MOSFET op-amps vs Tubes... signal characteristics?

[bluefoxicy]

Someone is telling me MOSFET op-amps (microchip) break up pretty much like vacuum tubes, and I can effectively use them for the same purpose-- that purpose including overdriving the tube/chip to get the same sound.

I thought silicon op-amps clipped and tubes generated harmonic distortion. How am I supposed to get vacuum tube like characteristics (i.e. signal warming in gain stage 1 of the 12AX7; pleasant harmonic distortion in the power amp; etc) out of a block of silicon?

[Enzo]

How am I supposed to get sound out of a dim light bulb with extra wires in it?

[bluefoxicy]

How am I supposed to get sound out of a dim light bulb with extra wires in it?

By putting sound in? The fundamental characteristics of a vacuum tube are it allows electron flow from a negative cathode to a more positive anode, where the relative charge is controlled by throwing a resistor between the cathode (negative source) and the anode (ground) to limit electron flow. A grid in the middle inhibits electron flow.

In silicon you don't have all these free-floating electrons and stuff, you have nice clean semiconductor paths. This should work extremely well until you hit the limits of the device, and then suddenly not work so well.

How are these equivalent?

[Enzo]

My point was that to dismiss semiconductors as a "hunk of silicon" is about like calling a vacuum tube a lightbulb with extra wires. Neither phrase describes the device very well. I hear enough mis-characteration in the political news this election year, so perhaps I am sensitive to it here, where we try to be objective.

Sure they are different beasts, but the FET works differently than the bipolars, and is a lot more like a triode. I am not into FETs myself, but there are a number of guys here who like them, and hopefully will chime in.

But roughly:
The fundamental characteristics of a FET are it allows electron flow from a negative source to a more positive drain. A gate in the middle inhibits electron flow.

[Steve Conner]

No op-amp really has tone. I bang my head against a wall every time I see people swapping different brands of 4558 into their Tube Screamers.

This might sound like a drastic statement, so I'll do my best to qualify it. Basically, op-amps are nearly unusable without negative feedback, and that same negative feedback makes them clip hard and cancels out any pleasant low-order distortion they might have had open-loop.

Any "tone" an op-amp might be heard to exhibit comes from oddities in its hard clipping behaviour: some of them clip asymmetrically, and some older chips even change the sign of their gain, turning the waveform peaks inside out, if you overdrive them hard enough. I forget which ones they are, but I've been tripped up by them before in my day job.

You can add diodes, LEDs, etc, to the feedback loop, but then you're hearing the tonal characteristics of the diodes and LEDs.

MOSFET op-amps are no different, and they're several times more expensive than the normal bipolar and BiFet types. Their main strength is rail-to-rail operation: I've designed them into products that need to run all their analog circuitry off a single 5V rail.

Now, for more pleasant distortion, some folks use CMOS inverter chips as amps. I forget the part number, 4069UB, 74C04 or whatever. Or JFETs. Their curves don't exactly correspond to any tube, but I've had happy experiences with two JFETs more or less just plugged where the triodes should be in a Fender preamp circuit. Driving it with a booster pedal actually sounded quite good. You'd probably be able to tell it wasn't tubes, but it wasn't unpleasant, just a little different. I later tried some mods inspired by the "Fetzer Valve" on runoffgroove, which made it sound a good deal more tubey. (Tubular, even?)

If the silicooties really start to bite, you may be interested in these sites:
http://www.ssguitar.com/
http://www.runoffgroove.com/

I still prefer tubes because they look great at night and impress the layman

[Enzo]

Ah, I overlooked "op amp." I was discussing individual FETs.

I have never heard anyone claiming op amps sound like tubes - regardless of their innards.

[gmoon]

Now, for more pleasant distortion, some folks use CMOS inverter chips as amps. I forget the part number, 4069UB, 74C04 or whatever. Or JFETs.

Craig Anderton wrote an article for Guitar Player mag (Feb, 1977), "The 'Tube Sound' Fuzz", a 741 driving two gates of a 4049 CMOS inverter. I'm sure you could sub another op-amp.

I built one in college (waaaay back), and liked it. But gave it to my roommate for some long-forgotten reason.

There must be similar circuits (I've seen the JFET ones) floating around. I've alway planned to make another one, so I keep a copy of the article handy...

[bluefoxicy]

Ah, I overlooked "op amp." I was discussing individual FETs.

I have never heard anyone claiming op amps sound like tubes - regardless of their innards.

MOSFET op-amps exist. The first op-amps described used vacuum tubes, then transistors. See US Patent 2401779 for a high gain summing amplifier using three vacuum tubes.

http://www.google.com/patents?id=H-N...BAJ&dq=2401779

I'm specifically concerned with MOSFET op-amps, not vacuum tube amps or other silicon or whatever.