Distortion / volume issues with amp when pot is on long leads

Other Hardware General Electronics
1

Hey guys,

Some buddies of mine on another forum have been having an issue with the sound systems in the props the build.

They have a backpack with a sound board, speakers, and an amp in it, and connected to that is a long hose which goes to a gun they hold in their hands. They've been trying to desolder the potentiometer from the amp so they can put the volume control on their gun, but they've found whenever they make the wires more than a foot long, they get all sorts of problems. Some have reported full volume constantly, others say the sound is distorted.

I'm thinking the long leads are picking up EMI and causing the problems, but I'm not sure what the best solution to suggest is. Do you think wrapping the three wires from the pot around a ferrite bead would be best, or should they use decoupling capacitors? If they use decoupling capacitors, can they just use ceramic caps and wire them up in parallel with each of the leads to the pot, or is there another way they should wire it up, or can they get by with just one cap per pot on the middle wire? Also, what do you think would be the best value for said pot for an audio situation where you don't wanna block much below 22khz?

2

Also, what do you think would be the best value for said pot for an audio situation where you don't wanna block much below 22khz?

It impossible to make a suggestion w/o electrical schematics.
The best way is not run AC/Audio signal via long lines, rather DC, which control audio at other end.
Alternative digital bus / digital pot control. I don't think caps or ferrite would solve a problem

3

The amps they're using are from a kit, so there's no way to change how the circuit functions. I'm not even sure whether audio is being sent thru the pot or if it's simply being used to adjust the voltage on some other chip. I'm just looking for a best guess for folks to try out to solve their problem. I don't have one of the kits myself to work with.

4

There are several things that could be happening. Certainly both the input and output sides of the pot should be connected via separately screened cables - that will reduce the EMI problem. But the stray capacitance of the leads themselves may be creating an unintended radio-frequency oscillator from the audio circuit. In general you keep analog signal leads short for this reason and use correctly terminated transmission lines if the runs are long.

It may be possible to improve things with some RF-filtering, but with an understanding of the circuit its tricky to propose a solution...

5

Try braiding the pot leads, not tightly like some camp craft project, just loosely so.

6

but they've found whenever they make the wires more than a foot long, they get all sorts of problems. Some have reported full volume constantly, others say the sound is distorted.

If you get constantly full volume then it is wired up wrong.
If you get distorted sound then you have too much capacitance on the lead.

The amps they're using are from a kit, so there's no way to change how the circuit functions.

In which case there is probably no way to make it work how you want.

However in place of guessing what is wrong you need to get a scope on the signals and actually see what is happening when you make this extension to the pot.

7

You could try using 3-core screened lead to connect the potentiometer to the amplifier. Connect the screen (braiding) to the amplifier ground point. Do not connect the screen to anything at the pot end. In fact insulate the end of the screening it to ensure it cannot accidentally connect to anything.

Working at audio frequencies is not rocket science but normal precautions need to be taken to prevent noise pick-up; screened leads usually prevent any problems.

As to cable length, when you consider that microphones, guitars etc have long connecting leads and work at very low signal levels, there's no reason why a remote amplifier potentiometer that is working at much higher levels cannot be made to work.

Fitting capacitors to filter out noise pick-up really isn't the way to go; better to avoid the noise pick-up in the first place.

8

As to cable length, when you consider that microphones, guitars etc have long connecting leads and work at very low signal levels, there's no reason why a remote amplifier potentiometer that is working at much higher levels cannot be made to work.

Yes there is, it is all to do with the impedance of the source and destination. You can only have long leads with a low impedance. If the amplifier at this point has a high impedance then fundamentally you can't have a long lead.
The lower the impedance the longer lead you can get away with.

9

Yes Mike, but he's only talking about the length of a human arm plus a bit, and in all likelyhood the circuit impedance will only be a few k so impedance mis-match shouldn't be an issue.

10

impedance mis-match shouldn't be an issue.

It is not miss-match that is the issue, it is the absolute impedance that is the issue.

he's only talking about the length of a human arm

It seems to be long enough to be causing problems already.

11

Grumpy_Mike:

As to cable length, when you consider that microphones, guitars etc have long connecting leads and work at very low signal levels, there's no reason why a remote amplifier potentiometer that is working at much higher levels cannot be made to work.

