A little noisy experiment regarding carbon comp resistors.

[Baguette]

Hello,

Just a quick note to inform you guys about a little experiment I made.

I have a build which is close to a Brownface 6G3 Deluxe Amp with added Presence control. All carbon comps inside.
When I fired it up for the first time, I had an annoying noise problem (hiss + intermittent noises), that I typically associate with a noisy plate resistor or noisy coupling cap (I've had a bad batch of 500pf silver micas that yielded many problems, as an example), or worn preamp tube.
The noise was there even with all controls turned to 0. Weirdly enough, the noise decreased when the Presence was turned up. So I thought the problem was coming from the NFB circuit.

Bingo.

The culprit was the 4.7k carbon comp across the Presence pot to ground.

I tried a couple resistors type and values and here's what I got:

- original 4.7k carbon comp = lots of noise
- another 4.7k carbon comp, same batch = even more noise, unbearable
- 4.7k carbon film = totally noise free
- 4.7k metal = totally noise free
- 6.8k carbon comp = totally noise free
- 6.8k carbon film = totally noise free
- 6.8k metal = totally noise free

Worth mentionning that the 4.7k carbon comp there was making an awful lot of noise and switching to any other type made the amp totally silent.

The morale of this little story : seems like some carbon comps are noisy, while others are not. And this particular resistor in the NFB / PI circuit can be very sensitive to type and composition, just as a plate resistor can be.

Thanks for reading, hopefully this little note will help some of you chasing a noise down the road.

Best,

Victor

[Merlin]

Very interesting- who would have suspected such an innocuous little resistor would be the source of the noise! 

[tony_hunt]

Thanks, that is interesting.
Which Watt value did these carbon-comps have?

[shooter]

anecdotally I have found pretty similar results, I find myself short a resistor, grab a molded oldie and chase noise to that R.  And I second the pF micas, they seem to be really leaky, worse than moldy oldie ceramics 

[Baguette]

All carbon comps were 1/2W.

Those 500pf silver mica that I have are like the plague. Every time I use one in a critical spot like a reverb driver input, PI input or interstage coupling, there's a fair chance the amp ends up with an annoying hiss.
I once had a customer bringing me back an amp I built saying "there's a herd of buffalos running in there". I replaced the 500pf mica cap at the PI input and the noise was gone.
They do fine in other spots though, like in a Fender type single knob Tone control.

As for those bad 4.7k carbon comp resistors, they raise hell in the NFB circuit, but do fine as a midrange fixed value resistor in a FMV tone stack.
Go figure.

[Merlin]

As for those bad 4.7k carbon comp resistors, they raise hell in the NFB circuit, but do fine as a midrange fixed value resistor in a FMV tone stack.

This is because it is 'excess noise', which is caused by DC current. In the NFB loop you presumably had LTP tail current flowing in the resistor, causing excess the noise, whereas DC is blocked in a tone stack.

[tompagan123]

Is there a need for carbon comps and / or silver mica caps in lieu other types?   I think the NOS carbon comps look cool because of the thick color bands, whereas the new ones don't look cool.   Fender amps have a lot of ceramic caps, and Marshalls typically didn't have carbon comps, right?  Just curious if we're into psychological factors w/r/t these choices.

[HotBluePlates]

It's probably mostly psychological. As in, "if I use all the same parts as the old amp did, my new amp will sound good like the old amp." I've been guilty of using parts for that reason before.

There are some aspects to how carbon comps function which may alter the amp's sound. Said as simply as possible, if the effects described actually occur in a guitar amp, there may be a subtle layer of distortion added to large signals (separate and apart from the tubes distorting).

Wirewounds are supposed to have the least noise of common resistors, but large resistances are difficult/impossible to find. I've used metal film resistors in my last couple amps, and don't feel like I'm missing out on anything by not using carbon comp.

You can and should try out different parts to see if it makes a difference to your ear. Doing that, you don't want to have to re-solder anything, because auditory memory is short and the difference (if any) are likely to be subtle. For example, I have picks of a certain shape made of various materials. Someone asked me recently if the material change made any tonal difference, and I couldn't be sure how to answer them, even though I felt there was a difference. I actually had to sit down with 4 different picks and briefly play the same thing quickly switching between the picks to be sure what difference I heard between them.

[tompagan123]

Based on this thread out of curiosity I switched out the mica signal caps for ceramics in one of my amps. (treble tone cap and mid boost cap).   It could be my imagination, but there seems to be a slight difference - for the better.   The values are really close so I don't think it's that.   BTW I was underwhelmed the first time I replaced a 250pf ceramic for a mica.   So the brain is telling me that the ceramic treble cap sounds more desirable.  Crispy is the word that comes to mind.   I've heard others say this, maybe it's power of suggestion.   As far as picks go, I can hear differences, but it may have to do with the way I play as far as  different picking angles, scraping the strings etc.   Fun stuff. 

[Voxbox]

Totally agree that there is difference between disc ceramic and silver mica in the tone stack.
In a 1960 Vibrasonic that I was working on, my preference was for the silver mica caps. I would describe the disc ceramics as "harsh and edgy". But to someone else that would be "bright and detailed".

One man's meat is another man's fish.

