PWM - Mosfet/motor introducing noise

Printable View

- 16th October 2007, 16:51
DynamoBen

PWM - Mosfet/motor introducing noise

I'm building a prop controller for Halloween. The circuit I've designed has two relays controlled via transistors and a PWM motor controller via a mosfet.

The mosfet gate is tied to the pic via a 1K resistor. The motor and relays all have a 1N4001 tied between 12V and common. I’m using the same 12V supply for the motor, relays, and the rest of the circuit (5V regulator to the circuitry). The PWM controlled Fet is connected to a windshield wiper motor from a car.

The issue I'm running into is when the motor is run I am seeing spikes on the power buss.

Any thoughts on how to correct this issue without dramatically changing the design? I could drop optos in and use a separate power supplies but that would require a ton of rework.
- 16th October 2007, 17:05
mackrackit

Are the spikes on the 5 volt buss? What do you have in the way of capacitors before and after the the 5 v regulator?
- 16th October 2007, 17:08
DynamoBen

1 Attachment(s)

Here is a working schematic.

I'm sharing a common between the 12V and 5V rail. I haven't looked at the 5V rail yet.
- 16th October 2007, 17:32
mackrackit

Try a 4700 or greater cap across the motor leads?
- 16th October 2007, 23:25
DynamoBen

2 Attachment(s)

I tried a 2200K, no change.

Here are some scope shots of the 5V and 12V rails with the motor running.
5V = motor controller2 sml
12V = motor controller 12sml

2ms/div, 1V/div (x10 probes)
- 17th October 2007, 02:51
DynamoBen

After some thought I may just leave things as they are. To get it perfect will require a major redesign which I don't have time for. At this point I'm not sure that this will be a big deal. Agreed?
- 17th October 2007, 08:54
Acetronics2

Hi,

First ...

Your diode parallelled to the mosfet must be paralleled to the MOTOR ... that will GREATLY limit the spikes ( 1N400x seems a bit Weak for this kind of motor ... try better 1N540x or a fast recovery TO220 model )

Now,

1) place a low ESR 1000µF or + capacitor as close as possible to the motor ... to lower the 12v power source impedance. ( by the way, I'm saying the mosfet could be placed close to the motor ...and the diode between motor terminals ...)

2) place a .1µF on your motor terminals

3) place a 10nF between each motor terminal and motor frame which has to be connected to circuit ground.

4) you can place a core coil ( 1mH range "big value" ) in series beetween 12 v and your 5v reg input

Alain
- 17th October 2007, 17:00
DynamoBen

Quote:

Your diode parallelled to the mosfet must be paralleled to the MOTOR ... that will GREATLY limit the spikes ( 1N400x seems a bit Weak for this kind of motor ... try better 1N540x or a fast recovery TO220 model )

The schematic isn't correct in this case, the diode seems to be drawn wrong. However I will try moving the diode up to the motor. I will try to located a different diode...

Quote:

1) place a low ESR 1000µF or + capacitor as close as possible to the motor ... to lower the 12v power source impedance. ( by the way, I'm saying the mosfet could be placed close to the motor ...and the diode between motor terminals ...)

The mosfet can't be moved. I don't know if I have a low ESR cap. I can dig through my parts bin. Anything specific I should be looking for?

Quote:

2) place a .1µF on your motor terminals

Is this in addition to the above cap?

Quote:

3) place a 10nF between each motor terminal and motor frame which has to be connected to circuit ground.

Will try.

Quote:

4) you can place a core coil ( 1mH range "big value" ) in series beetween 12 v and your 5v reg input

I don't have anything like this lying around.
- 17th October 2007, 18:17
Acetronics2

QUOTE=DynamoBen

The mosfet can't be moved. I don't know if I have a low ESR cap. I can dig through my parts bin. Anything specific I should be looking for?

> IF I tell you BC Components (ex. Philips ) Series 136 ... ESR <= 22 mOhms ( Milli ...yes ! )

Is this in addition to the above cap?

> The .1µF capacitor must be soldered ON the motor terminals ...

Will try.

> The 10nF are also necessary.

I don't have anything like this lying around.

> Not compulsory, yes ...but highly recommended !!!

Alain
- 17th October 2007, 21:32
DynamoBen

This seems to be what you are describing.

http://www.kronosrobotics.com/an142/GAN142.htm
- 18th October 2007, 08:39
Acetronics2

Quote:

Originally Posted by DynamoBen

This seems to be what you are describing.

http://www.kronosrobotics.com/an142/GAN142.htm

Hi, Ben

Yes ... that's a nice tutorial ... but some tips still missing !!! ( diode, low esr cap, inductor ...)

note here, in this example, motors can be reversed ... so, the diode can't be placed on the motor terminals, but close to the "reversing commutator" ( relay, transistors bridge ...)

Alain
- 18th October 2007, 15:37
DynamoBen

Quote:

1N400x seems a bit Weak for this kind of motor ... try better 1N540x

I picked up a 1N5406 at Radio Shack last night.

Quote:

1) place a low ESR 1000µF or + capacitor as close as possible to the motor ... to lower the 12v power source impedance.

I was unable to locate a low esr cap, so this one can't be done.

Quote:

place a .1µF on your motor terminals

done

Quote:

place a 10nF between each motor terminal and motor frame which has to be connected to circuit ground.

done, frame of motor is connected to common

Quote:

you can place a core coil ( 1mH range "big value" ) in series beetween 12 v and your 5v reg input

Still no luck locating a coil.

In the end my "issue" still exists. *sigh*