Throwing metal film resistors in Fender amps

[Alan0354]

Alan presented a short list of self-noise sources for the component. But a carbon comp resistor has a granular interior, which can become mechanically unsound, and that will cause noise due to vibration and even thermal expansion of the body. Also the wire leads have to bond to the resistive part, and that can become loose.

Well, I just assume things are in good condition, no mechanical problem. Or else, the source is endless!!!

[Alan0354]

What do you mean by 1/f noise ?

Regarding the tubes, I tried swapping with no improvment, unfortunately.

What make you think the tubes are all good? Try pulling out V1, v2 and V3 one at a time, see which one affect the noise. V1 is Normal channel ( no reverb), V2 is Vibrato channel and V3 is reverb. Then concentrate on the one that gives you problem.

[Enzo]

Or else, the source is endless!!!

But that was exactly the point. For a resistor to have a mechanical defect or weakness, it doesn't have to be beat up or in "bad shape." One can never know by looking what the inside of a resistor is doing. Your list was for theoretical noise sources in a component, but they are not the only way a part can be noisy.

[Alan0354]

But that was exactly the point. For a resistor to have a mechanical defect or weakness, it doesn't have to be beat up or in "bad shape." One can never know by looking what the inside of a resistor is doing. Your list was for theoretical noise sources in a component, but they are not the only way a part can be noisy.

So you think the defect resistor is making noise sitting by itself with no vibration. I assume OP had the amp on and heard the noise without touching it or playing through it. If you are right, then it can happen to all different types of resistors, capacitors, transformers, diodes......................... You have to eliminate things one at a time.

Theoretical noise works very good. I might not be working with guitar amp as much, but I did design ultra low noise circuit and had to deal with noise problem all the time.

To me, I am not even convinced totally that OP's amp has a problem. All amp hiss when cranked up, just how much. That's the reason I suggested pulling the preamp tube and reverb tube and try to narrow down the source first. Even if it's a defect resistor, narrow down the possibility before changing everything. Replacing all the components is a dangerous game for a cheaply made pcb. From pictures I saw before, it seemed to be a single sided board with no plate through holes. It would be very easy to pull the traces during soldering and de-soldering.

[frus]

Far as I know there are three type of noise:

1) current noise ( shot noise) where it generates as current passed through a device.
2) Thermal noise where it only depends on the resistance.
3) 1/f noise that is device dependent. It is mainly lower frequency noise. You can hand pick carbon resistor that don't have this noise.

there is more info on types of noise here:

https://web.archive.org/web/20080304...istorNoise.htm

[Alan0354]

there is more info on types of noise here:

https://web.archive.org/web/20080304...istorNoise.htm

Seems like the article call 1/f as contact noise. It did say it's 1/f. The main issue of carbon comp resistors is 1/f noise.

[Mick Bailey]

Yes metal film resistor is less noisy. That is proven fact.
Even better is Bulk Metal Foil Resistor. That is well documented.

And you can see that many manufacturers have started using Metal Film in production amp / audio equipment.

BUT in tube amp, you can also use snubber capacitor to reduce the noise. This is used widely.
3-7 pf between grid and plate is my preferred method. It works very well.
There are several other accepted methods also.
[ATTACH=CONFIG]26902[/ATTACH]

Metal Film has been around for a long time now, but manufacturers have persisted with CF because they're cheap and in a well designed amp don't contribute enough noise to warrant the (then) extra cost, other than in certain strategic locations. I can see that stocks of CF are being reduced (in the UK anyhow) by component suppliers in preference to MF. Eventually I can see CF drying up as it makes more economic sense for a manufacturer to rationalise component stocks.

Amp manufacture is a tiny fraction of electronics production and in a world of increasingly more complex circuits the stability and tolerance of MF is a better choice.

The DRRI isn't a high gain amp and every one I've worked on has been pretty quiet, other than the annoying vibrato tick. A more efficient speaker will emphasize hiss and the speaker characteristics are an important factor, particularly if the upper-end is better represented. I can't agree that taking an established, proven design and changing the resistor technology is a good move. There must either be a fault or less-than-perfect valve choice.

Adding snubber caps to an existing design shouldn't be necessary in low-gain triode audio stages - it may offer a 'fix' by masking symptoms, but I'd rather find the cause and fix that.

[J M Fahey]

Agree.
Killing highs sure will mask hiss, but I wouldn't call that a proper repair.

[Wil]

Seems like the article call 1/f as contact noise. It did say it's 1/f. The main issue of carbon comp resistors is 1/f noise.

By contact noise, should I understand cold solder ?

[Jazz P Bass]

Contact noise is inherent to the resistor.

