Audio noise with op-amp. Need recommendations. 1

[monkeysolder]

Hello -

I am using a LM324 op amp to amplify an unbalanced signal and am getting a large amount of hum. I suspect this is due to the power source for the op-amp because when we take the op-amp out of the loop and listen to un-amplified audio, it sounds fairly clean. Perhaps the power noise is passed directly through the internal FET of the op amp?

My question is: Is there an established way to power an audio op amp when noise is a concern? I can't use a battery for this application, but am open to any power filtering, etc schemes. Right now, it is being powered by a 9V linear regulator which is fed off of a switching regulator package from TI.

Thanks,

 

[IRstuff]

Both the linear regulator and the LM324 should have lots of power noise rejection capability. Your conclusion about the LM324 is flawed. Your introduction of the LM324 in the path is not strictly about amplification; there's also grounding and cable pickup.

TTFN

 

[monkeysolder]

Hmm. Perhaps I am creating some ground loop inadvertantly when using the opam circuit. I will look into that. Thanks for hte suggestion.

 

[BrianR]

If you're powering the Op Amp as you say then it is more likely that the hum originates from the grounding connections of the input signal and the op amp in relation to the power supply. How is it connected?

Having said that it is worth confirming that the regulator is not oscillating.

 

[MikeHalloran]

Opamps will run for a surprisingly long time on 9V batteries. You might consider doing so temporarily for diagnostic purposes.



Mike Halloran
Pembroke Pines, FL, USA

 

[monkeysolder]

I will try powering the opamp circuit with batteries, but it is quite likely that this is a grounding issue, as there are several audio devices muxed together and sent over a long signal carrying cable.

 

[itsmoked]

Tell me you have twisted pair of wires and an outer shield that is grounded on one end only. (as a starting point). Have you measured the hum freq? 60Hz?


Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[monkeysolder]

Heh - well we do now. Going to a balanced audio driver, but the noise is present even without the long cable. It is just exacerbated by the cable length it seems.

 

[itsmoked]

Any way you can post your scuzmatic?

Posting pictures: faq238-1161

This would probably propel the discussion forward.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[MrBananas]

Is the input AC coupled to the cable to block DC? DC bias from the previous stage if present could cause some hum in the output.

Also, are you using two supplies (pos and neg)? Opamps don't operate rail-to-rail so that could be messing up the signal if it is AC coupled or goes within about 1.5 V of either rail.

What are the ranges (amplitude and freq) of your input signal?

 

[BrianG]

The LM324 is better than many op-amps in that its inputs can go from 0V to V+ supply without damage. Common mode input voltage range is 0V to approx V+ supply, less 1.5V, so any d.c. offset is less likely to be a problem with this device.

What has not been asked, or stated, is the gain that the circuit is configured for: the 324 has an open-loop gain in excess of 90dB, so any mains hum will really get in there!

Also how is the input to the 324 configured (presumably single-ended) but where is the ground side referenced?

 

[monkeysolder]

We do have DC blocking before amplification. Can't post a schematic, but looking at it here, I can see many of the ground loops and impedance mismatches. Unfortunately the PCB and enclosure have already been spun, so it will be hacks and floating chips. You guys probably know what I mean...

Thanks for the help

 

[BrianR]

If you're down to hacks then try this.

1. Substitute an appropriate load resistor for the audio cable and optimise the amp for hum. Do this by ensuring that the input signal path does not share any common conductor path with either the power supply ground return or output signal return to power and filtering power.

2. If possible use a shielded twisted-pair for the audio input cable as this offefs much better opportunity for noise protection than single shielded/coax.

3 If the input signal source is isolated from ground then keep it that way, connecting the twisted pair to the signals and the shield to ground at the signal source only.

4. If one of the signal source connections is grounded, then connect the shield to this ground as well as one of the twisted pair.

 

[Noway2]

Regarding your comment about being down to hacks, be aware that sometimes hacks can make things worse rather than better and also don't spend more effort trying to hack something than it would take to redo it correctly in a proper way.

 

[OperaHouse]

If you are operating off a single end supply, You obviously have to offset the input with some voltage. I suggest you filter that offset voltage seperately. If you just made a V/2 divider, perhaps this is the source of the hum and you are amplifying it.

Of course, it may be because it doesn't know the words.

 

[monkeysolder]

Thanks for the help all. I've identified the source of the noise - ground loop resulting from the audio ground and opamp ground. Tied those two together and it seems to be clear over the entire cable length now.

Thanks!