Page 1 of 2

Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Wed Dec 27, 2017 7:01 pm
by nikdan
Hi,

I just bought the license from you and installed my UC300ETH to my bob, a 'CP0-10V' board from cnc4you.co.uk

Everything worked out with the installation of software, connecting through ethernet directly, finding the license and such. Everything boots up and I started doing the axis setup. Using the manual from CNC4YOU I set up the X-axis as on the pictures below.
Nothing happens to the stepper motor and while measuring the output voltages from the BOB I can see it's changing for Port 3 (Dir X) from 0 -> 5V when changing direction with the jogging arrows. The step pin, however, shows no sign of changing voltages.
Looking at the diagnostics I see no sign of life there either for pin 2.

I tried with pins 4 and 6 which should be for Y and Z, nothing.

I also tested the emergency stop for Pin 12 - Port 2, which works.

I have:
- 5V to the UC300eth
- Using port 2
- Using an IDC26-DSUB25 Parallel cable between the UC300eth and CP0-10V
- 15V separately PSU for the CP0-10V

What am I doing wrong? Can I isolate the problem somehow?
Can I measure anywhere to see if the UC300ETH is sending anything to the BOB?

Any help is appreciated!

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Wed Dec 27, 2017 8:37 pm
by cncdrive
What did you use for the measurement? A multimeter? An oscilloscope? A logic analyser?
I'm asking this, because the step signal is a low duty cycle pulse train and so you will not see it if measuring with a multimeter,
because a multimeter is a slow device, it will measure the avarage voltage and since the duty cycle is low the avarage will be also around the 0Volts or 5Volts if it is set active low.
Using an oscilloscope or logic analyser you would see the signals.

One thing to verify is the kernel frequency. Try to set it lower, e.g. start with the lowest 25kHz setting,
because it may be possible that it is now set so high that your BOB or drives cuts the signal off if there is a slow optocoupler or other slow circuitry in them which is usual with chinese BOBs and drives.
So, first check if setting the kernel freq. to 25kHz will make any difference.

And depending on how your BOB handles the enable signal it may be possible that the enable signal when not set cuts the step/dir signals, however I do not know the circuit of your BOB.
So, one other thing to try is to invert the enable signal (with it's active low swtting).
And don't forget to press Apply settings button when making changes and before testing them.

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Wed Dec 27, 2017 9:28 pm
by nikdan
Thanks for your reply!

Yes, I used a multimeter.
Still, using a multimeter, the output on port 2 gave Pin 1 (Enable), 2, 4, 6 and 8 changed varied from 0 to 0.02V, at most while running a program going all axes back and forth. Shouldn't there be any display on the multimeter?

I tried the 25kHz setting and there was some progress.
To test everything I run a program going from X10 to X-10 back and forth. Now the steppers "clicks" and switches between being locked a free at each line. They don't spin, however...
I after that changed the frequency back up towards 400kHz, the same result is now seen on all frequencies.
At diagnostics, I could see very few flashes on pin 2,4,6,8 while running the program. Once or twice per line of code.

Measuring from the CP0-10V output the voltage on the Step pins varied from 2.3V to 2.7V at max, switching at every new line of code. Dir pins changed between 0V and 5V at each line of code.
I've tried changing between Active Low setting, does not work.

Do you have more suggestions on what the problem could be? Should I try Port 3?

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Wed Dec 27, 2017 10:27 pm
by cncdrive
You may catch step pulses on the LED, but because the duty cycle is low it is like 0.001% of time when you can catch it with eyes.
Imagine that with a 400kHz kernel frequency one step pulse's active side is only 1.25 microseconds long.
Let's say you commanding to move a motor which has e.g. for simplicity makes one rotation from 1000 step pulses and you commanding it to rotate with e.g. 100 RPM,
that means 1.666 rotation per second. Which means 1666 pulses per second. This means approximately 2 milliseconds total time per second when the output is turned to the active edge and the rest of the time which is then 998 milliseconds the signal is off.
If you measuring avarage voltage e.g. a multimeter is doing that because it is not a fast device, so then assuming the output active edge is high level 5Volts and so the low level is 0 Volts,
then the avarage will be (5/1000)*2 = 0.01Volts which is 10 millivolts, this is what your multimeter will show for this example measurement.
However there are 1000 step pulses happening on the step output pin, but you will still measure 0.01Volts only which your multimeter could even show to 0V depending on it's resolution and precision and especially because the value is far from the upper range what avarage you will see can vary, but you will see an around 0Volts value for sure.
So, a multimeter is not good to measure this type of signal, because you will not see if there are step signals on the output or not.

And the enable pin state you should see on the output pins using a multimeter.
The enable pin changes state when you press and remove the rest condition in the software.
It should move between 0V and 5V states and the active low checkbox sets if it has 0V or 5V when the reset is removed and ofcourse the opposite state is when the reset is pressed in.

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Wed Dec 27, 2017 10:33 pm
by nikdan
Thanks for the explanation, I really lack knowledge in electronics unfortunately :P

I've now managed fixed so that the motors are spinning, changing kernel frequency and removing the enable cables, both ena+ and ena-, from the driver.
However. It's supposed to go 20 units back and forth, but it never ends up being at the same place each cycle, which it should since I'm not using any backlash settings yet.
You can also hear how the motors don't sound perfect, not sure what it could be.

