Replacing Transistor with Tube

[Hatredman]

"then simply add another "battery"

Well, the point is that he does not want to add anything. He wants the same exact circuit. That particularity is what makes everything impossible.

[Thecomedian]

Pardon me if I didn't understand correctly. I assumed he meant "change nothing except the required parts for a tube in place of a transistor", and that he felt he was being told he'd have to redesign the entire circuit (all stages, op-amp, etc) from the ground up which is something balked at.

[PRR]

> too different to work in the same Circuit without anything else changing.

You "can" set up the circuit so that "ANY" device will work.

Here's a design:



This will "work" for any device with these minimum specs:
Mu > 20
hFE > 50
Idss > 1mA, Vto < +/-3V

But...

It is not a great use of any of these parts. The input impedance is lower than a tube/FET allows. The voltage gain is just 9 or 10, which is less than a tube or BJT allows. The current gain is just 9 or10, which is far below the intrinsic current gain of any of these devices.

The distortion is quite small up to a point and then it clips cleanly. No "soft compression" such as many people think a tube offers.

Tube or JFET cathode resistance is ~~1K; input overload typically near a Volt. Transistor emitter resistance (at 0.2mA) is 150 Ohms; input overload uniformly 25mV-50mV. They are very different devices. To work them in the "same" circuit I slagged it with so much resistance that the device differences become small.

> how can I objectively compare the difference in using a tube to compress a signal or using transistors to compress a signal.

Then build the BEST transistor mangler you can. Build the BEST tube mangler you can. For completeness you should fool with FETs also.

All of this will keep you busy for three lifetimes.

And your taste for what is "best" may change in the process.

Note that while amplifiers "do" mangle large signals, this is often not the best way. For predictable steady-state manglement it is often better to use biased diodes. For plucked music it is often fun to have R-C time constants which change the mangle over the duration of a note.

Steal a happy circuit and wring the changes on it.

Ah.... you want a thesis to research and defend. Not a box you will play-through down at the Drunken Frog tavern. Then build the circuit above, abuse it, write-up your findings, and note that the circuit may sound different when optimized for the device. While it may not be a good lesson in "compression", it is still a fine exercise of your fact-finding and writing-up talents.

[Hatredman]

Comedian, I wasn't criticizing you. Sorry

PRR, I think the OP mentioned that he wanted to use some already existing tube compressor, no a specially designed one, and just drop in the FET he had.

[wavley]

edit: never mind

[midwayfair]

PRR, I think the OP mentioned that he wanted to use some already existing tube compressor, no a specially designed one, and just drop in the FET he had.

The Op only mentioned the BYOC 2-knob compressor (which is a Dynacomp/Ross) here on the second page of the thread, but he talked about it specifically when he asked these same questions in a couple BYOC posts. (It's hard to remember what's been answered where when the same question is posted in multiple places.) He may have abandoned this thread because he said he was "having trouble viewing threads on DIYSB" on a more recent BYOC post.

Basically, he wanted to simply drop a tube into the BYOC compressor and was asking where to put it. Then he wanted to replace a transistor in the BYOC, presumably because it was explained to him that a transistor was used as a variable resistor in the BYOC/Dynacomp/Ross. I linked to RG's explanation of the Dynacomp a couple times, but I don't know if he read it or understood it, because he posted again shortly after on BYOC asking how the compression works in the 2-Knob/5-knob. He hasn't posted in either place in a few days, so maybe he's reformulating his question or doing more research.

PRR, didn't you post a variable mu tube compressor design somewhere? I seem to remember your name being used in connection with something I saw or something that was referenced somewhere.

Anyway, wavly brought up a great point when he said:

"Yes, he mentioned compression, but when engineers speak of compression (which we did have that information) they aren't speaking of a compressor like a dynacomp, they're speaking of the onset of clipping where the tops and bottoms of waveforms start being squashed and you have a drop in output power."

Where the op might be completely misunderstanding what the compressors we're talking about are. He might be thinking of general lack of dynamic range when the device is overdriven (which all of us in pedal land would refer to simply as distortion), instead of a device that rectifies the input signal to a control voltage that's used to vary gain or volume. Huge difference between them.