PWM motor noise getting into audio

[darren adcock]

Hey all

I have been working on motor driven mechanical sculptures that are controlled by a PWM signal, which is being controlled by control volatges from my synthesiser. The motors are Pololu 12v cylindrical brushed geared DC motors (Low power). The one in question is 260rpm.

The sculptures produce acoustic sound which I am amplifying with Piezo transducers and high impedance JFET pre-amps.

There is a high pitched frequency that comes from (i'm assuming) the PWM signal and the motor, but I don't understand this exactly. So I need to understand what this is so I can attempt to reduce or eliminate it from creeping into the amplified sound from the piezo's.

What I've tried so far; (and for some of the contraptions I've made it's worked with one or some of these combined)

>As standard I use coaxial cable in as short a length as possible from the Piezo to the pre amp. This is usually no further than 40cm or so.

>Shield the piezo inside copper tape

>Shield the pre-amp

>Use shielded cable from the pre amp to mixer (I am yet to make the ciruitry to send a blanced signal from the pre-amp to mixer, am awaiting components to try this)

>Ground the motor casing

>One instance with a 12v worm motor a 100uf cap across it's terminals, but on others it just sped the motor up.

>An op amp after the pre-amp set as an inverting op amp impedence buffer with a eg, 10nf, capacitor from the negative input to it's output

So far grounding the motor has reduced the noise the most, I am yet with this sculpture to have fully enclosed the pre-amp circuitry, but that should be done today which I'm sure will make a difference. As it stands there as not much bleed through and it's such a high frequency that it's almost inaudible. My fear is that when I get to the stage of wanting to trial amplifying this through a larger P.A the squeal will be intolerable, so I need to understand what is happening so I can address it at this stage of the build.

But clearly I am lacking some knowledge here, and am floundering around, have tried to search google to understand my problem fully but I think I'm possibly lacking a bit of terminology to unlock this puzzle.

Thanks in advance

Darren

 

[Bluejets]

Use optocouplers with NO common ground connection.

 

[darren adcock]

Thanks Bluejets

Use an optocoupler on the output of the pre-amp?

Confused by 'No common ground connection'.

So keep the ground from the Led (pre-amp) side of the optocoupler not connected to the ground that goes to the mixer?

 

[Audioguru]

A piezo has an extremely high impedance so the high input impedance preamp picks up all kinds of interference.
A piezo has a horrible frequency response making it a poor microphone.
Why not use a normal electret microphone and a preamp with an input impedance of 10k to 20k instead? The electret mic has a Jfet impedance converter inside so its output impedance is fairly low at about 2k ohms. The frequency response of an electret mic is almost perfect.

Simply, can't you increase the PWM frequency so that it is not heard?

 

[darren adcock]

Hi Audiguru thanks

I'm using piezo's as a means of amplifying the sounds that are too quiet for us to hear, from the vibrations within the mechanisms etc, but also there is also audible sounds that certainly can be amplified with electret microphones. But also, maybe it's convenient for me to use piezo's as i'm familiar with them, but then it's certainly not convenient having lot's of issues. So a rethink is possibly needed. Is there a better way to amplify the sounds/vibrations?

I'm sure I'd thought of increasing the PWM modulation frequency and tried it before but then found I got less response in terms of the PWM tracking the incoming control voltage. But this was a while ago now and my memory is vague at the best of times. I'll try this on Monday when I'm back in my studio and report back.

 

[darren adcock]

Sorry for delayed conclusion here. Increasing the PWM frequency to out of hearing range worked. Thank you.