Double Triodes in parallel

Hey folks,

I'm planing a Siemens SM204 build and spoke to a really knowledgeable person who advised me to not build it in the original configuration (6072 in single triode with the other side grounded) but to build it with both triodes in parallel. In conjunction with a lower ratio transformer (7:1 instead of 9:1) this should decrease noise a bit (I read only by about 3dB) and it should also be better for the tube as the unused side can somehow deteriorate after some time.
Also he said, that the Siemens in original configuration goes kinda deep in the lows and "rumbles" as a result. This is not the case with the tube in parallel.
But I believe the folks at AKG had a good reason to run two of their top mics in single triode mode(C12, Ela M 251).
Or maybe not?
So I just wanted to ask for some opinions around here who have experimented with parallel triodes and also maybe a hint on why AKG did it the way they did it.

I always wondered the same thing as you (see here), but Jakob might be onto something in relation to grid current. But we'll never know unless somebody tries  .

Oliver Archut used both triodes on CS-1. I think Brian Fouxman has a paper on his site about this as well.

Yeah, the man who told me that, also said that they made a lot of tests and also prefered the parallel triode. But I read also a few comments about single triode sounding "better".
But don't have the both halves to be perfectly matched? Otherwise wouldn't it introduce distortion?
Also how important is the mid heater tap? Not all double triodes have it. 
As I'm also looking for substitutes for the 6072 as the prices really have gone up and good specimen getting scarce...

It all depends on tube and circuity. 12a*7/ECC8* double triodes not always works superior in parallel mode.
I have pretty good experience with submini double triodes connected in parallel. Definately worth to test.

Is the purpose of paralleling to reduce Miller capacitance?

If so, then I would assume the capsule capacitance, gain of the amp, the amount of miller capacitance in the tube, etc would be some of the determining factors in how paralleling affects the mic. Would it not also change the plate to grid "feedback" path of the tube? If that's the case, it completely makes sense why some like or dislike paralleling as more feedback generally means cleaner gain and more low end and some prefer the higher distortion of less feedback in an amp.

Delta Sigma said:

Is the purpose of paralleling to reduce Miller capacitance?

Click to expand...

On the contrary. Cin is doubled, rp is halved, the same gain, the noise is lower for 3dB.

I don't see how unused triode (no applied heater nor B+ on it) can deteriorate after some time.

Delta Sigma said:

Is the purpose of paralleling to reduce Miller capacitance?

Click to expand...

On the contrary! It doubles Miller capacitance, and I bet that's the reason why not many use parallels. Indeed the circuit also determines the capactance. Cascode and cathode-follower are good at eliminating Miller capacitance.

moamps said:

On the contrary. Cin is doubled, rp is halved, the same gain, the noise is lower for 3dB.

I don't see how unused triode (no applied heater nor B+ on it) can deteriorate after some time.

Click to expand...

The gain only stays the same when lowering transformer ratio, right? Otherwise you would have less gain because of higher input capacitance.

Concerning detoriated unused triode; this was not something the man said but something I read in the internet. I'll try to find the page.

Murdock said:

The gain only stays the same when lowering transformer ratio, right? Otherwise you would have less gain because of higher input capacitance.

Click to expand...

Depends. You should calculate in lower rp too.

AusTex64 said:

I think Brian Fouxman has a paper on his site about this as well.

Click to expand...

Do you mean Brian Fox or Mark Fouxman?

abbey road d enfer said:

Do you mean Brian Fox or Mark Fouxman?

Click to expand...

LOL! Elderly moment. I meant Brian Fox. Thanks for catching that Abbey.

Would it not be true that an unused cathode would still emit electrons from heating, but only half the expected plate area for attraction from the vacuum?  I don't know if that's entirely true, or if it could be a problem.  I'm sure there are people who've experimented with leaving any possible sections out.  There's a number of tubes more recently used in power amps where they use screen grid as plate to get different characteristics, can't recall what they do with the plate, float or connect elsewhere. 

