Question for The Tube Experts

Paul Marossy · March 22, 2006, 04:06:11 PM

I have a "stompbox" of sorts that uses a 12BA6 pentode RF amplifier tube, which has about ten times the gain as a 12AX7 as I have it wired up right now. It sounds really great, but there is also a lot of extraneous noise when I'm not playing...

So, here is my question: What ways are there to make the tube appear to have less gain in the circuit than it does (or is there a way)? The part that makes this confusing to me is that when I use a resistor to lower the plate voltage, the current also drops and then the tube runs out of headroom real quick and sounds like crap when the gain is turned up past 9:00 or so. The following stage, a 12AT6 Dual Triode/Duo Diode, just makes matters worse because it amplifies the circuit noise more. There is a 500K gain pot between stages. I was thinking that maybe knocking that down to 50K or so might dump some of the extra gain to ground.

Any suggestions or insights?!

MartyMart · March 22, 2006, 04:18:25 PM

Sounds interesting Paul ... what is it ??
Sounds like you're on the right track, as ever "high gain = High noise" !!
I'm sure Ton will know exactly what's required

Marty.

Headbuttking2 · March 22, 2006, 04:25:07 PM

Can we see a schematic or something....?
Even though I probably have no idea the answer

Paul Marossy · March 22, 2006, 04:34:31 PM

http://www.diyguitarist.com/PDF_Files/RadioHead.pdf

Note that R4 is now a 33K and R5 is now an 82K. The gain pot (R6) is the one that I am considering reducing in size.

TryingToDo613 · March 22, 2006, 04:42:53 PM

You might ask this on the hoffmanamps board Paul. There are a couple of very good guys there. I do know that one of the guys has dropped voltage with a big diode instead of resistors becuase the resistors have issues. It's like a diode that bolts to the case. -ph

JimRayden · March 22, 2006, 04:47:30 PM

As I understand, you want less signal to hit the second stage? If so, why not a resistor between C4 and R6? A 470k resistor will knock off about half of the gain available. Or is that too obvious?

----------
Jimbo

John Lyons · March 22, 2006, 04:48:15 PM

Hey Paul

Probably just need to dump some gain to ground with a voltage divider like you say.
Good place for a treble bypass for a sound option as well.
John

Paul Marossy · March 22, 2006, 04:51:38 PM

QuoteAs I understand, you want less signal to hit the second stage? If so, why not a resistor between C4 and R6? A 470k resistor will knock off about half of the gain available. Or is that too obvious?

Well, I guess that would work as well...

Sir H C · March 22, 2006, 04:53:57 PM

Try shrinking R2 and be sure it is a metal film resistor. Try 100k. That is a noise source and at the beginning of all the gain, it contributes. Since R4 is reduced in size, you can reduce the size of R6 in parallel. Don't make it too small or it will start cutting your gain.

To reduce the gain, you can make R3 two resistors and put the cap between them, much like is done with the Fuzz Face gain control.

Tubebass · March 22, 2006, 05:06:22 PM

One way to reduce the stage gain would be to simply configure the pentode as a triode! Just tie the screen and suppressor grids to the plate instead of wherever they're going now. Of course, this may not give the sound you're after, but at least it's an easy thing to try.

Paul Marossy · March 22, 2006, 05:08:29 PM

QuoteTry shrinking R2 and be sure it is a metal film resistor. Try 100k. That is a noise source and at the beginning of all the gain, it contributes.

OK. But how does this resistor add noise? I don't quite understand...

QuoteOne way to reduce the stage gain would be to simply configure the pentode as a triode!

Yeah, but I like the sound of the pentode.

markr04 · March 22, 2006, 05:10:14 PM

I didn't see a grid stopper in your schematic. Have you tried this? Soldering directly to the grid pin also helps reject RF.

Paul Marossy · March 22, 2006, 05:17:25 PM

QuoteI didn't see a grid stopper in your schematic. Have you tried this? Soldering directly to the grid pin also helps reject RF.

You're right , one's not shown, but I added a 100pF cap to ground right off of the input jack because I was picking up some radio station intermittently!

JimRayden · March 22, 2006, 05:20:23 PM

Quote from: Paul Marossy on March 22, 2006, 05:08:29 PM

OK. But how does this resistor add noise? I don't quite understand...

Yeah, 1M input resistor is quite a noisy thing. You know, the bigger they are...
But I don't think lowering it would be a good idea in this case. a grid stopper would indeed be wiser.

-----------
Jimbo

Paul Marossy · March 22, 2006, 05:21:57 PM

Quotea grid stopper would indeed be wiser.

Pardon my ignorance, but what exactly does a grid stopper do? Is it not the same thing as adding a cap to ground to shunt RF?

will · March 22, 2006, 05:36:02 PM

Hi Paul,

If the sound is great at low volume maybe you don't need so much gain on the 2nd stage. You could just disconnect C9 to reduce the gain of V2. If it is a case of overdriving the input of the 2nd stage then the resistor before the R6 gain pot is also a good idea.

I don't think the 1M input resistor will contribute too much noise as there is no voltage drop across the resistor.

You might consider adding grid resistors to the input of V1 & V2 to reduce HF ringing. 1K directly connected to pin 1 of both sockets.

Regards,
Will

Paul Marossy · March 22, 2006, 05:48:56 PM

QuoteIf the sound is great at low volume maybe you don't need so much gain on the 2nd stage.

I want to agree with that, but all the noise is generated in the first stage. The second stage simply amplifies it some more. But, it couldn't hurt to disconnect C9 and see what affect it has.

EDIT: Hmm... or maybe get rid of C7? That should definitely have an effect on the amount of noise.

markr04 · March 22, 2006, 05:51:24 PM

A grid stopper is a resistor in series with the grid. It's usually applied only to the first stage's grid to lower input and reject RF. A common value is 68K, but you can play with that value and maybe get some results in your particular app. I'm interested in hearing your results with this, should you decide to try it.

Paul Marossy · March 22, 2006, 05:53:00 PM

Thanks for the explanation markr04.

sta63bmx · March 22, 2006, 05:53:10 PM

I could be way off base here, but I think that a grid stopper is in there to prevent high-frequency parasitic oscillations. I think that was the sentence from the Torres book. I think that a grid stopper performs the same function as a gate resistor in MOSFET circuits.

If you DON'T have gate resistors in something like a charge pump/switching power supply, you essentially have this capacitive "tank" on the gate pin where signals are free to bounce around and since the resistance of the lead is low, the resonant frequency is high. They can get bad enough that they'll actually screw up your gate voltage and sent MOSFETs into the linear region instead of being on/off, and then you get resistance and then your transistors melt down and/or latch up and then you will see crazy voltages later on in your circuit way past what your electrolytics are rated for, the electrolytics will blow, your transistors will melt, you will burn yourself poking around in there, and you'll feel like a moron for destroying a bunch of parts...all because you didn't use gate resistors. In some cultures the sky turns black and the rivers turn to blood if you omit them.

Uh....that got off topic. But they are there to prevent high-frequency oscillations and kill RF noise, I think.

Don't ask about the charge pump/gate resistor thing. I've got the TO-220 shaped burn scars to prove it.

News:

Question for The Tube Experts