Practical application of noise reduction

[Gibsonman63]

You guys have been really helpful in the past helping me with my Fender amps and I have gotten tons of great information reading through the threads.

I have read the rather exhaustive and somewhat humorous thread on noise reduction in high-gain amplifiers. What I glean from it is that most are in agreement that metal film resistors of larger power ratings will reduce the hissyness.

I am using a Marshall JMC800 Model 4210 (50 Watt, channel switching combo with one 12" Celestion) as my main gigging amp. I really love the sound of the amp, but it does get hissy on the higher gain settings. I have already changed out the power supply filter capacitors and tubes and rebiased the amp. Prior to this the amp was anemic, so I am happy with the results, but better is better.

The specific componants mentioned in the previous thread were the grid-stop resistors. There was also some reference to changing out electrolytic capacitors to capacitors with lower ESR.

Rather than arbitrarily changing out componants in my main gigging amp, I would really appreciate some input or guidlines on replacement of specific componants that would give me the most bang for the buck without changing my overall sound.

[Gibsonman63]

A little research

I took a read through Mr Aiken's site and found some useful recommendations under the question and answer section.

Plate Load Resistors - 2W 750V Metal Film
Power Supply Decoupling Resistors and Screen Grid Resistors - 5W 1% 750V Metal Oxide

I am having trouble sourcing resistors rated for 750V, but at least I have a starting point.

[defaced]

Not that this will be too much help, but I have often wondered how much of the hiss was design related. I have built two high gain circuits, a VHT Deliverance and a Soldano Avenger (SLO 100 overdrive circuit). Though the VHT isn't working 100% perfectly, the VHT is significantly less hissy than the Soldano (this has been my experience when I've played them in stores too). Why? I don't know. I'm using the same brand/size/rating of filter caps, resistors, and coupling caps.

[Joey Voltage]

I took a read through Mr Aiken's site and found some useful recommendations under the question and answer section.

Plate Load Resistors - 2W 750V Metal Film
Power Supply Decoupling Resistors and Screen Grid Resistors - 5W 1% 750V Metal Oxide

I am having trouble sourcing resistors rated for 750V, but at least I have a starting point.

It's not necessarily the wattage that's the important aspect as far as noise is concerned, It's the physical size of the resistive element. Typically the larger the resistor, the more it will suppress contact noise. Generally you would expect a 2W resistor to be larger in size than a 1/2W one, but this is not the case a lot of the time, so the 2W rating is meaningless in this application.

[Joey Voltage]

Not that this will be too much help, but I have often wondered how much of the hiss was design related. I have built two high gain circuits, a VHT Deliverance and a Soldano Avenger (SLO 100 overdrive circuit). Though the VHT isn't working 100% perfectly, the VHT is significantly less hissy than the Soldano (this has been my experience when I've played them in stores too). Why? I don't know. I'm using the same brand/size/rating of filter caps, resistors, and coupling caps.

could be a lot of things honestly, most super high gain amps like that will have some degree of hiss, and yes some of this is design related, and some of it is component, and layout/lead dress/grounding related (not so much hiss on the second set, but noise certainly). Sometimes it can even be tube related.

If you examine both topologies, you will see that the S*l*o may be slightly noisier by design

[Gibsonman63]

Successive Aproximation

Thanks for the replies. I ordered some 2W metal film resistors from Mouser so I can experiment with the first stage to see if I can reduce the noise. It is not terrible and not very noticable when I am playing live with drums and bass, but it is more noise than I would want in the studio for recording.

[Joey Voltage]

Thanks for the replies. I ordered some 2W metal film resistors from Mouser so I can experiment with the first stage to see if I can reduce the noise. It is not terrible and not very noticable when I am playing live with drums and bass, but it is more noise than I would want in the studio for recording.

well like I said, the 2W power rating on most newer parts is really meaningless, and you really want to consult a spec. sheet for both dimensions, and voltage rating. If you are trying to minimize contact noise by using a larger part, it really becomes an exercise in futility if the new part is no larger physically than the one you are replacing.

[Gibsonman63]

But wouldn't changing to metal film resistors be quieter? ... at least theoretically?

[Joey Voltage]

But wouldn't changing to metal film resistors be quieter? ... at least theoretically?

The metal film part yes, the higher power rating no, unless the power rating actually corresponds to the physical size of the component. on a side note, you may not notice a huge difference in noise floor when comparing a 1/2W metal film, to a 1/2W carbon Film.

[Robert M. Martinelli]

"It's not necessarily the wattage that's the important aspect as far as noise is concerned"

Joey,
allow me to respectfully disagree on this statement of yours.

If we're talking about thermal, or Johnson-Nyquist noise, at a given value all kinds of resistors have the same noise figure, expressed by the equation En=Sqrroot(4*K*T*R*deltaF), regardless of the material. By examining the equation, it is clear that, for a given value of R(esistance), in a given frequency interval, (deltaF), its thermal noise is directly proportional to the Boltzmann constant, (K), and to the resistor's temperature in degrees Kelvin (T), therefore, using a higher wattage resistor is the only way to reduce thermal noise, all other factors being equal (unless you have the power to reduce the Boltzmann constant ), because the resistor will then operate at a lower temperature, thus producing less thermal noise.

If we are talking about contact noise, then a difference indeed exists between the various kinds of resistors, carbon comp being the noisiest and metal film being the less noisy.