The boundry between the actual carbon grains can produce a voltage when a current is running through it.
There is also a temperature reliance.
The higher the temp, the more voltage is produced.

In reality, we are talking about nano volts!

If every amplifier that Leo or Jim made (all used carbon comp resistors) had a 'real' issue, they would not have been as successful as they became.

[soundguruman]

In this amp, you will find the quieter operation by adding a capacitor between plate and grid of the first input stage.
This kills the "white noise."
This is the method used by Fender and Marshall in factory production. Either in the layout or the actual PARTS design.
This is by far, the best compromise.
No matter what design, It's ALWAYS a compromise. There is NO perfect noiseless design.

As has been said by me and others, the noise of the first stage is more important because it is amplified by all the following stages.
That "ALONE" is probably the most important consideration of all.
Resistor noise, guitar pickup noise, filament noise, power supply noise.....it's ALL there in the first stage, and then amplified. (garbage in, garbage out)
So it's a very good practice to reduce the noise of the first stage to as low as possible.

The plate and the grid wire located next to each other FORMS a capacitor.
Any time you locate two wires in close proximity, you are also making a capacitor between them.

Example

is layout of Boogie amp. (as much as you may hate it)
The circuit tracks are laid out very close to each other, on purpose. Almost too close for my taste...
This forms capacitor between the circuit tracks, and kills noise and oscillation.
Whether you install an actual capacitor or locate the plate and grid layout next to each other, to FORM a capacitor...
or twist the plate and grid wires together...

In Fender amps, the grid wire is routed right under the plate circuit, to FORM a capacitor. It's ALL absolutely deliberate.

You separate this layout, and the noise is going to increase. The ringing and tendency to oscillate will also increase.

FIRST stage, Metal film resistors, plate to grid capacitance, DC heaters, Humbucking pickup, etc... will all lower the
noise of the first stage and produce a quieter result in the end.

Reducing the high frequency response? It's a necessity. It's a requirement.
The guitar does not go to 1,000,000 Hz. But the amplifier WILL unless you deliberately limit it by design and components used.

In ALL well designed high gain guitar amps, the high frequency is deliberately limited.
If it was not, you would have major irritating noise, ringing and oscillations. The amplifier would be UN-useable.

SEE, you START with a p-90 pickup.
That's the noisiest pickup on planet earth.
Then you amplify it. Then you get NOISE. Of course you do. NOISE IN, NOISE OUT.

Please don't insist on using the noisiest pickup, and then complain about the noise.
and don't blame the amplifier, it's only amplifying the noise, that YOU present to it. It OBEYS.

[Alan0354]

I would not put any cap from grid to plate just yet, that has a miller effect of multiplying the cap by the gain. If the cap can reduce audible hiss, it will reduce the audible highs of the guitar signal. This is particular bad in Mesa. But I think they did that because of the uncontrollable feedback when cranked. To me, Mesa amps do not sound good, you use their cascade channel and roll down the volume of the guitar, it is all muffled because of the filter caps in the signal path.

More important is as Bailey and Fahey said about using a P90 and complain about noise and masking a real problem.

To me, amps do hiss a little, it is not important in real life situation compare to other noise. If it is excessive, something is wrong. That's the reason I suggested removing V1, V2 and V3 one at a time to pin point the noise path. If all three are the same, chances are the amp is ok. You don't have 3 sources going back at the same time. If one one gives you the problem, then chase it down.

[Mick Bailey]

Can you hear any difference in sound quality between the MF and carbon in the two amps?

Not really, I switch between them and think I hear something different but then ask myself if I'm trying to convince myself of something that isn't really there.

[Mars Amp Repair]

I would only add that if a carbon resistor is 'noisey' and not necessarily 'intermittent' as a result of the glued in leads being loose, dropping freeze spray on them can help locate the ones area actually adding 'noise' to the system. BTW, if you do have an old amp that actually has the bad glue carbon resistors, freeze spray is excellent for ferretting them out.

Also you can limit the upper frequency response of the stage (only recommended if you're stage is oscillating) by adding a 'Snubber' capacitor across the plate load resistor. This is not to limit the audible frequencies ie; limit noise, but to remedy undesirable oscillation.

BTW the reason for placing the cap across the load resistor & not to ground is that the cap would need to be rated at the plate voltage plus a safety margin & the power supply on the other side of the plate load resistor is ground to AC signals by virtue of its heavy filtering. So a lower voltage cap is sufficient. glen

[mhuss]

FWIW, 1/f noise is so called because it is inversely proportional to frequency (less noise is produced the higher the frequency, and vice-versa).

I have a mid-90s DRRI and it's pretty quiet. I did add the vibrato-ticking-reduction capacitor.