Noisy PWM Circuit

[jeffreylotz]

I have a 12VDC circuit that uses an Atmel 8535 to drive two DC motors @ 14Khz PWM locked anti-phase. I have blown a couple of my on board 5VDC regulators and when I put a scope on the 12VDC power within the board it was very choppy. The voltage spikes seems to go over 50V until I put a 0.1 uf cap across the motor terminal and 2 more from each terminal to the motor casing. This reduced the spikes to 40V....still too high I would think. I have ultra fast clamping diodes in place. Does anyone know what to try next? I have added the schematic and photos of the scope when at 50% duty cycle (motors stopped) and the worse case duty cycle (around 80%) I notice that the spikes get larger when the there is a load on the motor and it is being asked to increase speed. Twisting the motor leads made no change.

Thanks in advance for the help.

JL

 

[itsmoked]

No scope pictures.?

Try adding a choke with enough current capacity to supply the motors just before pin 4 of your drivers. You might also try a good sized electrolytic at the driver supplies (after the choke).


Keith Cress
kcress -

http://www.flaminsystems.com

 

[jeffreylotz]

Here is the picture of the spikes.

Do you think a snubber on the motor terminals would do the trick?

JL

 

[blacksea]

Still aren't pictures...
What type of switches you use for power stage (i.e. MOSFET or something else)?
How you connect ultra fast recovery diodes to power switches?
What is motor inductance?

 

[itsmoked]

Look at his first post blacksea. There is a complete schematic. OTS driver chip.

No I don't think a snubber would work. See my comments above..

Keith Cress
kcress -

http://www.flaminsystems.com

 

[jeffreylotz]

Can you see the photo now? Not sure what's going on.

 

[blacksea]

OK with scope's photo.
Do you put electrolytic capacitor (470uF at least) to 12VDC supply?
What's pulses frequency?
Can you check voltage on current sense resistor(i.e. motor's current)?

 

[jeffreylotz]

Your right the snubber did not work. I put a 470 uf 100V LYT cap on the 12V rail and found that the spikes were reduced greatly, but still have spikes up to 19V and down to 9 volt with an unloaded motor. Not sure what the inductance is, here is the motor though:

http://www.mabuchi-motor.co.jp/cgi-bin/catalog/e_catalog.cgi?CAT_ID=rs_555pxvx

Thanks for the help.
Pulse feq is 14Khz and seems worse when the motor amperage is increasing.

JL

 

[Watthour]

An input reactor (choke) is probably a good idea.

 

[waross]

Do you have a free wheeling diode?

Bill
--------------------
"Why not the best?"
Jimmy Carter