My advice would then be to use over-dimensioned ( say, 2 times the needed wattage ) metal film resistors in the most important resistors' positions within a gain stage, which, BTW are the ones needing relatively high values of resistance ( see the very exhaustive recent thread on this matter ), though, with the soldering iron in my hands and the amp gutted out on the bench, I'd probably swap all of them just for the fun of it anyway.....

( is there something wrong with me, doctor? Pleeeaaase, tell me!.... )

Hope this helps

Best regards

Bob

[Joey Voltage]

"It's not necessarily the wattage that's the important aspect as far as noise is concerned"

Joey,
allow me to respectfully disagree on this statement of yours.

If we're talking about thermal, or Johnson-Nyquist noise, at a given value all kinds of resistors have the same noise figure, expressed by the equation En=Sqrroot(4*K*T*R*deltaF), regardless of the material. By examining the equation, it is clear that, for a given value of R(esistance), in a given frequency interval, (deltaF), its thermal noise is directly proportional to the Boltzmann constant, (K), and to the resistor's temperature in degrees Kelvin (T), therefore, using a higher wattage resistor is the only way to reduce thermal noise, all other factors being equal (unless you have the power to reduce the Boltzmann constant ), because the resistor will then operate at a lower temperature, thus producing less thermal noise.

You do have a point, and I retract, however this would mostly be relevant in the circuit where temperature is a factor. In a typical signal path, it would probably be best to reduce the resistance instead in order to reduce thermal noise.

If we are talking about contact noise, then a difference indeed exists between the various kinds of resistors, carbon comp being the noisiest and metal film being the less noisy.

My advice would then be to use over-dimensioned ( say, 2 times the needed wattage ) metal film resistors in the most important resistors' positions within a gain stage, which, BTW are the ones needing relatively high values of resistance ( see the very exhaustive recent thread on this matter ), though, with the soldering iron in my hands and the amp gutted out on the bench, I'd probably swap all of them just for the fun of it anyway.....

( is there something wrong with me, doctor? Pleeeaaase, tell me!.... )

Hope this helps

Best regards

Bob

No question about resistor type, however (and this is what I was mostly trying to point out) using a physically larger resistor also helps reduce contact noise, and most people buy higher wattage resistors thinking they are going to get a larger footprint which isn't necessarily true, so the 2W rating is kinda useless for this application. you can buy a 2W resistor and be disappointed when you find that it's no bigger than a 1/2W.

[Gibsonman63]

Marginal Results

I replaced the plate resistors with Metal film. Mr. Voltage is correct, the new resistors were only marginally larger than the old ones. Everything is the original values except for the plate resistor for the post gain output buffer (V2A) which I reduced from 150 to 100 ohms.

The clean channel is nice and quiet, but I have a still have a steady hum in the gain channel at the same frequency as a B on my guitar. The amount of hum is directly related to the amount of gain.

Another observation is that I seem to have more high treble on the front of my notes now. I am sure it was there before because BFG style pinch harmonics have always been effortless with this amp, but now it is more pronounced. This is directly related to my pick attack on the string.

It is still a great sounding amp. I have been gigging with it, but I am careful to immediatly switch back to the clean channel between songs. I would still like to reduce the hum on the gain channel. I looked at the heater wiring and it is run with all of the other wiring to the tube sockets, so it looks like a challange to twist those wires and I didn't want to do too much at one time without playing the amp for a while.

Is it possible that reducing the plate resistor value is contributing to the additional high end? Any suggestions on reducing the hum on the gain channel?

[Robert M. Martinelli]

.....Another observation is that I seem to have more high treble on the front of my notes now. I am sure it was there before because BFG style pinch harmonics have always been effortless with this amp, but now it is more pronounced. This is directly related to my pick attack on the string.

Is it possible that reducing the plate resistor value is contributing to the additional high end? Any suggestions on reducing the hum on the gain channel?

Hi Gm63,

1) Yes, as reducing the plate resistor value results in a gain reduction, and the Miller effect ( the multiplication of the input capacitance by the stage's gain) gets reduced too, reducing the stage's input capacitance (which, together with the grid stopper resistor forms a Low-Pass Filter), thus allowing more HF to pass through.

2) What about running the preamp heaters by means of a well-filtered DC supply?

Hope this helps

Best regards

Bob

[pdf64]

Have you tried some different tubes / tube brands in the distortion channel - a bad tube might pick up hum, some brands are worse than others?
Just moving the grid wiring well away from the heater wires can reduce hum interference.
Of course the hum might be 120Hz from the B+, rather than 60Hz from the heaters. Which might be a bad B+ cap, poor ground return connection etc.
See
Excessive Hum

[Gibsonman63]

More contemplation

Thanks everyone for the suggestions. I need to use the amp tonight for my gig, so I won't have time to do anything major until next week, but I do plan on changing the plate resistor back. I may even try going the other direction a little to see if I like it.

I replaced the filter caps and tubes in this amp about six months ago. Before then, the amp sounded anemic. The 1rst stage preamp for the clean and crunch channels share a common tube (V1A/B). Is it possible that one side of a dual triode is noisier than the other side or do they tend to be even for both sides?

There seems like I lot that can be done on this amp on the lead dress although I am a little leary about doing it for fear of making things worse. All of the wires to the tubes are run together and the heater wires are not twisted which is very different than my Fender amps. Am I missing something?

I have read good and bad about running the heaters on DC. Good as far as sound and bad in terms of implementing it. I'll do a little more research on the forum.