Question for teemuk - Valvetronix

[jazbo8]

I was looking at the schematics for the older AD100 and saw it used HV and an actual OPT to emulate the power amp distortion, but on the newer designs, the 12AX7 runs on low voltage without the OPT. If you have any information or insight on this design change, could you please share it? Also, why use a triode preamp tube to simulate the power amp distortion, couldn't Korg use at least a low power pentode?

[Enzo]

I am sure teemu has more insight into this than I, but my impression would be mainly one of cost. If the tubiness of the sound comes from the tube rather than the transformer, why add the expense of a transformer? Likewise, if one dual triode makes the tubiness - and not least providing the justification for having a tube in marketing - why use two pentodes in push pull? Or even a single one? Pentodes cost more and take up more space.

I seem to have the newer one, since I have no transformer in my schematic. But I note the 12AX7 seems to be running on only 50v. I don;t know, but starving a triode might be easier than starving a pentode. Not my area of expertise though. But since the solid state power amp uses 51v rails, using one of those existing supplies for the tube eliminates the need for a winding on the power transformer, and all the power supply components of the high voltage supply. A considerable savings.

[teemuk]

on the newer designs, the 12AX7 runs on low voltage without the OPT.

The voltage actually varies quite a lot depending on model. Some use only 15VDC while some use about 50VDC. Since the point is to get those tubes to distort by clipping I don't really understand why folks obsess about "proper" high B+. Even many modern all-tube designs now incorporate circuitry that deliberately decreases supply voltages to achieve earlier overdrive.

And yes, you can perfectly run tubes at low plate voltages with proper circuitry. Believe me, Korg has that covered and has patented a lot of related stuff.

The transformer really isn't all that neccessary. It really just sums two opposite phase signals together and reflects secondary load to primary. It's very easy to fake its operation with operational amplifiers. The "virtual transformer" circuit is kinda difficult to explain but luckily Korg patents (e.g. United States Patent 6,350,943) cover a lot of Valvetronix -related stuff and are definitely nice read with nice illustrations of circuit concepts and theory behind. My book also had a chapter about Valvetronix design, and in the 1st edition it's all about the second revision of Valvetronix... the one without that MOSFET PI, high B+ and real transformer.

Do note that pretty much all Valve Reactor series products have used somewhat different circuitry. The original blues were what they were, the following AD series was different, VR series was different from the aforementioned and even the ToneLab pedals had some minor differences. I don't even want to try to cover all that in a forum thread especially when schematics for just about all but the newest Valvetronix amps can be found for download with just a few minutes of searching from proper places, like this forum or Valvetronix forum.



Why did they not use a pentode? Well... The 12AX7 is a dual triode so for making a push-pull amp it understandably made more sense as a device choice than to use two pentode tubes. ...And pretty much all what enzo wrote. Higher voltage design is more expensive, and power pentodes cost more than preamp dual triodes.

Vox actually uses a pentode in the newest of the newest Valvetronix amps but to do that they have to employ an automated switching scheme that chops off parts of signals coming from PI, sums together only the halfwaves of interest and then post the pentode gain stage does the same in reverse, generating two "half wave" signals, which are then summed with a transformer or other suitable circuit. The concept is actually quite simple but in practice it's a pretty complex setup for achieving something that, IMO, could just as realistically be replicated using nothing but solid-state devices ...or if you really want to put a tube in it, those preamp triodes. ...and a single dual triode. There's a lot more to tone of the Valve Reactor circuit than the type of tube used in it, but folks like to obsess about tubes and a real pentode inside definitely makes better sales than all-solid-state circuit even if anyone couldn't tell the difference eyes closed. It's not always things that make sense that dictate how things are designed.


Got to go now. And damn this forum has run slowly for the past week.

[jazbo8]

Thanks guys, what you said makes good sense which I often lack Cost reduction is always on the minds of the manufacturers, so eliminating the HV supply and transformers no doubt played a key role in the circuit topology. While working/playing with tube amps all the time, I overlooked the fact that tube emulation technology has come a long way, and there are in fact many amp sims and/or pure silicon designs that do away with tubes completely - I guess, it is inevitable, perhaps a few decades from now, no one would bother with designing or building tube amps, because the sims will be so good, and not even the golden-ears and tone-snobs can tell the difference from the real ones...

[jazbo8]

@teemuk - I just read your book's chapter on valve emulation, combined with your comments above, it would seem it is more about pre and post distortion equalization, asymmetrical clipping and not so much about emulating the tube's unique transfer characteristics, nor OPT saturation, etc. All those cherished "tube amp characteristics" turn out be less important than initially thought. Interesting! Great book btw - very well done.