Analog Tube Power amp simulator: studying possibilities

cris_mas

Active member
So, here's the deal:

i've tried many digital amp simulations, none convinced me. Fine, so i've found tube preamps, that to my ears are a step ahead.

BUT, yes, but, there's always this feeling that something just doesn't work with digital and my ears. In very specific, the power amp simulation. I believe is one of the hardest points to get right, because you have a physical factor (a big speaker moving air) and also an electronic one, that has some parameters like sag, the impedance curves that are generated by different power tubes, etc. I've tried Torpedo Cab M, sounded "hollow", somewhat didn't closed the sound right. Sounded ok, but something was missing. Had a Helix Stomp, at least in the moment i had it, was an advance compared to previous designs (POD HD), but was the worse of the amp models i've tried. When i changed to a hybrid approach, i was more pleased, but always missing that je ne sais quois factor.

I might be just a crazy bud that wishes an old good tube amp, but i've moved on of that for practical and technical reasons. So, in my quest for a portable rig, miniaturizing has been a search of years and lately i've been toying with this idea: as a reference, is an schematic of the power amp section of a Soldano 50 watter
WhatsApp Image 2024-07-28 at 20.28.27.jpeg
I've done this very basic drawing to illustrate the idea to replace the tubes for FET o MOSFET and keep an inductor/capacitor to recreate the iron at the output. I'm still a noob on kicad to resolve how to make a proper schematic there, but explain a bit further what i want is:

-An analog representation of a tube power amp, in a stomp box format.
-Have a capability of hitting an Impulse Response loader pedal with a better/richer tone, saturation and compression that complements a digital part of a hybrid setup.
-Approach a correct impedance EQ curve that doesn't saturates or overloads the digital converters of the IR Box.

This could in part play a key role on the headroom that I've fought with in my attempts to make it sound authentic, to both record and play direct to PA in live applications.

I know there is an alternative https://axiomeffects.com/products/power-amp-emulator-pae-1
But i think is just too much, i want a simple design, not needing too many options. Besides, haven't seen an schematic of it.

Another possibility is Baja Reactor, but i don't need the speaker simulator.

What can be done here dear community? I would appreciate all help to start a design. I have very basic skill for electronic design but i can manage to figure out things.
 
Part of the sound you're missing might be the transformer. IMO it has the biggest effect on the sound of any part of a tube amp outside the circuit design itself. If you want to keep the portable format it might be worth trying out some small transformers for output coupling, I think Hammond and Edcor have some models that will fit inside a pedal enclosure, and theres the little Xicon ones that cost $2 and are in a bunch of pedals as pickup simulators.
Here's a page I just googled with a simplified marshall 18w circuit in diy pedal format, no transformer but might provide some inspiration to get going.
 
I might be just a crazy bud that wishes an old good tube amp, but i've moved on of that for practical and technical reasons. So, in my quest for a portable rig, miniaturizing has been a search of years and lately i've been toying with this idea: as a reference, is an schematic of the power amp section of a Soldano 50 watter

I've done this very basic drawing to illustrate the idea to replace the tubes for FET o MOSFET and keep an inductor/capacitor to recreate the iron at the output. I'm still a noob on kicad to resolve how to make a proper schematic there, but explain a bit further what i want is:

I strongly recommend breadbord or ltspice first.

There are many threads on this topic and so far the search is ongoing :)
I don't want to discourage you, because I'm interested in such a circuit myself, but maybe it's worth trying something existing? Remember that a power amp simulator doesn't have to be a 1:1 copy of an existing tube amplifier. What's important is what happens under the hood..

Take look at some schematics

Mofeta preamp (starting from R12).

Brownout Overdrive (starting from R15)

Chucks version of the Subdecay Variac (Brownout Overdrive)

Runoffgroove Tunderbird (with nice explanation of the circuit).

DIYStopmboxes thread

DSM Simplifier

Vox AC30 power amp simulation

Another Vox power amp.

Another possibility is Baja Reactor, but i don't need the speaker simulator.
It's not the speaker emulator (cabsim). Quote from the freestomboxes thread "pedal designed to mimic the performance of a typical negative feedback tube amp power stage driving a Celestion quad box or similar". You still need the cabsim/impulse response.
 
Last edited:
Some more stuff from Bajaman. There's power amp sim in every schematic.

 
thanks guys! it never occurred to me that a distortion pedal or preamp could carry the "power amp" part. I will look further into it more, and see what goes on in this. I will share my findings as things come to my brain
You rock!
 
I've been thinking about this a bit today. In order to simulate this could we not actually just make a ~0W power amp? We could even include an inductive load to sink a couple watts if necessary.

JFET / BJT LTP Phase splitter, into MOSFET push pull output, into a transformer with negative feedback and inductive load.

Some circuits for inspiration

1749098629348.png

1749098699531.png

1749098721531.jpeg
1749098741294.jpeg
 
To be honest I think the bottom picture is basically the entire design? The only thing I can't figure out is where the signal exits the circuit? It's laid out in a funny way but it's basically a LTP into push pull. If we could modify it to use MOSFETS or power BJTs for the output to drive the inductive load it could sound pretty great maybe?

The inductive load could obviously be scaled way back too in terms of power handling to make it pretty cheap and small.
 
Last edited:
To be honest I think the bottom picture is basically the entire design? The only thing I can't figure out is where the signal exits the circuit? It's laid out in a funny way but it's basically a LTP into push pull. If we could modify it to use MOSFETS or power BJTs for the output to drive the inductive load it could sound pretty great maybe?

The inductive load could obviously be scaled way back too in terms of power handling to make it pretty cheap and small.

The output power is 100mW. This is a fully functional circuit - I don't see the point of modifying it and messing with MOSFETs, increasing the output power, just to use a load instead of a speaker. In this case, the lower the power the better. You will still need a load if you only want line output, but again - the lower the output power the lower the requirements for coils and transformer power rating.
It's necessary to match the transistors q3 and q4, so it's better to have some spare pieces. I don't remember exactly which model of 2SK170 is needed.. BL probably.
 
Yeah for me personally I would only be using this system line out. I guess my thinking is the reactive load will make the transformer and feedback network a little more amp like. I'm guessing the output power scales with the input? Or is that not how it works?
 
With a reactive load it will be as if a speaker was connected, only without the speaker :)
Reactive load emulates interactoin betwen output transformer and speaker.
Here's the frequency response of the above circuit: with and without the reactive load.

1749122537824.png

I recommend trying another circuit - Baja Reactor.
 
Would baking the impedance curve into an IR using something like the Cabinetron and using that after a preamp be any different from building a Baja Reactor?

At least while the PAE-1/2 are still on the pedalpcb Wish List.
 
Back
Top