Carbon Comp ?

[Randall]

Who uses carbon comp resistors in their builds? And why? I just did a build and found at least one expensive cc making a lot of noise. Maybe I overheated it, I don't know. I put them in the plates supply because, just because retro says so. But I replaced one channel with MF, leaving the other channel CC to see what happens. Is there a sonic difference? They just seem like a PITA, and expensive.

[g1]

Here's an excellent article about why, where, and when to use carbon comps: http://www.geofex.com/article_folder...carboncomp.htm

[Randall]

Yes, that article has much informed my use of CC resistors. I only put them on plates, yet still I have trouble. I am leaning away from the CC is good camp.

[bob p]

Resistor Mojo be damned, I hate CC resistors and I won't use them.

They bring lots of trouble along with them. They get hot and drift upwards in value faster than anything else. Of all the resistor choices, they hold their value the poorest over time. Then there is the thermal noise is a problem -- especially with Fenders, which hang the tubes from the sockets, which have CC screen resistors mounted across them. The heat from the tubes rises straight up and just bakes the resistors, and then you're subjected to all sorts of thermal noise. No thanks.

Yes, CC resistors have theoretical mojo, but it's important to remember that the only reason that CC resistors were used in the Golden Age is because that's what they had available. If they had had the manufacturing technology to produce higher tolerance carbon film or 1% metal film resistors back in the Golden Age, then we'd be looking for MF resistors for our vintage amps. Make no doubt about it, nobody would have used CC resistors back then if there were something better available.

As an example of how bad CC resistors can ruin a good sounding amp, I recently brought a dead Super Twin back to life. What was wrong with it? Some of the CC resistors in the Pi filter just failed outright, cracking due to thermal damage. Some of them drifted upward and were out of spec. The net result was that the amp just didn't have proper voltages. A simple problem. Once I replaced the out-of-spec resistors with MF or MO resistors, the amp tested fine. Easy repair. Or so I thought. Then once the amp was up and running, it started making all sorts of random intermittent crackling noises at idle. I couldn't stand it. This is the sort of problem where you could rip your hair out trying to track down the root of the problem if you don't know what to look for. What did I do? I immediately ripped out the 6 carbon-comp screen resistors on the output tubes and replaced them with metal oxides. While I had the sockets exposed, I also replaced the 1k5 CC grid stoppers with MFs. Now the amp is dead silent.

I've been burned by CC resistors enough times that I won't go near them. Even though I have thousands of NOS CC resistors in a variety of sizes, I prefer to buy brand-new MF resistors and use them instead. When I'm working on an amp, if I have to replace a CC resistor it gets replaced with Metal Film. When I'm building a new amp, I use Metal Films, and the occasional Metal Oxide in higher power applications like the Pi filter and screens.

I'm sure you'll hear plenty of people respond who think that CC mojo is real and is worth the trouble, but I'm not one of them. IMO the CC mojo just isn't worth the hassle. Just my biased opinion. YMMV.

[J M Fahey]

They just seem like a PITA, and expensive.

They are

And it's pure superstition.
Sorry.

Strictly applying the Dictionary definition:

su·per·sti·tion (s p r-st sh n). n. 1. An irrational belief that an object, action, or circumstance not logically related to a course of events influences its outcome.

How come so many believe in it?

By (wrongly) reasoning :

*if*
A) Old Fender amps sound great (which by the way is true)

*and*
B) Old Fender amps used CC resistors (also true)

*then*
C) Old Fender amps sound good *because* they use CC Resistors. (Wrong/untested/unrelated)

The real proof lies in doing the "experiment"

ex·per·i·ment (k-spr-mnt)
a. A test under controlled conditions that is made to demonstrate a known truth, examine the validity of a hypothesis, or determine the efficacy of something previously untried.

Which in this case would be:

Hypothesis: "CC resistors sound better than other types"

We must test a few things:

1) that 2 exact same Fender circuits, one with CCR, other with, say, MF , sound different (or not)

2) if there is a difference, that CCR sounds better

To be trusty, specially to avoid human bias, we should apply double blind testing, meaning neither the player *nor the experimenter * must know at any time *which* amp is being played at that moment (they may be behind a curtain and a 3rd experimenter switches them at random, taking note of which is which every time).


If you want to go 1 step further: get something (a pedal/preamp/power amp/mixer channel strip/MP3 player) which never had CCR inside, replace some modern resistors with them, and test for sonic differences, if any.

Unless sonic differences can consistently be found, sorry but the hypothesis will remain in the "untested facts" folder.

jm2c

[Joe L]

I don't use them as plate resistors, but in my own amps I do use a few in the signal path. Why? Well why not? They look "mojo" along with the os blue molded caps that I have collected over the years. And if one of them start getting noisy, I can rip thru a Marshall or Fender vintage circuit in minutes and have the equipment to test the parts.