The groove tubes 67 used the screen grid instead of the plate, a 20k load resistor and only a 5:1 transformer, that thing was pulling a Q of 6ma. And it sounded like pants. Supposed highend lift which just sounded thin, though that was probably the resonator ring. Very disappointing sounding mic that was.

The original C12 circuit is to me perfection in simplicity. The 12VR uses a dual triode as a boost/pad however you see it / and is an excellent sounding mic, considering the the bad rap it's nylon capsule gets, but I still found myself switching the second one out. Then we got a real c12 and the VR never got a look in again.


Isn't the 204 just a rebadged C12 with a red strip on the grill?


Was that person Jim Williams by any chance? He really doesn't like the 12:1 transformer, because it technically doesn't pass low end very well - especially the Haufe. However I
Can say absolutely that the original C12 running in original config into the haufe sounds sweet as can be, and missing low end is not something you'd say about it. Maybe the CK12 has a bass bump without it, but I've used several different T14s in builds from peluso and cinemag which are both excellent, but the haufe just sings. Couldn't tell you why.

The dual triode maybe more efficient but wth a different transformer you'll be altering the sound. Also you're adding a whole second cuircuit, I think the the reason AKG only used half was to keep the design simple and manufacture costs low. And I think that that adding a second amplifier into the mix is adding another chain in the link which may become weak.

Just my thoughts, but there are 4 real top end LDC and no one has topped them for sound *or* durability

T

Last year I restored a SM204, you can find all the necessary info in this topic, mine was really in bad condition, so I had to start almost from scratch(luckily the transformer was original, but the capsule was dead):
https://groupdiy.com/index.php?topic=62801.0

Matchless DC30 is a go to amp for many Nashville guitar players.  It's kind of an AC30 but it has the 1st and second stage tubes run in parallel.  I believe that has something with the preferred quality of it.  I've only played through one  once.  It had a very unique sound.

emrr said:

Would it not be true that an unused cathode would still emit electrons from heating, but only half the expected plate area for attraction from the vacuum? 

Click to expand...

That would be the same as having an amp in stand-by mode, wouldn't it? I don't see how it would impair the life expectancy of one section vs. the other. Cathode sublimation would be identical (maybe a tad better, even).

abbey road d enfer said:

On the contrary! It doubles Miller capacitance, and I bet that's the reason why not many use parallels. Indeed the circuit also determines the capactance. Cascode and cathode-follower are good at eliminating Miller capacitance.

Click to expand...

Why would that be so detrimental?

If we spitball a 50pF capsule into a 150pF triode with Miller capacitance (assuming 30dB of gain in the triode), we have a -12dB loss due to voltage division/padding.  Doubling the Miller capacitance yields 4dB less signal due to padding, however output impedance is halved, so we can reduce gain loss in the transformer by using a lower ratio.  Seems like it might be a wash across to me.

abbey road d enfer said:

That would be the same as having an amp in stand-by mode, wouldn't it? I don't see how it would impair the life expectancy of one section vs. the other. Cathode sublimation would be identical (maybe a tad better, even).

Click to expand...

I don't think it would impair life expectancy, the question would be if there's any performance change due to half the expected plate size, or from the other angle, double the cathode emission. 

Matador said:

Why would that be so detrimental?

If we spitball a 50pF capsule into a 150pF triode with Miller capacitance (assuming 30dB of gain in the triode), we have a -12dB loss due to voltage division/padding.  Doubling the Miller capacitance yields 4dB less signal due to padding, however output impedance is halved, so we can reduce gain loss in the transformer by using a lower ratio.  Seems like it might be a wash across to me.

Click to expand...

What do you do about signal-to-noise ratio? Gain is cheap, there's plenty of it in the mic preamp. But signal lost right at the source means more noise at the end of the chain. Not mentioning distortion increase...