Tube power amp simulation

[amptramp]

It is not clear whether you want to simulate an amplifier with self-bias through a common cathode resistor in the output stage or fixed bias through a negative supply to the grids.  The overload characteristics are quite different.  If your output tube requires -14 volts bias and you feed in a signal that that goes positive 15 volts, the grid will act like a diode and conduct.  When the level goes back down, the extra electrons pulled out of the grid will increase the bias and the amplifier may cut out briefly (sometimes called blocking or gulp distortion) with a time constant set by the RC time constant of the grid resistor and coupling cap as well as other source resistances.  If the amplifier is self-biased, this may increase the current through the cathode resistor and whatever capacitor is across it so you have another RC time constant that may overload.  It sounds like the same thing is going to happen in both cases, but consider a waveform with a positive spike and a longer, lower negative side.  With fixed bias, the tube conducting during the positive side will pull grid current but the other tube will not, so only one side blocks and you get a lot of even-order harmonics.  With the common cathode resistor, both sides block because bias on both tubes is affected equally and you get mainly odd-order harmonics.

What do we intend to simulate, fixed bias or cathode bias?

Similarly, the long-tailed inverter (basically a differential stage with a cathode follower feeding a grounded-grid amp) and a triode phase splitter will have different harmonic generation.  Any imbalance adds to even-order harmonics.

What do we intend to simulate, long-tailed or phase splitter or maybe one of the more esoteric ones like the Van Scoyoc cross-coupled inverter or the humble paraphrase inverter with its built-in delay and frequency limitation on one side only?

Also, tube capacitances are quite low and do not vary all that much with the bias point.  Semiconductor capacitances are so high that even MOSFET output stages require high drive current just to charge the capacitances.  Furthermore, the capacitances get higher at lower voltages meaning that a waveform that looks OK at or near DC deteriorates at high frequencies.

Finally, tube gain varies with 3/2 power of bias whereas FET's are square law and bipolar transistors have an exponential change of gain with bias.  You might get something that sounds vaguely like a tube amplifier with solid-state devices, but only empty-state devices give you the sound of tubes.

[tca]

> ... to emulate the sound of a tube power amp.

As someone said, long time ago, tube amps are a continuum of things. You can emulate some of those things, or others, but not ALL those things, but wanting all, you really need ALL the tubes (and btw the output transformer).

Note also that not all tube amps sound the same.

Emulation of "tube power amp" is a ill defined quest.

This question pops up, every now and then, to everyone that builds stompboxes... You need to go beyond that, be creative and use low voltage devices to create something different (it may sound like any other existent device ).

A simple way of getting power amp distortion is to build an amplifier, the speaker plays a important role on the final outcome.

Or, simply, go digital.

> Semiconductor capacitances are so high that even MOSFET output stages require high drive current just to charge the capacitances.  Furthermore, the capacitances get higher at lower voltages meaning that a waveform that looks OK at or near DC deteriorates at high frequencies.

I'm a big fan of high power MOSFETs, they are in a certain way simple to work with and with simple math (only need to solve quadratic equations no need to use the Lambert function). You can work out a SRPP in a A4 piece of paper (anything really, A4 is big).

A simple way of getting power amp distortion is to build an amplifier, the speaker plays a important role on the final outcome.

Another experiment: get a power resistor 22Ohm, an IRF something MOSFET (510, 610, IRFP240, ...) and a laptop power source (18V, 6A). Two caps: input cap 100n-1u, ouput cap 470u-1m. One 10k to 100K resistor and try:



Use any booster/OD pedal as a pre-amp.

Does it sound like a tube amp?

Cheers.

P.S.
The booster/OD pedal will complaint.

[amptramp]

It is similar to a statement made about high fidelity: show me a transistor stage that will swing 50 volts at less than 1% distortion with no feedback.  The 6SN7 can do that.

Also, trying to get a clean push-pull stage with little crossover distortion.  Not difficult with tubes, not easy with transistors.

[FiveseveN]

Yes, it's not as easy if you insist on following the pointless design rule of no feedback. Are we trying to attain high fidelity or some characteristic power tube distortion? If the former, feedback is your friend, and a $5 chipamp will gladly swing 50 V @ 0.1% THD. But I think this thread is about the latter.

[disorder]

The booster/OD pedal will complaint.

How so?

[tca]

^ The input impedance is too low for what a "regular" OD would expect, but the opamp of the Shaka HV would do the trick...