Yes there is, it is all to do with the impedance of the source and destination. You can only have long leads with a low impedance. If the amplifier at this point has a high impedance then fundamentally you can't have a long lead.
The lower the impedance the longer lead you can get away with.

Would a larger gauge wire lower the impedance?

12

It's got nothing to do with the resistance of the wire. Simply get some fine screened three core and give it a try. Theory and practice are not necessarily the same thing. We get bogged down in theory far too often. Experimentation and making things work is much more satisfying that trying to theorise why a suggestion won't work.

13

I'll suggest the guys try the screened wire thing, but where would you get that sort of thing?

I asked them to describe the behavior some more and this is what they said:

Well, I'll tell you how it sounded to me.

There was the audio of the SFX, which would be LOUD, but faint (as in, not a lot of clarity), and the rest of the sound was this crunching static noise. The video I posted at the start of this thread has it, to a small degree. There was a definite lack of bass or treble, I suppose, when I ran it through my piezos AND 3-way speakers.

The potentiometer itself; as someone else mentioned, the best way to describe that would be "little to no control". When turned all the way DOWN, the sound was still blasting and staticy. When turned about HALFWAY, it's at it's quietest, and then when turned all the way UP, it was blasting and staticy again. Also of note, the volume would shift erratically when turning the pot from lowest to midway to highest. Like, "LOUD faint, getting louder LOUD LOUDER faint LOUD" all in the span of a few seconds, and all with that crunching static. Little to no control!

It's really odd that the volume would go up, then down, then up again as they turn the pot. And several guys have had the same problem and they said they had three wires going to the pot, so I doubt they all wired it wrong. They also tested with the shorter leads and said that worked so again, probably not wired to the wrong poles.

But I guess the leads acting as antennas would explain that and why the volume might change erratically as the resistance on the pot is shifted.

So nobody thinks adding a simple ferrite core will solve the problem? What are those for if not to reduce noise in the lines? Or are they only good for DC circuits?

Oh, and one last thing... Some of them were having problems with the amp even before they tried to extend the leads on the volume control, and they managed to solve the problem by adding a pre-amp designed by a another member of the board. Here's what he had to say:

I looked at the signals from the sound module and it is a differential PWM output. A differential input on an amp should work OK, but may also need a low pass filter, depending on how good the amp and speakers are.

The sound from the module is just not loud enough outside, so I added an amp this weekend. The amp is a simple fixed gain amp with single ended inputs. The sound was very distorted if it was driven directly from either side of the sound module's speaker driver. I looked through my parts bins and found an LM386 dual differential op amp, so built a small circuit to convert from the sound module's output to a more normal single ended output.

Here is the circuit diagram that converts from the sound chip speaker driver outputs to a single ended usable by most amps output. I am running the sound module at about 4.5V, so needed to reduce the output by a factor of 2 to get the maximum volume out of my amp without too much distortion.

If your amp can take a different range, or the sound is too loud at max or too soft at max, just change R5 and R6 to the desired gain. Gain = R5 / 200K, so in the schematic, the gain is 1/2 (100K / 200K).

799×504 42.5 KB

800×600 97.3 KB

I don't know if any of that information is helpful in any way.

And again, I don't have one of these amps to play with, so I can't test anything, I'm just trying to get suggestions for others to try.

14

Volume going up, down and then up is defiantly mixing up the three wires of the pot. No doubtbabout it.
Screened wire from any hi-fi shop.

15

jackrae:
Yes Mike, but he's only talking about the length of a human arm plus a bit, and in all likelyhood the circuit impedance will only be a few k so impedance mis-match shouldn't be an issue.

At RF frequencies the mismatch will be significant - just because you think its an audio amplifier doesn't mean it can't be an RF oscillator too... Parasitic oscillations cause various forms of audio distortion and this could be what's happening due to extra capacitances and inductances that are present with long leads.

16

I don't know if this will help at all, but I found out the amp they're using actually has a datasheet and schematics:

http://www.quasarelectronics.com/3047-8w-mono-amplifier-module-tda2003.htm
http://www.quasarelectronics.com/kit-files/electronic-kit/3047.pdf

I sent them the suggestion about the braided wire.