Phase 45 w/Maillet/Univibe mods build report

[audioguy]

I used Torchy's layout- and it sounds great, its going to be a permanent addition to the arsenal.
- one complaint.... the bias knob only has about 10% usability. Meaning that 90% of the sweep is useless- there is no change... but when you get to the last 15 percent of the sweep, you get the effect, but then for the last 5% or so the effect is biased out.
Would a lower value pot work a little better?

Audioguy

[petemoore]

You could change the pot value or taper to get sweep in the full range of the bias, for a boxtop 'sweepcontrol' knob...because varying the resistance Does change the sound of the sweep my 45 has the sweep knob [with only 12%-15% range of the knob settings producing noticable sweep sound] box surface mounted.
 I think the schematic assumes this knob is probably board mounted, or at least meant to be set and forgotten. The wide range [with small usable portion of that range] is probably chosen because when taking into account circuit to circuit inconsistancies, being able to go past ['above/below] the usable ranges [varied resistance that do produce sweep sound] insures that this knob will find a sweep range with Jfet VGsoff and other variations from one circuit to the next.
 I should probably redraft this, but basically they want that knob wide so it'll make the different encarnations of the circuit work, you can trim it down so the range fits your circuit, or one exactly, precisely like it.

[Eb7+9]

This can be fixed by varying the adjusters on each side of the Bias pot ... once you've hit a region within the sweep that does anything the side-adjusters can be tweaked to even out the active region ... if you like it to exhibit a way wider transition, go for a smaller pot value and use higher value trimmers ... my schematic shows a LIN 100k pot with two 47k side-adjusters - instead try a LIN 50k pot with two 100k trimmers and set those so that the full range of the pot now covers pretty much the whole range of active bias ... if that doesn't get you there try a pair of 250k trimmers ...

If you just want to center the response of the Bias pot because it's too low and you don't want to mess with trimmers try adding straight signal diodes (1n4148, 1n60a) in series with the Zener to increase the ref voltage until it's where you want it ... or try a different ref diode ...

the Zener sets the upper limit on bias range and should be chosen according your jFET's turn on needs - and yes, the original circuit wasn't meant to be bias swept for selecting darker/brighter shades of the filters so a change might be in order anyway depending on your devices ... this is also an alternative to consider if you're stuck with FET's that get no action from the full range of  available bias sweep, ie with the top adjuster shorted ...

~jc

[audioguy]

Thats great info. I'll give that a shot!

thanks!

Audioguy

[MartyMart]

All of the above is very helpful, but before you try that please try some
more "pairs" of 2N5457's or 5458's  ( which ever you have gone for )
When I built the "test" version before Torchy posted the layout, he had
also found that selecting a closer "matched" set would greatly improve
the quality of "phase" and "bias" sensitivity also ....
I had one pair that hardly phased at all and after a few swops, I got a
great "sweet" pair with a wonderful range  
I recommend this process VERY highly ......
If you've "been there - done it"  - then please ignore this ....

Regards,
Marty.