Try and breadboard it... I could draw the complete schematic if you would like to have a go.

P.S.
Or: http://www.diystompboxes.com/smfforum/index.php?topic=107374.msg975651#msg975651

[mth5044]

Quilter:

Once again a complete power amp. This one simulates crossover distortion ("Zero crossing process") and sag modulated clipping at plates ("Soft clipping cell"). Low damping factor is again designed in to the final power amp stage with the aid of current feedback. Again no phase inversion per se but same goal achieved with bipolar reference voltages and diode polarities.

Hey Teemuk, where'd this schematic come from? Is that part of the Tone Block 200?

[J0K3RX]

Modern amps have most of the distortion done with preamp.
My aim is to build the more compact as possible pedalboard to use instead of my amp head + cab for rehearsals or little gigs where I don't want to carry my amp+cab.
I'm building a tube stompbox preamp, I plan to buy a cab sim (ada gcs2?). I'm thinking of adding a compact "power amp sim" stompbox, not really for distortion, but only to sweeten/soften the preamp distortion (I don't know exactly which word to use, you know sometimes preamp distortion can be a little raw and harsh).

- A simple option can be the Runofgroove Thunderbird (only the last stage) with maybe adding teemuk embodiments (but as I'm a newbie I have to understand how all of these work!)
- KMG micropower seems another good option (and maybe if the embedded cab sim is good I can use it instead of buying an ada gcs)
- Johan project could be another good option but initially I didn't plan to use tube to make it as compact as possible. And I already have the "tube sound" from 2 12ax7 in the preamp. The only thing I'm looking to achieve is to add a little bit of the power amp stage feeling in the sound to sweeten the preamp distortion and make the preamp + amp sim + cab sim act more like a real amp.

Run your tube preamp into a laptop and use impulse response with a good software cab sim. Or use something like JamUp Pro and just use the power amp and cab simulation, you can run that on an iPhone!  Or, buy a Torpedo setup. I doubt you will find better in a stomp with all of the options that these would have!

[teemuk]

Hey Teemuk, where'd this schematic come from? Is that part of the Tone Block 200?

http://www.google.com/patents/US20130136278

[Steben]

Why not run a 0,5 - 1 W tube amp (for example with ECL / PCL range) into a dummy load with high impedance output tap into SS?

I think it is the way Eddie Van Halen did it, but smaller.

[cd-rom]

Hello everyone!

I'm looking for ideas and information about tube power amp emulation. There are some brilliant thoughts here regarding the subject by teemuk and others. It's a pity that teemuk's schematic example from the post #6 (first image) is no longer available. I'd really love to have a look at it.

Something I've been cooking up lately. This simulates voltage sag, sag modulated soft clipping, and effects of grid conduction clipping such as crossover distortion. Low damping factor is also simulated.

Wayback Machine doesn't help (recovers only the third image).

If anyone has that schematic, please repost it here or PM me. Any help would be appreciated.

[megatrav]

Something I've been cooking up lately. This simulates voltage sag, sag modulated soft clipping, and effects of grid conduction clipping such as crossover distortion. Low damping factor is also simulated.

Diodes after phase inverter stage simulate grid conduction, the other series diodes clipping at the plates. Reference voltage for this clipping threshold is derived from a differential amp that substracts output of an envelope follower from a steady DC reference. Last stage simulates response of an amplifier with low damping factor drioving a reactive speaker load. The circuit probably needs some tweaking, such as some pregain and some crude voltage limiting so that the input stages sans diode clipping won't get overdriven. Signal should never be allowed to swing above +/-13Vpeaks.

Something Peavey uses ("T-Dynamics"):

This is actually a complete power amp that simulates effects of grid clipping and associated bias shifts (crossover distortion) and the moderately soft clipping at plates. Sag is not simulated. Low damping factor is designed in but it requires that the final power amp section and a loudspeaker load is included. You could likely convert that to run at lower voltages and currents to a simulated high-impedance reactive load. This circuit does not use phase inversion but relies on bipolar reference voltages and polarities of the diodes.

Quilter:

Once again a complete power amp. This one simulates crossover distortion ("Zero crossing process") and sag modulated clipping at plates ("Soft clipping cell"). Low damping factor is again designed in to the final power amp stage with the aid of current feedback. Again no phase inversion per se but same goal achieved with bipolar reference voltages and diode polarities.

I came across a post you made from 2015. I am curious what you found out and if anything ever became of this?