Was curious to see why his thing wasn't kicking over for some folks ... in reading the comments, many of the right observations were present
I think that (i) the (output) gain block is in the wrong place, (ii) a one-sided training wheel won't fix things if the stronger device is on the wrong side ... it needed a very large input (or a booster pedal preceding it) depending on devices used ...
---
the following arrangement should solve these issues, making it easier to build ... while also make it work better as a stand-alone guitar FX ...
(https://viva-analog.com/wp-content/uploads/2020/10/jFET_DOUBLER__REDUX__JCM2020_TINA-Vgs-off.jpg)
https://viva-analog.com/jfet-doubler-redux-jcmc2020/ (https://viva-analog.com/jfet-doubler-redux-jcmc2020/)
should generate a not-too-fizzy up-Octave sound ...
The idea of using a misbiased class a stage made me think of bjts as well, but this did not lead to any satisfying results.
i have tried Keen's jfet doubler and after a lot of tweaking to get the octave i threw it off my breadboard to do something else with my time, it was a good fuzz but i couldn't get the octave working right back then i think it also was because i had unmatched jfets. ill try this one out when i have another available breadboard (all 3 are filled up at the moment).
Quote from: Steben on October 23, 2020, 05:27:11 AM
a misbiased class a stage
i believe that's called class C
cheers, Iain
Quote from: iainpunk on October 23, 2020, 08:26:43 AM
i believe that's called class C
cheers, Iain
True that 😂
Very exciting. Is it so sensitive that we'll need multi-turn trim-pots?
Quote from: daeg on October 23, 2020, 12:34:10 PM
Very exciting. Is it so sensitive that we'll need multi-turn trim-pots?
no, the volume of the signal gets balanced, not the bias point, there is no DC on the pots. and the rest is all self biasing
Good to see you're actually tinkering, not only casting aspersions. I notice you can't quite stop criticizing, though. You need to work on that.
There are several ways to run this thing. The point is, however, to bring up the native nonlinearities of the JFET, not the clipping behaviour. Nor any attempt to fake a full waved rectifier. To that end, you want to keep the JFET in the square-law region, not overdriving it. The phase inverter and opposing JFETs are intended for this purpose.
I suppose I should start sifting through your stuff, looking for "mistakes" I would correct. That would be fun, wouldn't it?
Shall we?
Er, I mean, do you need any help with your stuff? Maybe I can offer some technical advice. :)
Anyone breadboarded it yet?
I did several breadboards of things remarkably like that while I was messing with the doubler. They had various things that they did that might be interesting, but the particular variety of interesting wasn't what I was looking for.
I wonder if J.C. breadboarded it.