Fan causing noise in audio circuit

[The Master]

Hi, I've made an audio circuit that gets quite warm so I want to add a fan to it. I've tried a 12V DC brushless fan but it causes audiable interference. Is there a way to stop it?

I have tried:-
Wrapping the fan wires around a toroid with no effect.
Placing a diode backwards accross the wires. This stops the fan from working.
Running the fan through a separate regulator. No effect
Adding capacitors. No effect or stops the fan depending on the configuration.

Unfortunately I don't have the option to run the fan from a completely separate supply. Is there anything else I can try or maybe a different type of fan?

[mister_e]

What kind of audio circuit is that to be that sensisitve and heat at the same time? a brushless shouldn't make ANY noise.

High likely some kind of Faraday cage 'round the high gain audio stage would help.

[The Master]

It's a simple 14W TDA2030 setup as per the test circuit in the datasheet

[sayzer]

Have a bobbin serially connected to GND of the circuit.

Try different bobbins if you see no change.

[Ioannis]

Totally agree with Steve. Brushless should make NO noise. Also TDA2030 is a little insensitive to noises since it is an automotive chip.

But on the other hand you may have a ground loop that is causing noise. Connect all grounds to a signle point preferably at power supply 0.

Ioannis

[mister_e]

Ground routing could be a problematic, a poor layout as well. Disconnect your audio source, Short the input to ground and see if you still have noise.

[The Master]

Sorry, I'm not sure what you mean by "bobbin" (not in an electronics sense anyway).
All grounds are connected to the same power supply. The audio source is a SOMO-14D which is also powered from the same power source. I tried disconnecting the input and grounding it anyway but I still get the noise.
This circuit is currently on a breadboard so the layout could be a problem but it shouldn't affect it that much.

[cncmachineguy]

Analog is not my thing - and definitely not audio, but how bout this. Since it doesn't make sense to be getting noise from the fan on the wires, maybe it is radiated from the motor. After all even in brushless there is still magnets spinning and such. Try moving the fan around and see if the noise changes. Now if it does change, I will still be of no value to get rid of it, but at least you will know the source.

[mister_e]

Audio and breadboard are poor bedfellow. worst when you have hi-gain/power stage. you're running into trouble for sure. I wouldn't use a fan though.. just a big/bigger heathsink, period.. it's just a couple of watt.

[sayzer]

Sorry, I'm not sure what you mean by "bobbin" (not in an electronics sense anyway).


What can I say now?
http://en.wikipedia.org/wiki/Bobbin

"in an electronics sense", the noise is generated at the ground of the circuit.
So, "in an electronics sense", a "bobbin" is used as the last component going to the ground of the power supply.
In other words, the "bobbin" is placed between the ground of the circuit and the ground of the power supply.

Usuallly, people make mistake by placing the "bobbin" to the positive supply and it makes no difference.


Sorry, I'm not sure what you mean by "bobbin" (not in an electronics sense anyway).

Are you clear what I mean by "bobbin" now?

[Charles Linquis]

Brushless fans DO generate a fair amount of noise. They might not have brushes, but they do switch the coils via FETs. Since the fan probably draws less than 200mA @ 12V, I would put a 10 ohm 1W resistor in series with the fan, and a 100uF + .1uf across the fan leads. You should put the resistor and cap close to the fan (not at the end of 10" leads!). The fan will run a bit slower than normal, but I assume you can tolerate that. Cheaper and smaller than an inductor, and (for this purpose) probably just as effective.

Also be aware that your fan not only has switching coils, it has rotating magnets. Any nearby conductor (less than 3" or so). will have induced voltages.

I can't imagine getting much noise if you use the filter I described and have the fan well away from your circuit.

And I'm with Mister_e. Breadboards are no good for low noise - or high frequencies..

[The Master]

Thanks for all of your advice. I decided to go with bigger heatsinks and lose the fan completely.

Moving the fan etc didn't seem to have any effect. I will test the full circuit out with the volume right up and check that the heat is ok. I can always add the fan later if I really need it