[kleuck]

They are And it's pure superstition. Sorry. Strictly applying the Dictionary definition: How come so many believe in it? By (wrongly) reasoning : *if* A) Old Fender amps sound great (which by the way is true) *and* B) Old Fender amps used CC resistors (also true) *then* C) Old Fender amps sound good *because* they use CC Resistors. (Wrong/untested/unrelated) The real proof lies in doing the "experiment" Which in this case would be: Hypothesis: "CC resistors sound better than other types" We must test a few things: 1) that 2 exact same Fender circuits, one with CCR, other with, say, MF , sound different (or not) 2) if there is a difference, that CCR sounds better To be trusty, specially to avoid human bias, we should apply double blind testing, meaning neither the player *nor the experimenter * must know at any time *which* amp is being played at that moment (they may be behind a curtain and a 3rd experimenter switches them at random, taking note of which is which every time). If you want to go 1 step further: get something (a pedal/preamp/power amp/mixer channel strip/MP3 player) which never had CCR inside, replace some modern resistors with them, and test for sonic differences, if any. Unless sonic differences can consistently be found, sorry but the hypothesis will remain in the "untested facts" folder. jm2c

I did in my VHT special 6, switching a single metal oxyde for a CC, high value (470K) on the second triode's plate (high voltage too) One of the resistors gave a more organic and compressed sound...bet ? The CC.

[J M Fahey]

I did in my VHT special 6, switching a single metal oxyde for a CC, high value (470K) on the second triode's plate (high voltage too) One of the resistors gave a more organic and compressed sound...bet ? The CC.

If you add a relay inside the chassis, which connects either a CCR or a Metal oxide, same resistive value,

and somebody ouside the room pushes a button 100 times while you play, which either actually switches from one to the other or momentarily switches off wichever is being used but comes back to it,

so you will never know which one is connected at any time and in spite of that

you are right significantly more than 50/50 (or 50% of the time) , as in, say, 80% of the time or better,

*then* I will accept you can reliably hear the difference between them.

Mind you, I'm not even asking for others to agree with you; with an open mind I can accept, as a working hypothesis, that you can hear things others don't, and also don't demand 100% accuracy on your side, simply that you are *way* above chance.

But until such experiment can be made under controlled conditions, I reserve the right to consider such statements as placebo effect or similar.

[kleuck]

Why a relay and why somebody outside the room ? My friend was in the room and i used a simple switch, so what ? I did not know the position, nor did he, and 100% of the time when played crunchy the sound i preferred was with the CC on. It was the first time i deliberately used a CC in a circuit (i sometime use salvaged ones) , for the very purpose of test, i did not expect such a clear result, i had the same in another small amp (not so noticeable, but tis amp is lower in gain, the resistor is of lower value, and the supply voltage is lower too, anyway the added liveliness is of the same kind) but i did not use a switch this time.

[J M Fahey]

Why a relay and why somebody outside the room ? My friend was in the room and i used a simple switch, so what ?

Then your experiment is flawed, sorry.

I did not know the position, nor did he, and 100% of the time when played crunchy the sound i preferred was with the CC on.

You are contradicting yourself: if you didn't know the switch position, how on Earth could you know which is the one you preferred?

[Randall]

well, I just pulled them off my plates, and the amp is quieter. I'm not going to bother with them in the future. there are so many variables in what makes an amp sound this way or that way that I am going to concentrate on the stuff that is quantifiable.

[JoeM]

The CCs on my SFSR (plates) got real noisy making the egg frying sound. I changed to 1W CF (they're cheap and available), and the amp still sounds great, just without the noise. For a high gain amp MF would be a better choice IMO.

[bob p]

this is the first time that I've ever heard someone say that CC resistors cause compression.

i can't come up with a mechanism to explain that. anyone?

[loudthud]

The mojo of CC resistors is said to come from the fact that they change resistance slightly as the voltage changes.

I had a bag of about 20 1/2W 100K 10% Carbon Composition resistors. I measured each one, only one was out of tolerance at about 120K. I chose one that was 100.2K. Then I sorted through a bunch of 100K Carbon Film resistors until I found one that was 100.1K. I soldered one end of both to a tube socket pin 6. The other end was left free so I could clip B+ to either one without disturbing the rest of the setup. A 1.5K resistor with 100uF bypass was used in the cathode. The generator was connected to the grid. A Heathkit power supply was used for B+ and heater power. B+ set to 300V.

Because the distortion analyzer has an input impedance of 100K ohms, I used the signal out of an oscilloscope to acquire the signal from the plate. A 10Meg probe was connected to the plate, AC coupled at the oscilloscope input. To prevent clipping, the scope input attenuator had to be set at 5V/div (50V/div with the 10X probe).

The graphs were identical at 20Hz, 1kHz and 20kHz. The rise in distortion below about 100mV signal is due to noise from the oscilloscope. This is the 1kHz graph.

[Randall]

Well done Mr. thud! Science wins again.