Bias tremolo MOSFET emulation, need advice

[iefes]

Hi guys,

as already discussed in a separate thread in the simulation section I am working on a MOSFET emulation of the princeton bias wiggle tremolo to use as an effects pedal. I got it to work really well already but PRR pointed me to one issue my circuit has and I would like to ask for your advice on how to work around it.

The schematic is attached below.
It's basically a boost running into a JFET cathodyne phase splitter. From the phase splitter each signal runs to one of the MOSFETs. The gate of each MOSFET is biased to a fixed value using the BIAS2 trim-pot (around 2.5V on the breadboard). To create the tremolo effect, the bias voltage is modulated by an LFO. I have used a phase-shift oscillator just like in the EA Tremolo, you can't see it on the schematic. It should put out a sine wave with an amplitude of around 4 - 5 V I think.
The two modulated out of phase signals are then sent to a differential amplifier to sum them.

As I said it works really well using the schematic attached. However, at higher tremolo rates the depth decreases significantly.
As PRR explained me this is because the phase splitter stage has a quite low impedance, so R8 C5 and R5 C4 form filters that roll off higher oscillation frequencies (-3dB at 1.6 Hz). I have tried decreasing the coupling caps to 39n , which helped but I felt that it started to cut down the bass-response of the guitar-signal.

Is there a way to feed the modulated bias to the gates of the MOSFETs without "loosing" the higher frequencies of the modulation at the phase splitter? For example I tried to replace the phase splitter by an opamp that splits the signal and inverts it in the simulation, but this didn't help as well. Frequencies of up to 10 Hz would be perfect.

The circuit works great so far as it is, the sound is nice and very smooth but I would like to be able to use some higher rates without having to adjust the depth-pot too much. And it would be good to have the same choppyness at high rates as I can dial in at low rates.

Any suggestions are highly appreciated. I'm planning to get some boards fabbed, so I want to get the most out of this idea :-)

Ah, one more thing: What value volume pot would make sense after the differential amplifier? I still didn't really get how to choose the correct volume-pot value. At the moment I'm using 10kB which seems to work fine. But 250kB worked fine as well. Then I tried a 100kA pot which seemed to roll off bass frequencies, but this doesn't really make sense to me. So that's why I'm asking.

Thanks heaps!

[j_flanders]

However, at higher tremolo rates the depth decreases significantly.

I would like to be able to use some higher rates without having to adjust the depth-pot too much. And it would be good to have the same choppyness at high rates as I can dial in at low rates.

I don't have a solution for your problem but I have owned a couple of tremolos where I had to adjust the depth according to the speed.
It was most noticeable on my Surfy Trem (bias setting).
If I left the depth at max it would be too choppy (too much off time) at slow speeds.
If I left the depth at the 'correct' max setting for slower speeds, it would be too shallow at higher speeds.
To quote myself:

The seemingly too wide depth range is needed for when you have the speed at higher settings, otherwise you couldn't get a maximum depth at those high speeds.

This post, including a picture, on tgp might explain it more clearly:
https://www.thegearpage.net/board/index.php?posts/24758266/

[iefes]

Thank you very much. It actually is already a nice tremolo effect, so you're probably right and I just should live with it. If I was to sell this thing I could certainly sell this behaviour as part of some special mojo of the effect   

However, if somebody comes up with a solution for this it would be very interesting after all.

[bool]

If you have it working so far, the only thing left to do would be IMO to hand-tweak the two caps to arrive to a "can live with it" compromise. You tried 100n and 39n, and this leaves the following common values to try: 47n, 68n, and 82nF.

Alternatively, you can try to tighten up the values of R5 and R8...

[iefes]

Yep, I think 47n should be a good compromise. In theory there should be no audible bass-cut down to 10n but for some reason for me it sounded a bit thinner. Do I have to take the feedback R into account for the cut-off frequency?

j_flanders I also found your contribution in the surfy-trem topic over at the surf guitar forum where you showed some recording you did with audacity (great idea!). I will try this myself later today and see if it behaves similarly to the surfy trem (what I expect). One problem with the PSO is that the amplitude is lower with extrem settings (very slow or very fast), so this contributes to the problem I am having.
However, lets see how it looks recorded :-)

[Lauri]

I just tested a tremolo circuit based on your design. I used mosfets for the first gain stage and phase splitter because these days a good jfet is hard to find. This is maybe the best sounding tremolo I've ever tried. It sounds very similar to a bias vary tremolo on a tube amp.

Thanks for sharing your tremolo design!

schematic

[Lauri]

Here's a sound sample of the tremolo

[Kipper4]

I like it. it's Swampy.

[Kipper4]

Are r15 r16 really 4 Ohms 7  or 4k7?
What led for D5?
Thanks Lauri.

[Kipper4]

Rv4 depth?

[Lauri]

R15 and R16 are 4.7 Ohms but they are unnecessary and you could just connect sources of Q3 and Q4 directly to ground. I put resistors there because I didn't even breadboard the circuit, I just made a pcb layout for it and wanted to have enough things there to fiddle with in case it didn't work properly.

Any red led should be fine for D5.

RV4 is depth pot.

You can get the Gerber files for my test PCB here

[Kipper4]

Thanks