I also did an experiment with carbon comps versus metal film in the same amp in the vibrato blender part of the circuit where there are high impedances of 470k in the audio mix circuit. Yes, there is less noise with the MFs but I was drawn to the sound quality of the CCs. It was very hard to quantify. Sweeter, perhaps? It was compared many, many times as I am aware that you can really only remember sound for 20 seconds. I did get really fast at swapping them over.....


Cheers, VB (Speed soldering specialist)

[stevehoover]

Bingo !!!


Thanks BIG TIME for this one! Great catch, turning down preamp and then testing presence pot.
I probably never would have found the noisy resistor on my presence pot.


Couldn't figure out why my current build (Normal channel of a 6g6B Bassman)had more hiss than normal.
After switching 4.7K Carbon comp for 4.7 Carbon film across the presence cap, higher than normal resistor hiss is gone.


YES!......Thanks Again, 

[Mike_J]

I have read in many places that the carbon comp resistor is beneficial in the NFB circuit.  Secret it seems is to use a resistor with a higher wattage rating.  Will quiet the noise.  A 4.7K 2W Allen Bradley carbon comp would be perfect if you could find any.

Thanks
Mike

[HotBluePlates]

I have read in many places ...

Information always varies in quality. You may wish to assess for yourself whether more sources of the same story equal confirmation of fact, or whether it means many repeat the same poor assumption.

I have read in many places that the carbon comp resistor is beneficial in the NFB circuit.  Secret it seems is to use a resistor with a higher wattage rating.  Will quiet the noise.  A 4.7K 2W Allen Bradley carbon comp would be perfect if you could find any.

Did they say why a carbon comp is specifically better there? Merlin mentioned above that the d.c. in the NFB loop probably contributed to the noise.

Don't take my word for it, look at EE Times. Keep in mind the article is written by an engineer from Vishay, who is probably wanting to sell you more of their foil resistors.

But do note the equation in the 3rd paragraph, as it's inescapable. The "T" in it stands for temperature. The theory of "use a 2w resistor" probably amounts to using a physically-larger part to dissipate heat better to attack that part of the equation. A different article claims that cooling a resistor from "... room temperature to liquid nitrogen temperature, its noise voltage is still more than half its room temperature value," which is an enormous change compared to simply using a 2w part instead of a 1/2w part.

So it's much easier just to use resistors that seem to exhibit less noise from the get-go (though sometimes you do have to go to other extremes, like engineering low-resistance circuits to attack the thermal noise equation from a different angle).

FWIW, I looked up Vishay's bulk metal foil resistors at a distributor website. $$$$$$$$!! A 100kΩ 0.6w resistor will set you back $23/each plus shipping (there is a slightly cheaper alternative-package device at ~$20/each). A 100kΩ 1w is almost $54/each. However, looking at Vishay's site, I may have found prices for a series of resistor that's not representative of their entire line of bulk metal foil resistors (they have at least 74 different series of this type made for through-hole applications).

[Mike_J]

I have read in many places ...

Information always varies in quality. You may wish to assess for yourself whether more sources of the same story equal confirmation of fact, or whether it means many repeat the same poor assumption.

I have read in many places that the carbon comp resistor is beneficial in the NFB circuit.  Secret it seems is to use a resistor with a higher wattage rating.  Will quiet the noise.  A 4.7K 2W Allen Bradley carbon comp would be perfect if you could find any.

Did they say why a carbon comp is specifically better there? Merlin mentioned above that the d.c. in the NFB loop probably contributed to the noise.

Don't take my word for it, look at EE Times. Keep in mind the article is written by an engineer from Vishay, who is probably wanting to sell you more of their foil resistors.

But do note the equation in the 3rd paragraph, as it's inescapable. The "T" in it stands for temperature. The theory of "use a 2w resistor" probably amounts to using a physically-larger part to dissipate heat better to attack that part of the equation. A different article claims that cooling a resistor from "... room temperature to liquid nitrogen temperature, its noise voltage is still more than half its room temperature value," which is an enormous change compared to simply using a 2w part instead of a 1/2w part.

So it's much easier just to use resistors that seem to exhibit less noise from the get-go (though sometimes you do have to go to other extremes, like engineering low-resistance circuits to attack the thermal noise equation from a different angle).

FWIW, I looked up Vishay's bulk metal foil resistors at a distributor website. $$$$$$$$!! A 100kΩ 0.6w resistor will set you back $23/each plus shipping (there is a slightly cheaper alternative-package device at ~$20/each). A 100kΩ 1w is almost $54/each. However, looking at Vishay's site, I may have found prices for a series of resistor that's not representative of their entire line of bulk metal foil resistors (they have at least 74 different series of this type made for through-hole applications).

Those prices seem a little high for a resistor.  The 2 watt theory is more of a listen to what sounds best argument.  I don't have any 4.7K 2 watt Allen Bradley CCs so I have never personally tested the theory.  Seems common that a good solution to lower noise is to increase the size of the resistors.  Don't think it matters where metal films are concerned.  Applies more to carbon comp and carbon film resistors.  I try to use 1 watt carbon film instead of 1/2 watt and have had good results with lower noise ever since.  Of course I started using metal film resistors on the preamp and tonestack plates at the same time so who knows how much each contribute. 

Always thought if you find something that works it is a good idea to repeat it.

Thanks
Mike