How to properly set phaser trimpots

[disorder]

Hi everyone, I just picked up a Maestro PS-1A phaser. I rebuilt the power supply with new filter caps and diodes and would now like to try and understand how to properly set the two trim pots. Here is the schematic. One trim is for oscillator amplitude (depth), the other biases the FETs. Other than plucking my guitar and wiggling both until it sounds correct... is there a proper way I can do this with a multimeter and scope?

I understand the oscillator is just slightly modulating the FET bias... and so it would be good to have FETs biased very close to the middle of the operating range... just need to understand how to find that. Any ideas?

[Govmnt_Lacky]

You did see the adjustment procedure on the very schematic that you linked... right? I think if you followed that you will be happy 

EDIT: Sorry if that sounded harsh. It was not meant to be. I am sure there is an exact science to getting the results you want however, it is a pretty good bet that those results won't be much better than using the instructions on the schematic.

[disorder]

I will likely use that procedure (which is basically what I already did... turn oscillator output off, manually move FET bias trimmer until you find the range that phasing happens, set that trimmer in middle of range, then set oscillator amplitude to taste). Just thought it would be cool to find the theoretical means of calibrating this as opposed to empirical.

[alanp]

The theoretical doesn't really add much.

For bias... this goes back to why six matched FETs are used. FETs have a very, very wide range of spec'ed Vgs. Insanely wide, compared to other parts.

With transistors, a lot of the time you can get away with using a couple transistors on the same tape for a circuit that calls for a matched pair. It won't be amazing compared to a properly matched pair (in a filter ladder, it might be a little bit muddier (not by a vast amount), in an exponential converter, the tracking won't be as good...), but it will do the job. Transistors made in the same batch tend to be fairly close in spec.

FETs, on the other hand, are almost willfully wayward.  R.G. Keen tested a batch of FETs, and got a range of Vgs values from 1.3V to 2.7V. A phaser is composed of a series of all-pass filters, swept by a LFO signal (usually a slow triangle.) That distinctive phaser "sound" comes because all of the filters are moving at the same time -- try using deliberately mismatched FETs in a phaser. It'll sound like either mud, or like no phasing is happening at all. Ensuring that all of the FETs have the same Vgs means that  the phaser filters will all "turn on" (so to speak) at the same time.

Tuning the bias for the PS-1A sets the circuit's LFO dc offset (bias) to match the Vgs of the selected FETs, so that the LFO will move around the Vgs for that set. A different sextet of FETs will have a different Vgs, and therefore the LFO's dc offset will need to be different, to match that different Vgs.

LFO Amplitude is there because the amplitude of the signal that the LFO puts out is bigger than what the phaser filters are all happy with -- clipping is most likely to occur (I strongly suspect that the clipping and distortion will happen at the FETs in the filters.) To get the biggest bang-for-buck, the trimmer is there so that you can set the LFO amplitude (how deep the phaser moves) for as much as possible, before horrible-sound happens.

http://www.geofex.com/article_folders/fetmatch/fetmatch.htm

Disclaimer: I am just a meatworker, not an EE.

[disorder]

I appreciate the reply alan. I did notice that at a certain point on bias trimmer the phased signal starts to take on a large DC offset. See my attached video... This is me probing at pin7 of U3 while slowly moving FET bias from 0V and it occurs around -2V. This also occurs if I set the bias properly (around -1.3V) and then set the osc amplitude too high. Any idea what causes this?