Gator - next obstacle, and swellers

[blackieNYC]

There are a few threads on lowering the threshold on the Gator.  They all steer towards Scott Lee's suggestion of taking the +9v via a 6M resistor to the inverting input of the first op amp in the side chain (IC5A in this schematic http://gaussmarkov.net/layouts/gator/gator-schem.png
This did not work for me at all, and I don't think the posters from the old thread sounded content with the mod either.
I've got it working quite well, but I do believe I am seeing limitations. Setting the threshold pot clockwise, so that a higher voltage is presented to the sidechain, is lowering the threshold at which the gate will open.  But still some notes from my single coil won't get thru.  Here's what I've tried so far with little or no results.
    Quadrupling the FB resistor R16, reducing the resistor to ground R13 to 1 Meg, increasing C4 to 1uf. -
I would think any of these would raise the gain in the side chain, and get the threshold down to accommodate softer notes.
Why didn't any of these work?

And what will? Any suggestions?
You can't raise the gain at IC1, because you'll distort the LM13600. (matter of fact, even with my low output single coils, I can't turn the gain past 2:00 - the LM13600 will give you a little distorted mosquito singing along with your melody. If you are incorporating fuzz, you might not notice at all.)
i've stopped trying to squeeze a longer release out of it (an earlier thread) Found no answers via keyboard or breadboard. But i really want to lower this threshold, to handle a clean guitar.

Thus far, I gotta swim against the current and say I prefer the Slow Gear.  The Gator has a longer attack than the Slow Gear, but I find the Gator's quicker release a bit frustrating. On the Slow Gear, the attack is quicker, but you can play with it more easily and the release and threshold allow you to swell your first note and keep on playing above the threshold, or damp the strings, swelling each note. With the Gator, you're waiting a good long time for the Attack to finish if you want each note to swell, and if you don't want the following notes to swell, you're going to have problems with that threshold. The Gator is better if you want to have a long dramatic chord swell, but watch out = as your note or chord decays (and the string starts to decay while the Gator swell is just getting started) that quick release/threshold is going to cut you off.
But I still want one.

[Kipper4]

I'll be honest I'm not used to working with OTA. So take this with a pinch of pepper.
I normally work with a more "analogue" manner.
The OTA is looking for currant form the rectifier. I believe you may need to shift your approach form thinking about voltage gain.
All such rectifier designs will have limitations. There's only so much one can do.
Maybe someone more knowledgeable will chime in sorry I can't be more help
Rich

[blackieNYC]

Thanks - Yes it is current the OTA pin 1 is looking for. I don't know how to squeeze more out of the side chain.
Another idea (voltage-wise, still) is to get some gain out of the input, and attenuate the audio input to the OTA.

[Mark Hammer]

I wonder if playing with the diodes will help address some of the problems.  Standard 500-600 Vf silicon types are used throughout.  Is it possible that raising or lowering the Vf in some places or even a single place would be helpful?

[Kipper4]

Hhhmmm Mark might be not something there Blackie.
Time to get the schotkys and leds out on the breader.

[anotherjim]

Hhhmmm Mark might be not something there Blackie.
Time to get the schotkys and leds out on the breader.

I think I know what you meant there Rich, but are you sure it isn't just the kippers you've been smoking?

I'm no OTA veteran either, but that method of working things is just weird - everything I was going to suggest was invalid when I realized the neg supply of the OTA is the part that's swung.

[blackieNYC]

I did string three Si diodes in series across the attack pot. I can go down to schottkies, or try either way in the FB loop diode locations.

Yea - that neg supply thing. Same with the original 3080. Time tested circuit. One thing time had not yet encountered though - me.

[Kipper4]

Perhaps I misunderstood Marks remark.
I was thinking of swapping out the diode in parallel with the attack decay pot.

Smoke me a phatty I'll be back before dawn.

[blackieNYC]

Right, that I've done, but not reduced down to a shortly. Worth a shot, as I don't completely understand the circuit.

[Kipper4]

Measure the voltage with the std diode.  (maybe disconnect the next op amp input before)
Swap it for a schotky measure again.
It might tell you something before you try it connected to the OTA.
I'm just pissing in the wind.
Maybe Mark will pop back and clarify.

