would this work the same way at parallel opamps? mix-and-match the characteristics of different JFETs for example?
Possible but difficult due to variability between individual devices. If you parallel unmatched devices then an imbalance will result, with one device doing all the work while the other puts its feet up (or something inbetween).
Not to say it isn't worth experimenting with though. You can effectively multiply the transconductance at the expense of higher supply current.
If all are biased separately and AC coupled, you could get some interesting results. Without AC coupling, you could have 1/2 them off doing nothing, quite easily.
I have simulated two paralled fetzer valve type jfet gain stages in ltspice with different model jfets. It seems to work. The clipping is softer looking.
wouldn't it be almost like a darlington?
Quote from: Vitrolin on May 26, 2009, 08:32:01 PM
wouldn't it be almost like a darlington?
that's more like daisy-chaining. 1st output to the 2nds input.
I've seen parallel transistors in high power circuits, like high-current regulators. Often they will add an emitter/source resistor to each one though, to help equalize currents.
Another technique is to linearize differential pairs by having devices of differing "strengths" in parallel to increase the linear range of the diff pair.
Paralleling is done usually for a specific reason, like to decrease the output impedance of a linear power amplifier. I was looking at a Yaesu linear tube amplifier schematic last night (for ham radio, about a kW of power) that uses a common-grid configuration with the two tubes paralleled. This lowers the output impedance of the linear before the output matching network and while you don't have any significant voltage gain, there is an appreciable power gain (~15 dB I think).
Paralleling op amps is typically done to increase power as well, although I guess somebody else has figured out the the different characteristics of each op amp will come through as well. This is no surprise when you look at the probably very similar output impedance of the op amps, the outputs simply mix together and give you a bit of each. I should play around with this more, I've never actually tried it before..
reminds me of "piggyback" (who? me search?) :icon_biggrin:
Quote from: Lurco on May 27, 2009, 02:37:47 PM
reminds me of "piggyback" (who? me search?) :icon_biggrin:
piggyback got me thinking about parallel.
Quote from: aziltz on May 27, 2009, 02:41:18 PM
Quote from: Lurco on May 27, 2009, 02:37:47 PM
reminds me of "piggyback" (who? me search?) :icon_biggrin:
piggyback got me thinking about parallel.
discrete semi-conductors in parallel? got me thinking of piggybacking :icon_lol:
i must admit, i don't fully understand the piggy back thing yet.
Things I must explore:
parallel
piggyback
feedback (non-op-amp)
dip and dots ice cream
There's one commercially offered pedal I know of that uses paralleled JFETs (J201s), and it sounds really cool. I'm not sure how much of the sound comes from having paralleled JFETs, but I imagine that it would theoritically act something like the paralleled input triodes on a Matchless amp for a "chimey" top-boost kind of sound...
Quote from: Paul Marossy on May 27, 2009, 06:59:38 PM
There's one commercially offered pedal I know of that uses paralleled JFETs (J201s), and it sounds really cool. I'm not sure how much of the sound comes from having paralleled JFETs, but I imagine that it would theoritically act something like the paralleled input triodes on a Matchless amp...
i know that bwampler's cranked AC uses parallel FETs. I'm getting a kit next week i think.