Problem: Stepper motors not moving about fluidnc HOT 43 OPEN

The bottom line is that we probably can't give you a definitive "do this and it should work" answer. You might get lucky and try something that we or you think of, but don't expect anyone to just know. For myself, I have quit using 5160s. The 5160 setups I have tried are just too prone to shutting down in the middle of a job. As an experienced electrical engineer, I probably would have been able to solve the problems given enough time and motivation, but for my present use cases, the motivation to spend the time was not present, so I switched to less-troublesome drivers.

from fluidnc.

Many people are very happy with DM5xx series of external drivers from reputable suppliers, but beware of cheap knock-offs.

from fluidnc.

Again, doesn't explain the difference in behavior between NEMA 17 and NEMA 23 motors, which generally have the same steps per revolution.

Well, what does? If I understand it right, the driver limits the current, no matter what motor is connected. For me, if it "works" with the NEMA 17, it should work with the NEMA 23.Maybe not good, but something should happen.

For the record, I just double checked: the X and Y axis have a belt with a pitch of 2mm and a 20 teeth pulley, so 80 steps per mm is correct...

from fluidnc.

I know nothing about that. I get not paid for writing the software.

from fluidnc.

$100 Bounty info at bottom of post.

from fluidnc.

What are the power supply specs?

from fluidnc.

Try with the other drivers, like the X and Y. Also try using FluidTerm to send a motion command like "G1 Z20 F400". If the motor is locked but not moving, that usually means that the driver is not getting STEP pulses or the STEP pulse rate is too fast. I don't think that "too fast" is the problem here, because the acceleration and rates are reasonable, but it is always a good idea to start with slow movements when testing.

from fluidnc.

http://wiki.fluidnc.com/en/support/setup/testing-step-pins

from fluidnc.

What are the power supply specs?

I’m using my bench power supply up to 30v 10amps.

i was possibly thinking maybe it didn’t like the variable power supply so took a 12v 50amp psu and hooked up a 24v buck booster that’s good for 6amps and nothing changed… I do have a 24v psu on the way though.

from fluidnc.

Try with the other drivers, like the X and Y. Also try using FluidTerm to send a motion command like "G1 Z20 F400". If the motor is locked but not moving, that usually means that the driver is not getting STEP pulses or the STEP pulse rate is too fast. I don't think that "too fast" is the problem here, because the acceleration and rates are reasonable, but it is always a good idea to start with slow movements when testing.

I have tried every axis so far they all do the same thing. I’ll try the LED thing with the driver pins when I get back into town but I did try a smaller cheapo stepper motor and it worked fine…

from fluidnc.

Do you mean a cheapo stepper motor or a cheapo motor driver like an A4988?

from fluidnc.

That information (the fact that you tried different axes, and the fact that you got it to work under some circumstances) is very important. Had we know that earlier, that would have eliminated many different possibilities. Don't make us guess.

from fluidnc.

Do you mean a cheapo stepper motor or a cheapo motor driver like an A4988?

Stepper motor

from fluidnc.

That information (the fact that you tried different axes, and the fact that you got it to work under some circumstances) is very important. Had we know that earlier, that would have eliminated many different possibilities. Don't make us guess.

I apologize

from fluidnc.

Would it been fine to set the board to SD and try to control the steppers without using spi? What would cause it to work with a nema 17 motor I believe and not a nema 23?

from fluidnc.

I’m starting to think it’s my PSU my 24v one comes in today but I’ll be out of town still.. I was thinking maybe it’s fighting with the board in some way as I noticed even if I set it to 30v it would only go to 22ish volts.. I guess some bench PSU will only raise the voltage as the request for current increases.

from fluidnc.

Try leaving all chips installed and disconnect all motors, but 1.

The 5160 is notorious for current spikes. I suggest at least 4 amps per motor regardless of current levels in the config file.

from fluidnc.

Try leaving all chips installed and disconnect all motors, but 1.

The 5160 is notorious for current spikes. I suggest at least 4 amps per motor regardless of current levels in the config file.

I have tried one motor with all drives. With the config set to 2amp run will it still allow the spike or should
I bump it.

from fluidnc.

