12ax7 noise.. which side is best?

[guitarmike2107]

I seem to remember some one once saying that one triode of the 12ax7 type valve is quieter than the other side due to the way they are constucted.

Is this true, and if so which side is best.

Reason I ask is I have added an additonal valve to an amp and am only using one triode and its a bit hissy, I have used metal film resistors and have pretty good grounding layout, so the only source I can now think of is a noisy tube, duff resistors or my cable runs are too long, which is unavoidable with the current chassis.

Thanks

Mike

[tubeswell]

I seem to remember some one once saying that one triode of the 12ax7 type valve is quieter than the other side due to the way they are constucted.

Is this true, and if so which side is best.

Lots of people say unusual things. Even if it were 'true' (which I doubt), I wouldn't think you'd notice in a geetar amp anyhow.

[jimboyogi]

According to the data sheets, the triode of pins 6,7 and 8 has a lower noise level than the other one. I always try to wire them up this way, but I have absolutely no idea if it makes any worthwhile difference

[guitarmike2107]

interesting, probably not much in it.

I have used that triode so I guess thats not my problem then.

Thanks

Michael

[Robert M. Martinelli]

According to the data sheets, the triode of pins 6,7 and 8 has a lower noise level than the other one...

Actually, my old, trusty Philips ECC83/12AX7 Data Sheet says :

"With Vf applied to pins 9 and 4+5 and the centre (sic) tap of the heater transformer connected to earth, the triode section connected to pins 6,7 and 8 is the more favourable (sic) section of the triode with respect to hum"

The thermal, "Johnson-Nyquist" noise figure will be the same for both triodes, but, due to the physical construction of that kind of valve, one of the triodes will have less chances to pick up hum coming from an AC-supplied heater.

JM2CW

Best regards

Bob

[guitarmike2107]

In terms of noise and noise rejection would it help if the three pins for the unused triode in my setup were grounded.

[overtone]

guitarmike,

over at the ampgarage trainwreck discussion, I noticed the first triode chosen in their schematics was the 6,7 & 8 side. And I read that Ken Fischer always did it this way because it was "quieter" - but I never read which was supposed to be quieter, hum or hiss. As Robert pointed out, it is probably less hum.

I have tried both ways on a couple of TW "Rocket" circuit amps and did not notice any difference between the two options. Dead quiet either way, but I don't seem to have any golden ears, well I hardly have any hearing at all these days...

Could the additional hiss be purely due to the additional stage, maybe it is amplifying some hardly noticeable hiss that you already have in there? I am not aware how hiss could be added by a little more wire length, is that possible? Maybe someone here can enlighten.

If I was trying to chase that problem down I would swap different valves in there first.
Good luck.

[guitarmike2107]

I have tried several valves, I think that If I now found a low noise valve, it's because its a very good low noise valve rather than a standard one out the box sort of thing. I want the amp to sound good out the box so to speak.

The hiss isnt that bad, it is after the volume control and hardly increases with the volume control turned up. I have just become fussy about these things. I am sure many would say it is quiet. but I have had higher gain amps that are quieter. so thats what I am aiming for here.

Thanks

Mike

[Enzo]

I doubt you'd ever hear a difference. Certainly it would not be a substantial difference anyway. Like chosing competing gig offers, you chose the one that pays more. But if one pays $1500, and the other pays $1500 plus 50 cents, would anyone ever know the difference?

And as I like to preach, it is not about the parts, it is about the circuit. Let's say one side of the tube was a teeeeensy bit less noisy. OK, "one side of the tube is a teeeeensy bit less noisy." And don't call me Shirley. What if the connections to that side have to be routed differently - are the wires longer? DO they go past anything that might radiate noise? Are they closer to something else?

There are always more things to consider.

[Robert M. Martinelli]

If we're talking about hiss, then we have to talk about the valve as a noise generator.

A triode can be modeled as a noise source by thinking about it as an "equivalent resistor", that generates the same amount of noise.

The formula is : R(eq_noise)=2.5/gm

Where gm is the tube's mutual conductance.

In a typical ECC83, the gm is 1.6 mA/V, so its "equivalent noise resistance" is :

2.5/0.0016=1562 Ohms.

The equivalent noise figure for a resistor is en=sqrroot(4*K*T*R*deltaF), where K is the Boltzmann constant, T is the resistor's temperature in degrees K, R is the resistor's value in Ohms and deltaF is the frequency interval we're considering, but, even before you go ahead and calculate this value, I would like you to focus your attention on the fact that, in a typical gain stage, there inevitably are other resistors (grid leak, grid stoppers, anode load resistor ) whose values are way bigger and way more significant from a "noise" standpoint, so their contribution to the stage's total noise is way more significant than that of the tube alone as a noise source. I have neglected the cathode resistor because its value is usually around 1 KOhm in a typical gain stage, and also because, AC-wise, it's often bypassed by a capacitor.

Take note that we're talking about thermal noise, and by using metal film resistors you'll improve only the "shot" noise performance, but the thermal noise figure will be exactly the same, because thermal noise doesn't bear relation to the material the resistor is made of.

If you're SO concerned about noise, you could probably (only marginally) improve each resistor's noise behavior by using 1 Watt resistors, because they will operate at a lower temperature (remember the (4*K*T*R) part of the above equation) thus generating less thermal noise.

Hope this helps

Best regards

Bob

Edit: gm, though known as one of the main valve "constants" is not actually constant, but varies somehow with the operating conditions. The above example used the "nominal" value of 1.6mA/V (ECC83/12AX7), the point being the rather low value of the "equivalent noise resistor" if compared to the other resistances used in a typical gain stage.

[overtone]

(Edit) Ooops, I was slower than Robert. His post sums it up for me. I had written this:

Yes, for sure, not having to rely on a particularly good valve is a good starting point.

If you want to read up more background on noise in guitar amps you could check Designing Vacuum Tube Amplifiers, Chapter 4.1.1 Noise, from page 33, which mentions resistor values with regard to noise etc. Merlin and Tubeswell kindly posted a link to it, on another forum if I remember correctly, although Merlin is not the author. Be patient it is a huge acrobat file, some 318 pages long and takes a while to load. Maybe you have read it already?
https://acrobat.com/#d=UjgTcfpvfqvboIiQLKn5rQ
Designing Vacuum Tube Amplifiers, Charles R. Couch 2009

[guitarmike2107]

There are always more things to consider.

tell me about it.

I will take a photo tonight to show and tell.

Will also try a few other things too.

Thanks

Michael

[R.G.]

The hiss generated by a triode also depends on the operating current. I'd have to go look at the textbooks to dig out how big this effect is.

[Robert M. Martinelli]

The hiss generated by a triode also depends on the operating current. I'd have to go look at the textbooks to dig out how big this effect is.

Me too, R.G.! I raised the "equivalent noise generator" issue because It was the first that came to mind and because I wanted to emphasize that probably the improvement he can get is small, if not negligible...but now you just rang a bell inside my head so I'll have to go and do some memory refreshing....

Only...I will have to wait 'till I'm back home...no tube theory textbooks around here to be found...God knows why!

Geez...I love this forum!

Best regards

Bob

[tubeswell]

Yeah, thanks M8, I like totally forgot about that link