boss style fets in the cardinal: why can't I do it?

[duck_arse]

so I've been playing with the cardinal, and refraining from making jokes. instead of vactrols, I been trying to get fets to work. so far, I have the left hand section of the circuit below working as quite a nice tremolo, but I want to get it more like the original configuration. using the boss bypass arrangement shown lower right, I get something looking like an asymmetric bowtie on the cro, with masses of distortion. the bad kind.



[ggrrrr] imageshack.[/ggrrrr]

so, does the boss layout have a linear region I can use, or am I doing it wrong? I realise most of the cardinal circuit is missing, and I'll need inverted lfo's, and some values will change, but the parts currently of interest are shown. any help, ideas, corrections?

[midwayfair]

Hrm. I can't help with the distortion issue, but I don't think you'll need inverting LFOs with this set up. The lefthand side shunts to ground when the resistance is at minimum in series on the right hand side. In other words, one's up while the other's down.

But ... this means you have one side working where it just dumps signal to ground, right? And you just need one that gets louder when the resistance drops low? You're in luck: The EA Tremolo does what you need. You can rip out all the LFO-based stuff you have on the right hand side, stick the FET in series with C106 (I'd use a 47uF or 100uF there and make this the all-pass stage). Decouple the chip LFO at pin 7 and feet it to the FET's gate.

Keep in mind that doing this with FETs instead of Vactrols puts you back in the "balancing act" of getting the FETs to behave identically, and/or adjusting gains to compensate for the wider variability compared with the more reliable vactrols. In which case, you have to ask yourself if the chip-based LFO is worth reinventing RG Keen's version.

You'll have to go searching on the other place, but Catalinbread's Paradolia uses FETs in series with the output. There are mu-amps involved. Also, the LFO generator there is an XR2206, but again there's no need for an inverting LFO if you've already got one side working that gets quiet at minimum resistance.

[duck_arse]

bloody ea trem, I can't get away from it.

ahem. I think you maybe misunderstood. I have the lefthand side working, but using the same method for the hi-pass section of the cardinal would need the invert lfo, which I already have on the breadboard. (I now see it depends where the shunt goes, input side or output side.) I wanted to use the series resistance method to replace the shunt circuit, as in the orignal.

I had this warbling today, after various value changes, and have heard and cro-ed the balancing problem to some extent, even with close matched fets. a second trimmer on Vref is no great problem, though.

and while I had the original circuit values on the board, I looked at the output of the hi-pass section, and it had heaps of hummmm modulating the output of the fet. the cap values where increased in the version I listened to in order to get more sig to the fet, so I'd need to recalculate them somewhat.

[midwayfair]

I think you maybe misunderstood. I have the lefthand side working, but using the same method for the hi-pass section of the cardinal would need the invert lfo, which I already have on the breadboard.

Again, it might be helpful to look at the schematic for the Catalinbread's Paradolia -- it does the series resistance method using FETs as the variable resistance element. What I can't tell you is why it works as a source follower in that design and not a drain follower like you were working on.

[amptramp]

You don't always need FET's as a resistance element - check out the Univox U65 tremolo section in the lower left of the schematic where the phase-shift tremolo oscillator is connected to a transistor used to swamp the input when operational:



As amps go, it's pretty dismal, but you can't beat the simplicity.  And a simple stompbox with one switch is plugged in to short the tremolo oscillator output.

[amptramp]

This has all the resistor values for a similar circuit:

[duck_arse]

I have found the caitlinbread thing, which I can't reliably spell even when written down, and at the moment it looks baffling. it'll take a couple of days and a few redraws to make things sensible, then I'll see.

I think I'm nearly there, cardinal-wise, my way, but I'm not yet sure. only at redraw #56 so far.

amptramp, the univox trem method looks similar to the kay t-1. and "tremolo root switch" takes on a whole new meaning with the aust meaning for "root".

[midwayfair]

only at redraw #56 so far

I'm not sure I'd have the patience to do that with someone else's circuit! Thanks for your efforts on this.

[amptramp]

Yeah, I noticed the "ROOT" switch - kind of hard to miss with this size of drawing, but there's not much I can do to fix a pdf.  The tremolo in my U65 worked after I recapped it, but a dual pot in the oscillator may be better to maintain a constant amplitude.

[duck_arse]

constant amplitude from a phase-shift osc? why take all the fun away?

midway, my patience may just result in a pile of paper and eraser rubbings. it's about time I finished something to the photo-ing stage.