The 12AX7 Gain Stage and Boosting Amps

[James Freeman]

Note: Please don't quote this post.
I may (will) edit this post several times to clarify or add things.
Thank you.
--

The 12AX7 Gain Stage and Boosting Amps

I thought it will be interesting to study how a typical 12ax7 gain stage behaves and use that knowledge to boost amps (virtual or real) consciously.

Using this set of calculators from ampbooks.com I could study a typical 12ax7 preamp gain stage and see how frequency response and gain change by adjusting various parameters.

Guitar Amp Circuit Design Calculators

www.ampbooks.com

I also modeled a 12ax7 boost gain stage form the infamous Jose mod in LTSpice to measure it.

Jose Arredondo Mod:

Here's the interesting part;

The Jose Arredondo mod adds a 12ax7 gain stage and a 1Meg Log volume before a stock Marshall Super Lead like you see in the LTSpice image above and the Friedman BE100 amp has this exact same boost gain stage when you switch in the HBE mode.
This 12ax7 gain stage has a gain of about 60 (35dB) and can be attenuated by a 1Meg Log pot afterwards to control the amount of boost which also acts as a grid stopper resistor for the next gain stage that attenuates high frequencies due to Miller capacitance.

Friedman's HBE boost stage has a fixed 1Meg/68k divider after it which is equivalent to about 4 to 5 on a 1Meg Log pot, this drops the output voltage of this gain stage by a factor of 1/10.
*I'm slightly rounding numbers here to make it easier to understand.

Friedman BE100 HBE boost stage:

Considering that we have initial gain of 60 (35dB) from the output and we attenuate it by a factor of 0.1 when the Log pot is at 5, now we only have a gain of 6 (15dB) boosting the next gain stage.
We also roll off some high frequencies because we introduce a large series grid stopper resistor when the pot is not on 10, and we have the small cathode bypass cap and output coupling cap that combined roll off a lot of low end as you can see in the LTSpice plot.


Now we have some knowledge we can use.
A 12AX7 gain stage with 1Meg Log pot set on 5 we have 15db of boost and roughly -6dB/Oct rolloff from 400Hz down, and -6dB/Oct rolloff form 2-3kHz up.
15dB of clean boost without clipping is achievable with 12vDC pedals with a 1vPeak signal at the input and even more clean headroom with 18vDC pedals.

Interestingly, this this frequency response is very similar to some famous boost pedals, probably the other way around, the pedals were designed to sound like a tube gain stage.
For example, the frequency response of some pedals I use to boost amp models in the Helix, a familiar curve.

 

[James Freeman]

I made a mistake in the first Jose LTSpice simulation, I forgot to add a 68k+68k divider after the boost volume, and also the plate resistor should be 300k because there's a 100k dropper before the 200k plate.

So here's the new results with boost volume on 5, 7 and 10 on the second stage grid.

Boost Volume at 5:


Boost Volume at 7:


Boost Volume at 10, a huge 15dB just from 7.


--

This is actually lower than I thought, with boost volume at 7 there is only 12dB of boost at the grid of the next stage and with the 300k plate resistor the first triode is almost clipping.

Here's the voltage/time plot on the second stage grid with boost volume at 7, that's easily within a 9v pedal range, but not above 7.




And of course before the boost volume control you have the full 60v swing.



--

Another interesting thing is in this Jose mod video by Pete Thorn he is using the boost volume on 2 o'clock which is 7.

 

[TSJMajesty]

That's a killer tone! (That's all I got from the post. Wish I did understand electronics...)

 

[Bob Zaod]

I made a mistake in the first Jose LTSpice simulation, I forgot to add a 68k+68k divider after the boost volume, and also the plate resistor should be 300k because there's a 100k dropper before the 200k plate.

So here's the new results with boost volume on 5, 7 and 10 on the second stage grid.

Boost Volume at 5:
View attachment 8372

Boost Volume at 7:
View attachment 8373

Boost Volume at 10, a huge 15dB just from 7.
View attachment 8374

--

This is actually lower than I thought, with boost volume at 7 there is only 12dB of boost at the grid of the next stage and with the 300k plate resistor the first triode is almost clipping.

Here's the voltage/time plot on the second stage grid with boost volume at 7, that's easily within a 9v pedal range, but not above 7.

View attachment 8375


And of course before the boost volume control you have the full 60v swing.

View attachment 8376

--

Another interesting thing is in this Jose mod video by Pete Thorn he is using the boost volume on 2 o'clock which is 7.

Some of them look like a boobie so I'm a fan.

 

[Elf]

This is great! Thanks for posting this. I’m trying to learn LTSpice would it be possible to have the files?

Edit.

 

[James Freeman]

Thanks. Have you actually read a single word of it?

 

[Elf]

Thanks. Have you actually read a single word of it?

Believe it or not I did but reading comprehension is not my forte :) Can I have the files now?

 

[James Freeman]

There you go.

https://drive.google.com/file/d/1HzKJpempCbGNhqt7G4sqDb9fEc8GtTeY

I have included the needed "lib" files so the Triode and Potentiometer work, they are not included in LTSpice.
Drop the included 'lib' in AppData\Local\LTspice (Windows).

 

[Elf]

There you go.

https://drive.google.com/file/d/1HzKJpempCbGNhqt7G4sqDb9fEc8GtTeY

I have included the needed "lib" files so the Triode and Potentiometer work, they are not included in LTSpice.
Drop the included 'lib' in AppData\Local\LTspice (Windows).

Thank you very much! Very kind of you!

 

[TheTrueZoltan!]

Some of them look like a boobie so I'm a fan.

Boobie Boost FTW!