Sporadic spindle enable signal

[per.takman]

My troubleshooting efforts unfortunately came to an abrupt halt this evening. It seems like I've fried my UC400ETH while trying out a different power supply to see if the supply was picking up disturbances. I'm not sure exactly why it happened and I've emailed about purchasing a replacement.



/Per

[A_Camera]

That's very annoying. Perhaps it's just an "easy" fix? Could it be a diode or something similar on the power input side? I don't know if there is any, but I'd certainly try to find out and would do some measurements. Also, a stupid question maybe... did you try to reset the UC400ETH to factory default, or is it definitely dead?

[cncdrive]

My troubleshooting efforts unfortunately came to an abrupt halt this evening. It seems like I've fried my UC400ETH while trying out a different power supply to see if the supply was picking up disturbances. I'm not sure exactly why it happened and I've emailed about purchasing a replacement.



/Per

Hi Per, Yes, I've sent you an answer e-mail yesterday.

[per.takman]

I believe that I've let that magic smoke out from the 5V voltage regulator. I will send it back for repair/replacement when I'm back from a weekend trip. I feel a bit stupid about damaging it, but I'm still not sure why it happened. I hope that I did not kill the stepper driver in my Stepcraft at the same time.

The reason I was trying out a different supply and laborating with ground was that my pump spindle somehow induce noise, when plugged in, into the parallel cable that cause the stages to loose steps. Adding a couple of ferrites on the parallel cable seemed to solve the problem at first. But when connecting the second IO port to an opto-isolated relay box I built to allow me to turn on/off the pump as well as my vacuum machine from UCCNC the problem came back. I therefore need to find a way to make the UC400ETH less sensitive to these disturbances.

I'm considering solid state relays instead but also replacing the plastic box for the UC400 with one made out of metal so that I can ground the chassis.

Any information available on how to make it more robust is welcome. Keep in mind that I'm not an electrical engineer (just one of those x-ray scientists) so my knowledge about electronics is rudamentary.

/Per

[cncdrive]

I believe that I've let that magic smoke out from the 5V voltage regulator. I will send it back for repair/replacement when I'm back from a weekend trip. I feel a bit stupid about damaging it, but I'm still not sure why it happened. I hope that I did not kill the stepper driver in my Stepcraft at the same time.

The reason I was trying out a different supply and laborating with ground was that my pump spindle somehow induce noise, when plugged in, into the parallel cable that cause the stages to loose steps. Adding a couple of ferrites on the parallel cable seemed to solve the problem at first. But when connecting the second IO port to an opto-isolated relay box I built to allow me to turn on/off the pump as well as my vacuum machine from UCCNC the problem came back. I therefore need to find a way to make the UC400ETH less sensitive to these disturbances.

I'm considering solid state relays instead but also replacing the plastic box for the UC400 with one made out of metal so that I can ground the chassis.

Any information available on how to make it more robust is welcome. Keep in mind that I'm not an electrical engineer (just one of those x-ray scientists) so my knowledge about electronics is rudamentary.

/Per

As I described you in e-mail the UC400ETH is not sensitive at all to noise.
It is a board with a single power supply input and therefor where electrical noise could enter the board is via the powersupply or the IDC26 connectors.
However if the UC400ETH would pick noise up the symtons would be lost connections to the computer, because what noise would effect is the power supply voltage, but this is clearly not the case at you, your machine is just loosing steps when you operate your blower pump.
The other possible scenario is that the output drivers of the UC400ETH pick noise up, which are 74HC14 buffer driver chips connected directly to the microcontroller pins.
The microcontroller pins can sink/source about 8mAmps of current, so if noise entering the inputs of the buffer ICs would mean that the noise should generate 8mAmps of currents inside the PCB wires/the chip,
this I do not beleive is possible, because the chips are so close to the microcontroller that this amount of noise would sure lock up the microcontroller also which does not happen.
So, for me it is crystal clear that the noise entering to other parts of your controller and not the UC400ETH.
That you putting ferrits on the IDC wires reduces the noise is only because the control electronics in your machine has non-isolated circuitry on the step/dir lines and therefor the step/dir lines are common referenced to the other part of the circuitry and noise entering that board is therefor seen also on these wires and the ferrits damping the HF domain of the noise.

I think to resolve this issue the only solution will be to elliminate the noise at it's source which is to install the proper EMI filter in your blower.

About the damage happened: However you don't exactly know what happened, but you said you connected the power supply negative (GND) to the earth. These 2 points are not the same potential in switching mode power supplies, they are floating to eachother as SMPS are isolated voltage converters. The 2 points are usually only AC coupled, they are never DC coupled!
If you made this connection while the device was already operating I think a huge discharge happened between the 2 points while the potential difference tried to balance, this could damage the circuit, the voltage regulator...

[per.takman]

You may very well be right about those conclusions. What puzzled me and caused me to believe that noise entered via either the PSU or the d-sub of the UC400ETH was the fact that connecting the relay box to the second IO port made the problem come back after I thought that I had solved the problem using ferrites.

If I understand you correctly the circuitry from Stepcraft can be made more robust and I need some kind of filter for the pump. My problem is that I don't know how to design/pick out one myself.

/Per

[cncdrive]

Dealing with noise can be a hard task, because to elliminate the noise or at least to reduce it to a level where it causes no more problems to other devices you need to know where the noise coming from and you also have to understand the properties of the noise. For these informations you have to study where and how the noise is generated and you have to know the frequency domains of the noise to be able to select or design a successful filter for them.
So, it is not any easy task mostly to resolve this kind of issues.
Sometimes even some ferrit beads can help, if the noise has a high frequency domain only and sometimes a much more complex and more powerful filtering is required.

I saw a few older machines, e.g. Gerger Sabres where the VFD and vacuum pump filter contained a filter with larger size than the VFD itself, the weight of the filter was about 10kgs built from huge chokes and different types of capacitors and resistors networks...

[per.takman]

As you're aware of, I now have a new UC400ETH board installed and I've resumed both the stepper motor noise troubleshooting effort as well as the spindle enable signal.

1. Since I installed the new UC400ETH board I have not been able to reproduce the sporadic spindle enable signal. I've tried both with the dongle connected and the remote inside my drawer (Faraday cage) and next to the computer for a couple of days now. Nothing has changed on my side except the UC400ETH board.
2. I've tried installing a line filter between the supply and the spindle pump. It removed the problem with the steppers twitching when combined with two clamp on ferrites.

I will let you know if I ever observe the sporadic enable signal again.

Thank you for your efforts!

Warmly,

Per Takman

[cncdrive]

Hi Per,

The UC400ETH can't switch the spindle on with the M3 button goes on on the screen, except if an input trigger is configured.
From the debugging it was clear that the spindle is turned on on your system by the UCR200 pendant, so it can be that the pendant works erratic, or the plugin has a problem of the USB receiver pick up somekind of noise.

[per.takman]

Hi,

Don't get me wrong... I understand that the source of the problem is the UCR200. I just noticed that I can't seem to reproduce it. I will be watching out for it to happen again.

Cheers, Per