Cascading gain stages

[BenHoper]

Hi there. I've built a first small amp from the bits of a Pilot radio, 6Q7GT into 6V6. All works fine. Now I thought I'll add another gain stage for some more adjustable break up as a learning exercise. Now I come to the point I'm missing.

With a 6Q7 having my of 70, the overall gain will be 30ish. For a 0.5V guitar signal I now have 15V on the grid of my next preamp valve. Too much!

So what do amps normally do? I've looked at Champ 5E1 and they just have a 1M vol pot between the two stages. How does this work, how do you get the gain down? Just a potential divider? Using lower gain valves would help, but the Champ uses 12ax7.

Any clues?

 

[rayma]

How does this work, how do you get the gain down? Just a potential divider?

Yes, use a high value volume pot after the first stage.

 

[Enzo]

Or two resistors as a voltage divider. Done all the time in commercial amps.

 

[PRR]

You usually figure a guitar amp so that, full-up, 20mV in makes full output.

If you play 500mV(0.5V), either you expect heavy distortion or you turn it down.

Alternate clue: *every* small amp from Champ to Junior has *two* high-gain triodes in front of the 6V6/EL84. Study! Copy! Steal!

While 6Q7 has an idealized gain of 70, in real circuits it may be closer to 35.

6V6 likely needs 15-20V peak grid drive, 10V-14V RMS.

Present driver gain of 35 thus needs 280mV-400mV.

Another stage of the same thing, gain of 35, you need round-about 10mV at its grid.

This is actually "too much gain" for many players.

If you look at other two hi-gain triode g-amps, some omit preamp cathode resistors to get gain down (and save pennies), others pad between stages, and Fender AA-Champ adds a mild-loss tonestack and takes a little NFB around the output. Many ways to tweak gain.

 

[Gnobuddy]

How does this work, how do you get the gain down? Just a potential divider?

I'm not entirely sure I understood your question, but in case I did: attenuation through a potential divider is also "gain". It just happens to be a gain that's less than unity.

If you have a gain stage that multiplies the signal by 50 times, and then a volume pot set to multiply the signal by 0.1 times, the combined gain is the product (50 x 0.1), which is 5.

Alternately, you can look at this same situation in decibels - gain is +34 dB for the gain stage, -20 dB for the volume pot, and the total gain is the sum (+34 - 20), or +14 dB.

In audio engineering in general, there is no mathematical difference between "gain" and "volume" knobs, as both of them simply multiply the incoming signal by a number.

In the world of guitar amplification, though, the word "gain" is often (mis)used to mean "the amount of nonlinear distortion". So we turn up the "gain" knob, turn down the "volume" knob, and end up with more distortion than before, because some part of the audio chain has been driven into overdrive / distortion / clipping.

Ironically, when you drive a circuit into clipping by turning up the "gain", it's actual mathematical gain (output voltage / input voltage) decreases!

-Gnobuddy

 

[BenHoper]

Not got much time tonight but thanks for the replies.

PRR - yes two gain stages of 35, for 12V at the 6V6 gives 10mV at the input. The effect I’m seeing is sever clipping and it sounds like crap. In fact my latest tweets have producedd an LFO! I think I assumed too higher out put for a guitar so I’ll try and recalculate it all, reset the bias points and, hopefully get more sense out of it.

Thanks all.

 

[Gnobuddy]

The effect I’m seeing is sever(e) clipping and it sounds like crap.

Then by all means lower the gain to better suit your guitar, and your playing technique!

Typical Fender "classic" guitar amp topology has a very lossy tone control after the first gain stage, usually followed by a volume control, and then the second gain stage.

Between the tone control insertion loss, and the volume control attenuation, most or all of the gain from the first gain stage disappears before you get to the grid of the second gain stage.

I've had some discussion on the "How big is a guitar signal?" question, and the answer seems to be "Anywhere from 20 mV pp to 10 V pp". Play a single string lightly on a guitar with low-output single-coil pickups, and you may get only a few tens of mV output. Strum aggressive six-string chords on a guitar with high-output humbuckers, and you might see 2 V pp. And at least one person demonstrated (with oscilloscope screen-shots for proof) that he could get 10 V pp straight out of his guitar pickup. 😱

I have at least one guitar (Ibanez AS73 with stock ceramic-magnet humbuckers) that will quite easily overdrive the typical half-12AX7 input gain stage in my Princeton Reverb (Rg=1M, Rk=1.5k, Ra=100k). The only cure is to turn down the guitar's volume control, or otherwise attenuate the guitar signal before it reaches the amp.

-Gnobuddy

 

[thoglette]

Have a look at these explanations of how the JCM-800 circuit and the Soldano SLO circuits work. Lots of tricks to avoid having high-gain sound like crap.

In short, a whole bunch of things have to be done differently to a standard "Fenderish" amplifier

 

[BenHoper]

thanks

Have a look at these explanations of how the JCM-800 circuit and the Soldano SLO circuits work. Lots of tricks to avoid having high-gain sound like crap.

In short, a whole bunch of things have to be done differently to a standard "Fenderish" amplifier

Thanks, the Robrobinette page was very helpful.
I think one of my problems was that there was some hum in my original circuit (1 pre-amp) then when I added the second, this oscillated. Will look again.