Volume Control, ToneStacks, Potentiometers and noise

[TheGimp]

I'm still working on my guitar amp(s) and have found that one of my biggest problem is noise in the form of hiss.

I've got my second build down to what I consider reasonable. A friend who helps em evaluate the amp considers it much better than his Fender Baseman.

However in my quest, I have started to wonder how much noise a tone-stack (or volume control for that matter) introduces into an amp. And, does the potentiometer type make any difference? That is to say, are there some potentiometers that are quieter than others (given the same resistance).

In the case of the amp I'm working on the tone-stack is after the first gain stage. I can remove the first tube and the hiss starts increasing after I hit the noon position on the volume control. If I put the first tube back in, there is more hiss associated with it.

Is there any kind of standard for the procedure of measuring noise in amplifiers?

So far the best tool I've found is Audio-Tester in the spectrum analyzer mode as it allows me to see the noise floor as well as peaks and noise profile vs frequency.

[RTF671]

Different kinds of resistors will have a big impact on noise. Carbon composition are noisiest. Resistors typically make a hissing sound (like you described). Resistor temperature also plays a part in noise. Overrated resistors (in terms of watts) will reduce noise slightly although changing the type will have the biggest effect. Pots being resistors (typically carbon comp) can add noise. The noise will be cumulative, targeting only the tone stack isn't going to have that much of an effect by itself.

What amp is this and what kind of resistors are in it?

[TheGimp]

The amp is a high gain top boost lead amp which is taylored for small venues and studio use (only 10W).

The design evolved from the Train Wreck Liverpool design (info from The Amp Garage) but has been modified considerably.

The TrainWreck design uses the tonestack following the first gain stage. Since most of the gain of the amp follows the tonestack, noise is an issue.

I'm currently using Carbon Film pots, but am wondering if more expensive Conductive Polymer pots would contribute noticably less noise.

[DF96]

I'm no guitar amp expert, but that is a strange circuit. It uses high gain valves, then throws some of the gain away but in the wrong place (before most of the gain). U1B is a high impedance cathode follower, so the tone stack has to use high value pots so you get more thermal noise than necessary. Then you throw away signal in a potential divider formed by R101 and R2 at the second stage grid (BTW have these resistors been swapped -everywhere else 27K is used as a grid stopper but here it is the grid leak).

My guess is that the signal at U2A grid is not much higher than the input at U1A grid, so both valves contribute to overall noise. You have enough gain that even thermal noise from R2 will probably be audible. You need to get the gain distribution right: more before the tone stack and less after U2A. Try removing C7.

You might be getting noise from the zener at D7, although I assume that under normal conditions this is not supposed to conduct.

[TheGimp]

Guitar Amps are strange animals.

My understanding is that it is common to use excesive gain, then throw part of the signal away. This is done to enhance harmonic distortion at moderate signal levels.


The zener is for protection and should never conduct (or at worst case during power up).

[RTF671]

I am familiar with the trainwreck series of amps. And yes that kind of thing is normal for guitar amps. It's not so much about clean voltage gain as it is shaping and overdriving successive stages with too much voltage input.

May we get a gut shot to investigate further? Carbon pots could be adding noise and its possible that changing them for plastic or other material will reduce it. Do you have any other carbon resistors in the amp?

It's possible that D9 is adding noise in. It's typical (and normal) for guitar amps to have a 10k in series with the diode to limit current going through it if (more like when) the grid gets driven positive.

[ruffrecords]

The first thing to do is short the grid of U2B to ground and measure the hiss at the output. The remove the short, short the amp input, and measure the noise at the output again, first with the volume turned right down and then with it full up. |Report back with those figures and we'll have something to go on.

By the way, the values of the tone stack resistors are not that high and the noise they create should be negligible.

Cheers

Ian

[TheGimp]

The amp is all metal film resistors. I've recently started changing them to 2W even if they don't need it for power dissipation just to see how it effects the amp.

I'll have to pull the amp back apart to get the gut shots as I had it put together to take to the Spartanburg Guitar show (BEE-3 Vintage Guitar Show) in SC tomorrow to get some (hopefully) unbiased opinions.

I'll have the gut shots and frequency sweeps a little later tonight.

I simulated the tone-stack in LTSpice and it looked like about 100K with all controls at noon.

[wakibaki]

Originally Posted by TheGimp

I've recently started changing them to 2W even if they don't need it

Tube amps are pretty robust but be aware that more equipment is broken by unnecessary 'fixing' than most of us care to remember.

That's why they say, 'if it ain't broke, don't fix it'.

w

[TheGimp]

I should have said " ... if they don't need it for power dissipation/reliability purposes".


If it is broke, fix it.

If it isn't broke, modify it.

Changing the resistors was to see if I could measure a difference in noise between resistor sizes. It made a difference in some locations (where large resistance values were used).

Photos are :
1. Gut Shot,
2. Tonestack Disconnected,
3. Tonestack connected and all controls at noon except volume set to min,
4. ... vol set to mid,
5. ... vol set to max

Measurements were made with no input to the amp which has a shorting input through a 68K resistor.

Plots are from Audio Tester V3.

Sound card is M-AUDIO 192.

Input has no attenuation and is taken off a 8 ohm resistive load.