Schaller/Heartthrob LFO workings

[Joncaster]

Can someone maybe shed some light here, please?

I'm trying to connect the Heartthrob LFO to the Phase Splitter network of the Ersatz Fender Vibratone circuit, to drive dual BJT modulators (Harmonic Trem mash-up).
http://www.geofex.com/FX_images/FakeFenderVib2.pdf

I can get everything to work if its DC coupled: straight from the collector of the LFO darlington (Q4), to the Base of the Phase Splitter (that has a bootstrap input network).
To get that to work, I dumped the LFO output network (R7, R14 and C8), and the depth pot.
I can set depth by using small series resistance and variable resistance to ground, but that puts DC on the pot and isn't really the answer, obviously.

I can't seem to get AC coupling to work between the two stages.

Perhaps I need to change the bias points of the phase split?
Is it a cap size thing? I tried 1uf (edited) electro (polarity issue?), as i noted in RG's schem, its unpolarized. Tried 10uF too.
But maybe that's not big enough uF if the Q4 collector is only 2k2.

I feel like I'm missing something regarding the LFO, and I'm not actually sure what some of those components are doing:
This is a 4-stage PSO, right?
What are C8 and R14 doing?

Any ideas on how to AC couple those two blocks?
Been learning as I go.
Thanks

p.s. I hope it's ok to post the link and picture of the two schems?

[duck_arse]

I thimk: from the schaller, delete R14, R7 and points right of the C8. add a volume control to C8, feed your phase splitter via a cap from the wiper of the volume control.

but don't quote me.

[Joncaster]

Will try that, thanks!

In the mayhem that is this breadboard, i might have omitted a cap after any kind of control...that would have certainly messed with things.

[Joncaster]

So I think that:

C8 isn't the output cap from the LFO.
I think C7 is the actual output, through the 47k R14.
This means that the output is biased (correctly?) with 2k2 on collector and 2k2 to ground (R8, R12).

The 50k depth pot is in parallel to the 47k and 2k2, so that works as a divider with R7 470k/330k, so as not to overload the modulator.
C8 seems to be blocking DC to the base of Q1, and allowing feedback around the output, but I don't know why.

I got the AC Coupling working!
If we disregard that whole output network as I have been:
I can't get output to work by taking a cap directly from Q4 Collector, I have to pass it through a volume control first, then into a cap to the phase splitter.
I also need a large output cap, above 10uf...i'm using a 100uf for now.
That seems to work, but i'm sure i'm missing something still...

My problem now is the rest of the chain, I'm pretty sure i'm overdriving the modulators now, as it seems to be a choppy effect, a bit more square than sine wave.
I have to still play around with the biasing of the modulators after the phase splitter.

My other problem I think i have is the relationship of the modulator variable resistance messing with the filter networks of the two input stages.
But one thing at a time.
I'm buffering the input, and splitting the signal into the two seperate stages.
If I have too much trouble with that, I'll have to change to output modulating or bias modulating.

[duck_arse]

there is a catch with your theory - R12 2k2 is not connected to the transistor collector, it's part of the phase shift network. R8 is the collector load for the lfo output transistor. possibly - just possibly, the 47k is providing a positive bias to those electro caps, such that they are actually pointing in the right direction, also just possibly, looking at the original Schaller Tremolo Tr circuit, that 47k resistor is being used to kill everything in sight, osc and chopper, when the unit is "bypassed" - actually more like stopped.

the way that osc is drawn has always bothered and confused me. I sould bread it someday ....

but I still put my [your] money on the 47nF cap being the osc output. using that 47nF cap, you should be able to feed a volume pot [100k?], and a cap from the wiper, but I don't know what max voltage level you will get out of it. then all the rest of wot RG has from the pot wiper and 1uF [C16] cap.

[Joncaster]

That makes sense, thanks.
I think it's a bit of a touchy circuit, but it does have a nice rhythm to the wave, which is why I started down this road.
It turns out this is a long road.

