Reducing DC potentiometer noise. (N-channel Mosfet amplifier)

[iampoor]

http://www.modkitsdiy.com/sites/default/files/product_files/the_piledriver_schematic.pdf

Been playing around with this circuit for the past few days to build a  nice high gain clean boost for my guitar.

Issue im running into is a scratching sound whenever pot settings are changed. I believe it is because the pot has dc through it. I have tried different output caps (1-10uf), and using a cap from wiper to ground (from 22uf to 2200uf) to stop the noise, but nothing changes. Is there something typically done in this situation? I wasnt ale to find any real god resources 9Maybe my googling was bad)

My few buds locally dont really have a clue on what the issue is, and Im clueless! I have seen a cap to round before in this situation but it does not seem to work in this circuit.

ANy help would be highly appreciated!

[SeanB]

You will need a better quality pot, or to clean the one you have with Caig Deoxit. There is no way not to have Dc through the pot in that circuit, so either change it for another or try the cleaning first.

[Rerouter]

your only option is to get a quality potentiometer, if it goes open circuit for the smallest moment you will get as large as a 9V pop,

[Mark Hennessy]

This is simply DC on the wiper - a big no-no in audio circuits. The capacitor between the wiper and ground is the only thing you can do, but of course, the damage has been done now, so you'll need to fit a new pot at the same time as the capacitor.

Personally, I wouldn't waste your money with a more expensive pot - they are no better at coping with this. Well, perhaps a really expensive wire-wound job might be OK. The "right" answer is to design the circuit properly in the first place, I'm afraid.

Mark

[iampoor]

Thanks guys. I have tried a few different ports, results all seem to be similar. I am trying to stay away from anything estorstic.

This is simply DC on the wiper - a big no-no in audio circuits. The capacitor between the wiper and ground is the only thing you can do, but of course, the damage has been done now, so you'll need to fit a new pot at the same time as the capacitor.

Personally, I wouldn't waste your money with a more expensive pot - they are no better at coping with this. Well, perhaps a really expensive wire-wound job might be OK. The "right" answer is to design the circuit properly in the first place, I'm afraid.

Mark

Hey Mark

What is the proper way to build an amplifier like this? (I know I know, engineering requires context and tradeoffs, Im not asking for a golden rule, just a nudge in the right direction ) I can think of a few different configurations off the top of my head

A. VOlume control on output, fixed gain up front.

Pros: No noise
Cons: Increased noise, gain has to be too low without there being any drive... (Deal breaker)

B. Pot on gate  input creating a resistor divider that reduces input gain.

(Not sure if this would work)

C. Pot controlling a small 3 terminal regulator that sets the transistor voltage.

Pros: Circuit stays the same
Cons: Expensive

Thoughts? Im sure there are other ways of accomplishing this.

[vk6zgo]

For over 80 years,people have realised it is a silly idea to run dc current through volume control pots in audio circuits,so they must know something.

You could use a pot between switch "A" & the 1uF cap.

The only possible  downsides of this are that you would have no isolation between the pot & the source,so perhaps the input impedance may vary with pot setting,& that you would be adjusting the gain at a fairly low signal level, but you could use another fixed gain FET input stage to fix both of these objections.

[Odysseus]

The textbook method is to use separate resistors for AC gain and DC biasing.

[vk6zgo]

The textbook method is to use separate resistors for AC gain and DC biasing.

Simple & elegant!

[amyk]

For over 80 years,people have realised it is a silly idea to run dc current through volume control pots in audio circuits,so they must know something.

This is an interesting fact, I've never worked with audio circuits much before but it looks strange. Is this an electrolytic effect of some sort?

[iampoor]

The textbook method is to use separate resistors for AC gain and DC biasing.

How do you calculate the gain using this setup?

Seems to be the solution, although my in person results are quite off. It seems to have turned into a rather intense distortion device.

[ejeffrey]

For over 80 years,people have realised it is a silly idea to run dc current through volume control pots in audio circuits,so they must know something.

This is an interesting fact, I've never worked with audio circuits much before but it looks strange. Is this an electrolytic effect of some sort?

No, it is just that when the wiper moves, it 'chatters', like contact bounce on a switch, repeatedly making and breaking contact with the resistive element.  When a pot is configured as a 2-teminal variable resistor, the chatter is like turning the resistor to maximum value and back several times.  This is bad enough when you have just the AC audio signal going through it, but typically not too big a problem -- just a bit of amplitude modulation on the signal until the wiper settles down.  When you have DC flowing, the chattering modulates the 'DC' current, turning it into an AC signal which can get through the coupling capacitor and be amplified, producing the 'pops' that you hear.

Once the noise is there, you can't fix it with filtering.  The chattering is in the audio band, you can't filter it without blocking your signal.  The only solutions are to not put DC across the pot, and where possible to use pots in 3-terminal voltage divider mode.

[iampoor]

For over 80 years,people have realised it is a silly idea to run dc current through volume control pots in audio circuits,so they must know something.

This is an interesting fact, I've never worked with audio circuits much before but it looks strange. Is this an electrolytic effect of some sort?

No, it is just that when the wiper moves, it 'chatters', like contact bounce on a switch, repeatedly making and breaking contact with the resistive element.  When a pot is configured as a 2-teminal variable resistor, the chatter is like turning the resistor to maximum value and back several times.  This is bad enough when you have just the AC audio signal going through it, but typically not too big a problem -- just a bit of amplitude modulation on the signal until the wiper settles down.  When you have DC flowing, the chattering modulates the 'DC' current, turning it into an AC signal which can get through the coupling capacitor and be amplified, producing the 'pops' that you hear.

Once the noise is there, you can't fix it with filtering.  The chattering is in the audio band, you can't filter it without blocking your signal.  The only solutions are to not put DC across the pot, and where possible to use pots in 3-terminal voltage divider mode.

Makes sense. Better quality pots usually chatter less due to a better quality material and closer margins then right?

Right. I was hoping it would be below he audio band but of course, then we would not hear it! haha

Right, by using it in 3 terminal mode chatter is usually reduced because it is acting as a divider, rather then just a variable resistor? (Sorry if that doesnt make sense, Im very tired haha).

Thanks for the clear explanation BTW! Really  appreciate it!

[Mark Hennessy]

The problem is the DC currents cause damage to the track, and it's this that causes the momentary open-circuits when the wiper is moved. Applying switch-cleaner usually improves the situation for a while, but there is no permanent cure. That's why I said above that you will need to fit a new potentiometer along with a DC-blocking capacitor.

Calculating gain is just a case of gm.RL - find gm from the datasheet of the MOS-FET, and when determining RL, remember to account for the load imposed by the following circuitry, which is in parallel with RL. By RL, I mean the 5k1.

This gives you the open-loop gain - that's when the pot is set to minimum resistance. To calculate closed-loop gain, divide RL by the setting of the pot - that's accurate enough for rough purposes.

Interestingly, in this circuit, should the wiper go open, the gain should drop to unity. Did the end of the track get returned to ground, as the schematic shows?

Bear in mind that the DC conditions change when the pot is changed, which would normally be a Bad Thing in an audio circuit, but here, where it's just trying to sound as crap as possible, that's probably an advantage. *shudders*