There is no way to predict the spike behavior - or whatever is causing the shutdown - without detailed knowledge of every aspect of the circuit. Everything from the power supply through the wiring through the controller through the connector pins through the stepstick and its MSOFET through the motor wiring and the motors could be implicated in the dynamic behavior. This is the sort of situation that, in a commercial setting, could require days or weeks in the lab with expensive equipment, trained engineers, and perhaps involvement with vendors. 5160s were intended for commercial use by companies that have the resources to tune the circuits and settings for specific product configurations. Putting them on stepsticks and throwing them out into the DIY world has caused no end of trouble. Many people have had problems that were unsolvable with the resources available. The stepstick format is particularly ill-suited for a driver with this voltage and current capability, even if you have it dialed down to a relatively low current. The resistance and inductance of on-board traces and dupont pins can result in not only surges but also voltage dips that can trigger protection circuits.

from fluidnc.

There is no way to predict the spike behavior - or whatever is causing the shutdown - without detailed knowledge of every aspect of the circuit. Everything from the power supply through the wiring through the controller through the connector pins through the stepstick and its MSOFET through the motor wiring and the motors could be implicated in the dynamic behavior. This is the sort of situation that, in a commercial setting, could require days or weeks in the lab with expensive equipment, trained engineers, and perhaps involvement with vendors. 5160s were intended for commercial use by companies that have the resources to tune the circuits and settings for specific product configurations. Putting them on stepsticks and throwing them out into the DIY world has caused no end of trouble. Many people have had problems that were unsolvable with the resources available. The stepstick format is particularly ill-suited for a driver with this voltage and current capability, even if you have it dialed down to a relatively low current. The resistance and inductance of on-board traces and dupont pins can result in not only surges but also voltage dips that can trigger protection circuits.

Yes I understand and seen some comments about this specific step stick. I may just order the adapters to run external drivers and replace the step sticks with some drivers from steppers online

from fluidnc.

Hi, can I ask a picture of the TMC5160 jumpers? Can it be that you are in full SPI mode?

from fluidnc.

If it were in full SPI mode, it would not work with NEMA 17 motors.

from fluidnc.

You're right, i didn't see that information. Next guess, I found a couple of pictures showing that B1 and B2 were inverted on the new BTT 5160T V1.1. Did you try to invert the polarity of one of the coils?

from fluidnc.

Inverting one coil polarity makes the motor move in the opposite direction. Inverting both has no effect - the first first inversion reverses the direction and the second reverses it back to normal.

from fluidnc.

Ok, now it looks like I'm the one who need's help... But hey, quite new to stepper and still learning! ;)

Don't know how relevant this is, but your step per mm seems completely off: I'd go with 640, if all axes have a 5mm pitch screw, and 200 steps per revolution motor running at 16 microsteps. Can it be that you jog one millimeter and it's moving only 0 .125mm?

from fluidnc.

Again, doesn't explain the difference in behavior between NEMA 17 and NEMA 23 motors, which generally have the same steps per revolution.

from fluidnc.

I once had a problem with motors from different manufacturers where the four cables had the same colors as the two motors, but the coils were connected differently. So unplug the plug from the board and measure it.

from fluidnc.

I think it may be a board issue.. I had some a4988 step sticks laying around and they do the same thing..

from fluidnc.

A4988 step sticks require different jumper settings than TMC5160s. If "the same thing" means that NEMA17 works but NEMA23 does not, I suspect that you have the NEMA23s wired wrong. Did you check the coils as BlueOrangeLive suggested?

from fluidnc.

A4988 step sticks require different jumper settings than TMC5160s. If "the same thing" means that NEMA17 works but NEMA23 does not, I suspect that you have the NEMA23s wired wrong. Did you check the coils as BlueOrangeLive suggested?

I did not check them on the nema 17s but yes I did change the jumpers and configuration.

Yes coils have been verified both via ohm meter and jumping the wires and pushing the gantry and checking for resistance.

I’m really stumped I’m not new to this kind of stuff and I was hoping it was some kind of configuration issue.

I even said hell with it and switched pairs around etc and nothing new.

The steppers are free moving but if I jog them they lock up until the jog command is over and they go back to be free moving manually by hand

I ordered some external drivers. I wired up the OEM board back up for the time being.

from fluidnc.

