How crucial are these resistors in Maestro Phaser design?

[disorder]

https://s3.postimg.org/edrbcjuib/maestro.png

The 100k resistors that connect from pins 3 and 5 of the phaser stage opamps to ground. The ones in parallel with the JFETS. How crucial are they?

I ask because I have been beating myself up trying to fix one of these pedals that has a bad DC offset thump sound when the oscillator is on. DC thump sound occurs in frequency with oscillator. And this pedal does not have them. In my unskilled opinion I think these resistors are very necessary as they provide ground reference to the input pin of an opamp stage! The input cap will hold charge there throwing off bias of the stage and the output eventually latches... right? This pedal has never had these resistors installed??? You can see it on the circuit, the holes for the resistors are there but the copper side of the board shows resistors never existed!

[thermionix]

From what little I know...the JFETs are just there to act as variable resistors, and the 100k resistors in parallel might not be necessary with certain JFET sets, apparently that was the conclusion Maestro came to if they left them out of some units.

My uneducated guess is that it wouldn't hurt anything to try adding the 100k resistors.  No idea if that might have any chance of solving your thumping problem.  I guess you already tried tweaking all the trimmers?

[disorder]

The thumping occurs when the JFET's turn OFF, I think. I can disconnect the oscillator and manually bias the JFET gates and at a certain point around -2VDC at the gates, this DC offset starts creeping up on the opamp input pins and of course their outputs and it just continues down the line of stages until it hits the output.

I've tried setting the bias and oscillator amplitude trimmers jussssssst right such that it doesnt get near the thump area but it's really hard to do and the phaser sounds weak and anemic in that area.

[Rob Strand]

Well I'd say they are necessary.   In order to get a good sweep you usually drive the JFETs close to cut-off and because of matching issues you may need to let one or more get cut-off.  That occurs on the negative swing of the LFO.  If one of the JFETs cuts off and those resistors are not present then, like you observed, the biasing of the opamps cannot be maintained.

Leaving them off is just asking for trouble.

To test the theory, you could increase the bias voltage a small amount to get it out of the cut-off region but the result might be weaker phasing.   (That's if you want to upset the settings.)

(Another possible cause of thud is if the LFO swings too far positive and drives the JFET gate into conduction.)

[Edit:  you just responded while I was typing.  I guess that answers the question!]

[disorder]

Well I'd say they are necessary.   In order to get a good sweep you usually drive the JFETs close to cut-off and because of matching issues you may need to let one or more get cut-off.  That occurs on the negative swing of the LFO.  If one of the JFETs cuts off and those resistors are not present then, like you observed, the biasing of the opamps cannot be maintained.

Leaving them off is just asking for trouble.

To test the theory, you could increase the bias voltage a small amount to get it out of the cut-off region but the result might be weaker phasing.   (That's if you want to upset the settings.)

(Another possible cause of thud is if the LFO swings too far positive and drives the JFET gate into conduction.)

That makes sense and that's exactly what I'm seeing. I can bias around the thump but then the phaser is dull sounding. This thing has been bumming me out for awhile, I hope this fixes it.

[disorder]

Adding these resistors fixed my issue. I have another Maestro PS-1 phaser that also is missing these six 100k resistors. Also looks like the resistors were never installed, not like someone just ripped them out.

[Rob Strand]

Adding these resistors fixed my issue. I have another Maestro PS-1 phaser that also is missing these six 100k resistors. Also looks like the resistors were never installed, not like someone just ripped them out.

Thanks for the feedback.

It probably worked on one day on one sample then someone decided to rip them out (of production) to save on parts.