controller design. There's no easy win to this.

It's frustrating as hell to try to help someone a couple of hundred miles away over a few days or weeks when you realise that in the long run it'd be quicker to just jump in the car and fix the problem yourself. No disrespect to anyone intended - we all have our own areas of interest/experience.

And I agree 100% with yours! Everything is easy when you know how, that's why we employ specialists to perform specific tasks. I wouldn't trust a brain surgeon to service my car or a car mechanic to sort my head out. Actually I don't think there's anyone clever enough to do that, but I digress. One of the key elements of a hobby like ours is attempting to become an expert in several different and complicated disciplines in next to no time and with no formal training.

Knowing how to trace faults in electronic equipment is a skill in itself. I have followed some of your past efforts to help others at a distance and am genuinely impressed at your patience and perserverance in the face of dificulties. I do like the idea of a sticky thread covering some of these issues. What the various signals consist of, how you can trace them through the equipment with simple, inexpensive test gear and what you can and cannot properly measure with an ordinary digital mulitimeter (the one you simply cannot do without if you expect to build you own machine) would be very useful to many.

Kit

P.S. We spent today at Steamfest in Sheffield. One of the junk stalls (sorry, "vintage tools emporium" ) had a lovely old bakelite AVO meter for $80. I was very tempted. It was the same as the very first piece of test equipment the BBC trained me to use when I was a lad.