I did a quick test where I swapped motor wires so coil 1 was connected to driver A- B- and coil 2 to driver A+ B+. The motor locked, but when I tried a move, it just shook and did not turn. This is as I expected. It is a 6-pack universal board with A4988 step sticks driving a short-length NEMA23 motor with no load. It spun as expected with the coils connected correctly.

from fluidnc.

I did some DC and AC voltage measurements across the coils. I have $stepping/idle_ms=255 so the motors stay locked. My power supply voltage is 12V, using A4988 drivers.

The DC voltage across a coil is in range from -1.5V to +1.5V depending on the microstep phase. I tested that by going into G91 relative mode then repeatedly issuing G0 Y0.02 command to step by one microstep - since this system has $/axes/y/steps_per_mm=50. For 80 steps/mm the increment would be 0.0125.

I then ran long fast moves while measuring the AC voltage across the coil. During the motion, I saw an AC voltage of 8V. That would depend on the meter; a true RMS meter would probably measure lower.

from fluidnc.

The exact voltages would depend on several factors including driver current settings and coil resistances, but the qualitative behavior should be similar. The motor I tested is rated for 1.8 ohm coil resistance. The A4988 drivers are set for Vref=0.32V with Rsense=0.068R (pre-2017 Pololu md09b), for a current limit of 0.59A. That corresponds to a coil voltage of 1.8R * 0.59A = 1.06V, but that is the RMS AC voltage, so the peak DC voltage would be 1.414 time that, or 1.5VDC, as I measured.

from fluidnc.

I went through again and the coils are being paired accordingly nema 17s work fine but nema 23s arent i noticed when the board is powering up the stepper is locked but then goes limp does that mean they are holding then stopping?

Also even with my idle_ms = 255 the steppers are movable by hand?

from fluidnc.

It could mean that the driver is shutting down to protect itself. Perhaps you could find an alarm pin to probe.

from fluidnc.

It could mean that the driver is shutting down to protect itself. Perhaps you could find an alarm pin to probe.

One thing I do see is people are grounding the clk pin to use internal clock vs external is that needed with this board?

from fluidnc.

I know nothing about that. I get not paid for writing the software.

Update clk pin did nothing but soldering a jumper from the enable pin to gnd to force it to stay enabled worked.. now why isnt the enable pin working?

edit

well i guess the enable pin is working since the nema 17s work but why wont it stay working hmm.. should i jumper all my drivers?

from fluidnc.

Maybe the enable needs to be inverted - but that does not explain the NEMA17's working.

from fluidnc.

Maybe the enable needs to be inverted - but that does not explain the NEMA17's working.

Yeah strange.. Im just going to jumper all my drivers and call it good i guess!

from fluidnc.

Hi again, seems like neither your board nor the drivers have a pull-down resistor on the Enable pin. Floating enable pin leads to strange behavior I guess; had something very similar in full SPI mode with end switch pin left floating. If you add a resistor to the ground rather than a jumper, you'll still be able to use the disable function.

I also have the last FluidNC board for external drivers, which have all the Step/Dir/Enable pulled down and had no issue with the TMC5160 from BTT. By the way, the clk pin should be connected to ground. Again, on the drivers I have, there is a zero ohm resistor connecting the clk to the ground.

from fluidnc.

The enables on a 6 pack universal board are driven from 74AHCT595 chips which have "totem-pole" outputs, and the OE_ pin for those chips is grounded, so pull-down is not appropriate for such a circuit. The '595 chip always drives either solid low or solid high. A pulldown resistor would have no effect other than to waste current in the high state.

The same is true of the 6 pack external board. There is a resistor+LED on the enable signal going to the external driver, but that does not serve as a pulldown, but rather as an indicator. It does not affect the VI characteristics as seen by the external driver - and external driver current-mode signaling is quite different from stepstick voltage-mode signaling.

from fluidnc.

Thanks for the information! The datasheet gives a VCC value between 4.5 and 5.5V for the 74HCT595 used. If I read the schematic correctly, they are supplied with 3.3V depending on the VCC jumper, right?

from fluidnc.

AHCT not HCT. I did an analysis of the electrical characteristics somewhere. Maybe you can find it. The bottom line is that, even though it is not characterized at 5V (since it is intended for replacing 5V TTL), it works just fine at 3.3V.

from fluidnc.