PDA

View Full Version : Info about Snubbers



mikebar
- 25th April 2007, 13:20
Hello guys.
Just a question.
I built a tool machine that move by DC motor.
I'm powering a 24VDC motor by a PIC 16F877 with LM629 Precision Motor Driver Controller and the power stage is built with LMD18201T.
The motor is far away from the controller board.
Anything seems to work fine BUT SOMETIMES I see that OTHER boards that come closer to this motor when the tool machine is moving pick some disturbance coming sure from the motor (like a coupled magnetic field). Actually the hall effect sensors used as encoder and that are on a tiny motor inside the board that get noise are no switching at all when the structure is close to the big motor. If I move the board (and the tiny motor) that get noise, away from the motor while he is running, the board with the tiny motor start working properly.
The big motor seems to have filter capacitors.
I want to try something different.
What you thing about SNUBBER solution?
I'm asking because in this case, the inductive load (the big motor) is powered by a PWM signal coming from the LM629/LMD18201T also when the motor shaft is not turning. Actually the system hold the motor rock steady until the control board command some coordinates to move to.
I've experienced that I get noise (of course) also when the shaft is steady, because there is continuously a PWM signal about 50% duty cycle.
I saw, also, that modifing the path of big motor leads the problem LOOKS LIKE disappeared... but I want to be sure about this...

Shortly:
I have two boards. One is steady and have big motor controllers on it. Big motor is far away from this board (around 21 inchs). It work fine.

I have another board that is moving by the big motor above. This board have on top a tiny 24VDC motor with a 2 channel encoder built with 2 Allegro hall-effect chips (90° phase)
Looks that the 2 Allegro are like saturated when they are close to the big motor.

The big motor specs are:
V = 24VDC
Imax = 5A at startup only
I don't know the inductance...

Power for both big and tiny motor is coming from a 24VDC industrial power supply (very well done)

Power for the logic of the main controller board AND for the board that get noise is coming from ANOTHER 9VDC industrial power supply (also this very well done).
Grounds from the 24VDC and from 9VDC are completely separated.
Both boards have 0.1uF decoupling capacitors everywhere.

Actually I DON'T get microprocessor strange behaviour. Simply I do not see incoming pulses into RA.4 (setup as timer0) when the "moving" board get close to the big motor. In that condition microprocessor still work as expected, also still able to send back the error condition by the serial port and also still able to get other commands from his serial port.

Seems that I have to screen the big motor in some way.

Do you thing that a snubber (if applicable) could be the final solution?
Or, if the snubber is not applicable, do you could suggest a solution?

Thank you a lot.

mackrackit
- 25th April 2007, 17:47
I had a similar problem and it was solved by connecting earth grounds as close to the boards as possible and adding an earth ground from the mains before the transformers.

Sinked the inductance.

mikebar
- 26th April 2007, 16:01
Thank you Mackrackit.
I already have earth grounds as you suggested.
Also connected in one point only.

Any idea?

mister_e
- 26th April 2007, 16:03
can you post some picture and schematic of it?

mikebar
- 26th April 2007, 17:21
Thanks Steve.
Actually you already had some idea on this my project a while ago.
The schematic is actually already there:
http://www.picbasic.co.uk/forum/showthread.php?p=24177#post24177

If it will be needed, I will try to post also pictures as soon I get assembled another machine.

Thank a lot.

mister_e
- 26th April 2007, 17:34
Yeah some installation and board picture (top/bottom) would be nice

mackrackit
- 26th April 2007, 21:01
The big motor seems to have filter capacitors.

Is this the same motor you referenced in the other thread? I thought you added capacitors?



Actually I DON'T get microprocessor strange behaviour. Simply I do not see incoming pulses into RA.4 (setup as timer0) when the "moving" board get close to the big motor. In that condition microprocessor still work as expected, also still able to send back the error condition by the serial port and also still able to get other commands from his serial port.

I am sure you have but have to ask the obvious. Have you checked or changed the cable and checked the connections to RA.4?

mikebar
- 27th April 2007, 02:39
Is this the same motor you referenced in the other thread? I thought you added capacitors?


I am sure you have but have to ask the obvious. Have you checked or changed the cable and checked the connections to RA.4?

Capacitors was "supposed" to be already installed by the producer...

Yes, cable checked. Also changed just to be sure. For checked connection to RA.4, if you mean look at connector, yes I did it. If you do not mean this, what other kind of checking to look for on RA.4 or suggestion to connect to RA.4?
But anyway, I remember: moving far away the tiny motor from the big one, pulses are regularly coming to RA.4...

mackrackit
- 27th April 2007, 02:52
Just thinking that when the board moves close to the big motor a strain is put to the connector / solder joint / etc.

Sounds like you checked all that as I thought you would have.

Do I have this correct? The only problem is -- when board gets to close to the big motor, RA.4 does not receive a signal. Every thing else is working "OK".

mikebar
- 27th April 2007, 07:01
Just thinking that when the board moves close to the big motor a strain is put to the connector / solder joint / etc.

Sounds like you checked all that as I thought you would have.

Do I have this correct? The only problem is -- when board gets to close to the big motor, RA.4 does not receive a signal. Every thing else is working "OK".

Yes, exactly. When the board gets to close to the big motor, RA.4 does not receive signals coming from the Allegro hall-effect chips.
Anything else is working fine and as expected (serial communication, big motor's position other motors tracking etc. etc.).

mackrackit
- 27th April 2007, 08:40
You need magnetic shielding, not electrical snubber.

The motor case will provide some shielding, but if the case is thin or low permeability, the field will pass through. You have earth grounds, good cable, good connections, etc. This is all that is left.

Several ways to do the shielding, with out seeing the setup I can only make a couple of suggestions.

Wrap the BIG motor in a material with a high magnetic permeability and/or the same around the enclosure with the hal-effects.

Google magnetic shielding This site looks like it has good info http://www.lessemf.com/mag-shld.html.

Post a picture or two and maybe we can be of more help.

mikebar
- 27th April 2007, 09:56
You need magnetic shielding, not electrical snubber.

The motor case will provide some shielding, but if the case is thin or low permeability, the field will pass through. You have earth grounds, good cable, good connections, etc. This is all that is left.

Several ways to do the shielding, with out seeing the setup I can only make a couple of suggestions.

Wrap the BIG motor in a material with a high magnetic permeability and/or the same around the enclosure with the hal-effects.

Google magnetic shielding This site looks like it has good info http://www.lessemf.com/mag-shld.html.

Post a picture or two and maybe we can be of more help.

THANK YOU VERY MUCH for the suggestion.
I will put pictures as soon I will mount another machine.