Any ideas?
How should I start doing diagnostics? Where should I focus? Electronics really isn't my thing...

If I go above 25kHz:
X-axis only spins CW (skips every second line)
Y spins both CW and CCW as it should (still not correctly tho').
Z only goes CW, and when it's supposed to go CCW, it continues CW.

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Thu Dec 28, 2017 10:08 am
by A_Camera
nikdan wrote:Thanks for the explanation, I really lack knowledge in electronics unfortunately :P

I've now managed fixed so that the motors are spinning, changing kernel frequency and removing the enable cables, both ena+ and ena-, from the driver.
However. It's supposed to go 20 units back and forth, but it never ends up being at the same place each cycle, which it should since I'm not using any backlash settings yet.
You can also hear how the motors don't sound perfect, not sure what it could be.

Any ideas?
How should I start doing diagnostics? Where should I focus? Electronics really isn't my thing...

If I go above 25kHz:
X-axis only spins CW (skips every second line)
Y spins both CW and CCW as it should (still not correctly tho').
Z only goes CW, and when it's supposed to go CCW, it continues CW.

Sounds like a mixture of electrical installation problems AND mechanical design/installation problems. Really, it can be anything, maybe even configuration issues. In my opinion you should start at the mechanical part. Check for binding, alignment, squaring and so on. Backlash compensation is not something you should use unless you have MEASURED backlash. Check stepper motion without the screws attached. Is it running smooth and nice this way? If yes, then mechanical issues are definitely a problem. You can run one axis at a time to find out which one is causing issues. Which microstepping values are you using? How does your machine looks like? Describe it and post pictures. Am I understanding you right that you are using 15V power supply? That may be far too low for the steppers.

Really, at this stage, basically anything can be wrong. Don't measure signal levels unless you have instruments to do that with. Some multimeters can do that, some can not. In any case, all multimeters have specifications, read yours to see what it supposed to be able to measure.

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Thu Dec 28, 2017 10:40 am
by nikdan
Hi A_Camera,

The steppers are currently not connected to anything mechanical.
I'm running the BOB with a 15V 1A PSU, the steppers use a separate 48V 10A PSU.

I made a video of my steppers running a program going from X10 Y10 to X-10 Y-10, back and forth, this is the result. One way seams nice, smooth without losing steps.
The other direction, X is slamming quite hard when changing direction and is missing half the motion. Y does not even move.
https://youtu.be/IXhlJX7Hh78

The driver is set up to 10 microsteps per rev. i.e 2000 steps/rev
Steps/unit: 200
Velocity: 2000
Acceleration: 100
The motors are wired Bipolar parallel as recommended by CNC4you

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Thu Dec 28, 2017 10:45 am
by nikdan
Here are a few more pictures I couldn't upload due to restrictions.

https://imgur.com/yilpHYO
https://imgur.com/H6l2d0U
https://imgur.com/rp4t4Iz
https://imgur.com/VeHUq88

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Thu Dec 28, 2017 11:49 am
by A_Camera
Currently on a ferry between Germany and Sweden and my battery is almost discharged and I don't have a plug near me so this conversation will end soon...

Which units are you using? Your acceleration values are too small for mm and possibly too large for in. If you are using inches then you may not be able to have that high values at all...

What sort of motors are those? Looks pretty large...
What about your drivers? Are they up to those motors? What are the current settings?

It seems in the video that you have low acceleration, so it seems that you are using mm units, but I can be wrong. Run one motor at a time several times. Does it work? Maybe your speed setting are wrong as well... difficult to say, but to be honest, I don't know if I understand the actual problem. does it seem OK if you run slow? Or do you experience problems also at slow speed?

Re: Problems with stepsignals to driver, UC300eth - CP0-10V

PostPosted: Thu Dec 28, 2017 12:59 pm
by nikdan
Don't forget the taxfree! ;)


These are Nema 23's, 4Nm dual shaft motors. Measures 100mm length excluding the shafts.

This is the motors specs:
https://cnc4you.co.uk/resources/Stepper%20Motor%20Nema%2023%2060BYGH401-03%204Nm.pdf

The motors wired in Bipolar parallel is rated at 4 amps. Drivers are set to 3.8 Amps as recommended by the reseller (max is 5.6A).
I bought this from CNC4YOU as a kit, including motors, drivers, PSU's, and the breakout board so I do hope they fit together. They supposedly use them on all their machines.

I somehow managed to get them spinning nicely now, trying them separate as you said A_Camera and also increased speed and acceleration,
I use mm. But I have no real clue in what to enter.
These are the results with velocity 5000, Acc. 300:

X alone
X, Y, and Y1
X, Z
X, Y, Y1 and Z

As you can see on the X,Y,Y1 and Z combined, something is still wrong. But I have no idea what.

I learned that they only work like this if I start at the top of a new program, rather than starting in the middle of a program. Which is weird?
If I start in the middle of the program they only move CW. Not CCW, they just stand still.