[duck_arse]

this just in: the 470k and 47k load the cap, filtering the osc output so that the lower freqs look like ugly triangles. the output of the osc is quite low at the higher freqs, and quite large at the lower, but for the drag of the 47k.

when you lift that 470k//47k, the output jumps to around 8V p-p, but the waveform has a flat arse. it still drops in level as you go higher. the original schaller circuit also shows a centre off schnell//slow switch, giving a greater spread and overlap on the speed pot. I measured 2.9Hz to 25Hz [where the osc starts to die] with the values on the redrawn circuit.

the blokes at schaller propbably knew what they were doing, their osc works with their trem, but you'd be better off with a standard pso like in the EA trem, or as in RG's circuit, if you want to feed a phase splitter and stuff. at least we know what to expect from the standard circuit [and how to fiddle its range].

[Joncaster]

this just in: the 470k and 47k load the cap, filtering the osc output so that the lower freqs look like ugly triangles. the output of the osc is quite low at the higher freqs, and quite large at the lower, but for the drag of the 47k.

when you lift that 470k//47k, the output jumps to around 8V p-p, but the waveform has a flat arse. it still drops in level as you go higher. the original schaller circuit also shows a centre off schnell//slow switch, giving a greater spread and overlap on the speed pot. I measured 2.9Hz to 25Hz [where the osc starts to die] with the values on the redrawn circuit.

the blokes at schaller propbably knew what they were doing, their osc works with their trem, but you'd be better off with a standard pso like in the EA trem, or as in RG's circuit, if you want to feed a phase splitter and stuff. at least we know what to expect from the standard circuit [and how to fiddle its range].

Wow this is awesome! Thanks for scoping it. Interesting that is so squashed and almost doing its job under duress haha.

Its been a few evenings of hair pulling and head scratching trying to get it to interface with the phase split and the two modulators. A lot of that was just not knowing what the hell I'm doing.

I was victorious this afternoon, got it to wobble rather fetchingly on both sides.

After throwing resistors and caps at it, I've come back to a pretty straight forward thing going on.
At some point, I'll do a schematic and pop it up here for some feedback.

Now to tune it!

[Joncaster]

 I was also reading that a Darlington stage has a DC offset, and that has an influence on symmetry, something like that.

So they maybe picked the darlington instead of an emitter follower for low output Z, and got a wonky wave in the process, and then brute forced it out of the divider?

[Joncaster]

Ok, so the 470k/47k is also biasing the modulator base to about 0.45V.
I was having problem with the modulators on the other side of the splitter shutting off and acting weirdly.
I didn't fully understand what these bjt modulators needed to stay happy, but I managed to bias them to that same region, seem to be happy.

I'm pretty sure the phase splitter looks like a very different load than just a bjt modulator used to, so that's also changing the output of the LFO im sure.
Not quite sure what to do about that, but at least I'm slowly getting closer.

[bool]

That 47k/470k/47n network looks like a "designers' addon"; a tweak circuit; probably developed empirically, looking at the scope and "by ears" - for the best vibe and player interaction.

[Joncaster]

That 47k/470k/47n network looks like a "designers' addon"; a tweak circuit; probably developed empirically, looking at the scope and "by ears" - for the best vibe and player interaction.

That's a nice way to view it, because it definitely changes the "vibe" of the throb when I mess with it.
I was just making a note to try some lower values there, etc.
Now that it's decoupled from the modulator, there's a lot to fine tune.

[Joncaster]

I got everything sounding nice with the BJT modulators, but now I've got some LEDs and LDRs on the breadboard just to see.

I've got them to glow and do their thing in the circuit, but I'm not sure about this:
When the depth control is at minimum, the Led is lit halfway in its available swing, and as I bring the depth up, the signal gets louder, because it's lighting up (and going darker) around that point.
Is that the normal way to have an LED flashing from an LFO?

Do I need to bias the LED to a certain point?
I'm trying to learn about modulating LEDs but can mostly only find on/off switching explanations.

So now the chain is: LFO into depth (volume control) into Phase splitter, into bjt driving LED (emitter follower config).

LDR's are from signal input to ground (between an input buffer and the two amps)