[DoctorBit]

Hi All

I'm new to this forum. I had a very stupid question and I don't seem to find any answer on the net. Maybe I'm just not looking good enough?

Years ago I received a very small CNC from a friend. An old 2020 Chinese CNC,that I can't even find the model any more.

The problem is quite clear but still I don't know how to fix it. Whenever I have the stepper driver ON and I switch ON the spindle the stepper motors start moving by themselves, completely erratically.

Good, so I know what is causing the movement, is the electric noise coming from the spindle (or any other motor, I've tried the vacuum cleaner and a corded drill with all same result). Also noticed that when I plug the spindle to a different phase the motors stop moving erratically. Good, so it's the spindle. But just half a year ago I moved to a new studio and now I don't have 2 separate phases anymore and I'm stuck.

I've also done some measurements:

voltage between earth and GND (out of the Power Suply of the stepper driver) is about 1.5 V (should that be zero?)
voltage between earth (at the socket) and chassis (at the computer) is 0.06V when ON.
Voltage measured between the spindle and ground is incredible and with huge spikes


So the question is:

What is the best thing I could do in this situation?

I have some clues but I could be trying forever before finding the real one.

Options:
1- Filter the spindle to remove noise
2- Filter the driver to make it immune to noise
3- Opto-Isolate connection between computer and driver (don't know why but I have the impression that could be related)
4- Isolate Motors Power supply

and of course other options would be:
buy a new better spindle
buy a better driver
buy a better Power supply

So maybe you will scream at me cause the answer is obvious but still haven't find anyone that could help me. Please enlighten me

Many thanks!


My setup:

Since I got this CNC from a friend I never had the exact model but looking on internet I found a close match:

CNC driver controller: Chinese TB6560 CNC Stepper Motor Driver Controller Board 3 V3
Aluminum Box CNC Router 3 Axis 3.5A TB6560 Stepper Motor Driver Board Controller $62.30 Free Shipping @GoodLuckBuy.com
Aluminum Box CNC Router 3 Axis 3.5A TB6560 Stepper Motor Driver Board Controller

model 2020 mini 3D CNC router/milling
[*]X-Travel: 210mm/ 8.25"[*]Y-Travel: 190mm/ 7.50"[*]Z-Travel: 75mm/ 3.00"[*]Table Size: 250 x320 mm/ 10"x 12.5"[*]Interface : NcStudio breakout board(insert to computer Slot with 15 pins) or Mach 3 breakout board (25 pins port of computer)[*]Shaft size: X Axis:13mm Y Axis:16mm Z Axis:12mm[*]Lead screws: M10x1.5 pitch (mm)[*]Feed Speed: 1800mm/min (1500mm/min Recommend)[*]Repeat accuracy: Better Than 0.04mm/ 0.0015"[*]Machine Size: 385/380/390 mm 15"/15"/15.4" (Length/Width/Height)[*]Packing Size: 440/440/460 mm 17.4"/17.4"/18.1"(Length/Width/Height)[*]Machine weight: 13kg/29pounds[*]Gross Weight: 13.5kg/30pounds

[CitizenOfDreams]

I know someone will chime in soon telling you to throw away your TB6560 board... For some reason they get as much love on this forum as a red-headed stepchild.
I would start with grounding and shielding. How long is your parallel port cable? Is it shielded? What does your spindle setup consist of? Is it a DC motor with a PWM controller?

[DoctorBit]

hi CitizenOfDreams

Thanks for your input.

Well, parallel port is long, about 50 cm, shielded? I really couldn't tell but I would say it is shielded. I will check that. Since it is connected to the computer the shield (if any) is connected to earth (not ground) and at the TB6560 side seems to be floating (as the driver itself seems not earthed). Does that make sense?

The spindle setups is straight forward. It is a rubbish spindle that came with the KIT. I can not find any reference nor any comment on internet. It's ON/OFF brushed DC motor connected to a large relay that I use as a spindle control. If I'm not mistaken it is rated at about 600W and a low RPM, (never measured that).

I have to mention that the Power supply is also a Chinese one. So basically the combination anyone would like to have.

Do you recon I should start with a new parallel port cable? shorter and make sure that is shielded?

[CitizenOfDreams]

What powers your spindle? A separate DC power supply?

50cm parallel cable should be short enough not to cause any problems.

[DoctorBit]

O no, sorry. I just meant the power supply for the TB6560 :P

[DoctorBit]

Could it be the computer picking up all the noise from the spindle and sending erratic voltages?
Are the computer power supply good enough to handle all this noise?

If I connect the oscilloscope to a pin on the parallel port of the computer, will I be able to pick up the noise in the oscilloscope if there is any?

[DoctorBit]

I've done some mesurements today.

Connected the oscilloscope directly to the parallel port on the computer and measured the output pin of the X axis pulses. The PC is not yet connected to the driver. The spindle is ON.

Attachment 323334

You can clearly see spikes up to 6 volts.

And here some pictures of the ground rail at the stepper driver

Spindle OFF
Attachment 323336

With the spindle ON
Attachment 323338

So.. what can I do when my spindle is sending electric noise to reach another galaxy?
In the other hand, isn't the PC power supply suppose to take care of the spindle noise?

[Andreiir]

Try to use shielded wire for spindle , and ground chasis of cnc to earth ground .

[DoctorBit]

Try to use shielded wire for spindle , and ground chasis of cnc to earth ground .

I will do it today. Will post the results.

I've also read that if this fails I could make a ground plate. Where all grounds converge, also the PC ground.
Do they mean GND + Earth or joining earth and GND is a bad idea?

[Andreiir]

Join gnd to earth .

[DoctorBit]

Ok, It is working now.

I did join all ground and earth together: Chassis of CNC, Spindle and GND of the motors power supply. That did nothing, still there was huge spikes at the parallel port input of the TB6560 driver board.

Also changed the [power cable of the spindle for a shielded one. Nothing....

But I observed that all the noise was gone when measuring the input voltage at the motor driver while noisy at the parallel port optocupled section of the same driver board. So what I did is join inside the same board the input ground with the optocoupled section ground (previosudly isolated). And all noise is now gone.

So maybe now it's not optocoupled anymore, but at least I can work with it.

Hope that helps somebody with the same problem.