[Mark Hammer]

I won't pretend to fully understand the circuit, or OTAs for that matter, but Alan's issue involves an insensitivity to "softer" notes.  IC5A and B each provide some serious gain, so I am reluctant to think this is a problem solved as easily as "undersignalitis" in a Dr. Q, for example, by simply raising the gain of a rectifier stage with a larger feedback resistance value.

IC5A and 5B have diodes in their feedback path.  Regardless of what the gain of each stage is, the diodes have fixed forward voltage.  The feedback diode is flipped around in 5B because 5A is inverting, so in order to affect the same half cycle in 5B the diode needs to "face" the other way.

The question I ask myself is whether the goal of allowing easier re-triggering is achieved by raising or lowering the threshold, via adjusting a Vf at one or more of the three intervention points.  For instance, tacking another diode in series in 5A raises the maximum swing for that half cycle in that stage, but leaving 5B unchanged clamps things where D2 sets them, so monkeying with 5A doesn't accomplish much.

If the Vf of D3 is dropped, then the threshold of conduction for that half cycle is lowered, and the cap allowed to charge up without requiring a larger amplitude for that half-cycle.

The build documentation from PAiA, which can be found here: https://www.paia.com/talk/viewtopic.php?f=6&t=204 indicates that the 10uf cap discharges back through what gaussmarkov shows as D3.  That suggests that discharge time (i.e., decay) might be extended by placing a bit of resistance in series with D3.  I'm just guessing though.

Read through the design analysis on page 20 of the PAiA document.  perhaps it might suggest a remedy.

[blackieNYC]

I did try a large pot in series with the attack decay pot,D3

EDIT - I mean a new pot in series with the diode, and that is in parallel with the attack pot.

[Kipper4]

I misunderstood what Duck said too then because I thought he meant a big pot in series with the diode the in parallel attack/decay pot.
Oh geez.
Never mind I don't think I'm helping.
As you where.....

[blackieNYC]

Page 20 may be telling me to change D2, in the 2nd fb loop. It of course mentions the gain control of the OTA, but this can distort, and is beyond the point where the threshold can be affected.

[Mark Hammer]

I did try a large pot in series with the attack decay pot,D3

If I understood the documentation, the idea was a resistor in series with the diode, not the pot.  The documentation indicates that the cap discharges through the diode back to the preceding op-amp stage.  It would also do so via the pot, but the diode provides a quicker path, since once what is discharging has exceeded the forward voltage of the diode, it can pass through as if it was a zero-ohm resistance.  If one adds a small to medium resistance in series with that diode - smaller than the pot resistance, but big enough to add a couple of milliseconds - in theory it ought to make the turn-off less abrupt.  I'm gonna pull a number out of my, um, hat and suggest that a 10k trimmer might be called for.

I have a boxed Gator at home.  I'll see if I can dig it out and try out what I suggested this weekend.

[duck_arse]

kipper - what Mark said.

blackie - is what Mark said how you wired the pot-to-diode string?

[blackieNYC]

I misstated that - I had put a pot in series with the diode, and that pair of components put in parallel with the attack pot. As Mark said. Had high hopes.
You guys seem to be coming up with the same ideas that I've already tried, so I think I'll walk thru all of them again.  And I have not changed the fb loop diodes yet.

[blackieNYC]

I've tried doubling the Si diodes from 1 to 2, and I've tried replacing them with a single schottky.  I've tried changing the fb cap values in the side chain, to try to get the highs to help open the gate a little more.  I've tried again to get more gain by reducing R13 or increasing R 16.  And I've even tried giving the input op amp a feedback resistor/pot to Vbias gain adjustment followed by an attenuator to the OTA input.  And then added clipping diodes to the newly amplified audio path in an effort to compress the signal going into the side chain. All fun attempts, but no positive results. I could try inserting a gain stage after the threshold pot I suppose.

"It works well if I put a (compressor) in front of it." - I generally consider this statement to be a far cry from "Eureka! I've done it!", and much closer to "This has been miserable failure".   But.. it works well if I put a compressor in front of it.

[Mark Hammer]

So what yoou're essentially saying is: "If the signal level is kept pretty constant, it's a lot easier to find a Threshold that works consistently."?

[blackieNYC]

What? No!  I just...
You.

I just can't play the skinny strings hard enough for this gate, if I go straight in from the guitar.  I still think I'll solder-n- box it.