MXR Phase 90 with switch, volume and depth

[Mark Hammer]

sorry for OT, but just wanna make sure to prevent tons of debugging, crying, swearing ...

are the JFETs you intend to use matched? just wanna make sure... 

I actually wanted to ask about that... to make them matched is it enough to be the same thing (for example 4x 2N5952) or they have to be "specially" matching?

Matching is critical, proportionate to the width of sweep. 
ALL JFETs will have a drain-source resistance that can be varied.  The question is "How much?" and "with what current applied to the gate?".  Unmatched JFETs - even use of different kinds of JFETs (2N5952, 3819, 5457, K-30, etc,) within the same quarter - can produce phasing.  But once a bias voltage in common is found, how much can each of them change drain-source resistance beyond that?  They might all be able to show change with gate-current for a narrow sweep, but once you go beyond that narrow sweep this JFET stops changing then that one and the other one.

So, if a person was making a unit that was intended to produce a fast bubbly-and-narrow sweep, to mimic a fast-speed Leslie, matching would not be super critical.  Unselected JFETs would likely "work".  If the goal was to also be able to produce ultra-wide slow sweeps into the stratosphere, with visions of Alex Lifeson, JFETs would need to be matched, to assure that every unit is able to keep changing drain-source resistance over the entire sweep cycle, and not just for parts of it.

Note that it is not important for the actual drain-source resistances to match.  What produces the effect is the summed phase shift across all stages at frequency X, not what happens within each stage.  What IS critical within each stage that the JFET drain-source resistance - whatever it happens to be - is always changing in response to the LFO.  So what you're selecting for is equivalent changeability and responsiveness.

[Eb7+9]

3 - Depth knob. I read here https://www.diystompboxes.com/smfforum/index.php?topic=86588.0 changing the 1M resistor for a 1M pot should control the depth (purple dot on the image, lower left corner).
     Is it the right resistor?
     Must it be a linear pot?

it's not a very common mod // but it works ... though, only to some degree on its own

notice the voltage sent to the jFETs happens to be formed by a voltage divider set up by the (said) 1Meg resistor and the 3M9 leading to the big LFO cap ... this voltage divider divides both the DC and AC components seen at either end of the divider ... at one end is the bias trimpot with a DC voltage and no AC voltage component ... at the other end, at the (said) big cap, is a DC voltage that is roughly at half-rail and an AC component referred to as the LFO signal ...

since the likelihood that the trimpot (bias) voltage lies exactly at half-rail is low to non-existent you will very likely have different DC components at both ends of your divider ... which means that if you make this divider variable (either by making 1M variable, or combining both 1M and 3M9 into a 5M pot) you will be introducing variable DC along with variable AC ... which means, your Depth control will shift the biasing of your FETs - and so, won't be perfect in a sense ... in that case you might want to make your bias control externally accessible so you can optimize the response after adjusting the speed ... it's a drag but there's no way around that, save re-designing the whole thing

[Mark Hammer]

That's why I suspect that any attempt to increase sweep width via reducing the 3M9 resistor from the LFO, should probably be complemented by being able to increase the current-limiting resistance off the bias trimmer to more than 1meg, such the the JFETs don't get pushed into a range where one or more simply stop sweeping.

So, if sticking with 3M9 from the LFO, one can probably be comfortable with 510K in series with a 500k pot, coming off the bias trimmer, since we know that the standard 1meg fixed resistor works.  If one drops that 3M9 down to 3M3 for a wider sweep, then it may be wise to use 680k in series with a 500k pot for the bias/offset control, to compensate for the higher current coming from the LFO.

The over-arching objective is to be able to get some variation in sweep width, and where in the spectrum the notches are created, without having to pay too big a price for it.

[ghiekorg]

Hi guys, just a quick question. I am still waiting for the components but today some of them arrived. I mounted the trimpot but then i realized maybe i have mounted upside down.
Here is the factsheet http://www.uk-electronic.de/PDF/Copal%20FT63.pdf
and the product should be this:


i guess also with trimpots the 1 usually go to ground... In that case i placed it wrong (i just followed the leads. in my trimpot the nr.2 is centered and above the other 2 legs, while on the layout it seems the opposite).

Do you think i can just leave it as it is and "set it" in reverse or shall i unsolder it and place it properly? I am having a lot of problem with soldering, i can't understan why. The tin is always sticking to the iron and creates spikes/cold joints even if i keep it 5 seconds on the components. Moreover the copper is "peeling" off the board in some points (maybe i am overheating it). I never had this problems before...

Thank you guys

Nevermind, i flipped it

[ghiekorg]

hi guys, i started to solder the pots but i realized i am not 100% sure they are correct.
Is this image correct? https://postimg.cc/2qfDmzN1
they are all facing "up" (the shaft is up)
Thank you. That would be really helpful to avoid a lot of troubles later

[ghiekorg]

So... the stuff arrived but the pedal doesn't work. I only get dry signal. Maybe i did a mess with the dry-wet mod? All other pots are pretty much doing nothing, just slight volume changes here and there.
Can you guys please help me?
Thank  you
my bad, reversed 5457 now it works, i will make a clip

[Mark Hammer]

You won't get any phasing unless the bias trimpot is set appropriately.  Have you attempted to adjust it?

Also, I will note that, against all mathematical probability, it never ceases to amaze me just how many different pinouts there can be for the same 3-pin FET.  Get the datasheet for the brand of JFET you are using,and verify it is installed correctly.  I've had nonfunctioning phasers in past, only to discover the FETs needed to be rotated 180 degrees.

[ghiekorg]

So...here is a clip, i still have to box it.
https://youtu.be/21G8BTi3Xfs

Everything seems to work. just 2 marginal notes:
- as you can see the feedback goes crazy when passing a certain limit (which is expected) and in general doesn't really change the sound that much (but i didn't notice a bit change also when watchng comparison of the script vs non script)
- the sweep doesn't really change much the sound either, but probabily i should listen at both of them more carefully. It actually work good but seems to change the Q of the frewuency more than the depth. i will post a video when i box it
- The rate pot is already soldered so it's outside of the camera view
- the dry/wet doesn't really make it 100% dry, there is still some effect left even on 100%
- The Offset sounds to me more like a depth control (shouldn't the sweep control the depth?). but watching again the video it seems to actually effeect the bass frew a bit. does it? It seems to work like a "detailed" trimpot. I set now the trimpot so when the offset is on 0 e get just a bit of effect on lowest frequencies so when i use the pot i can move the freq center around. Before i had it half way so when on 0 the effect was not present (like having the trimpot wrongly set)

small edit: my quartet are 2.5mA 1.3V. I used a 5.1V 0.5w diode. Do you think other values  for the diode could work better?

[ghiekorg]

Here is the test

thank you guys

[Mark Hammer]

Now you know why I generally only include a 2 or 3-position toggle for Feedback and Sweep width.  The Offset control has lots of settings in between min and max that are useful and audibly different.  The sweep width and feedback less so.

I also usually include a dry-lift for vibrato, and wet level for less intense phasing.  Personally, I don't find much of use for myself with wet levels between 50% and 100%.

But the main thing is that you got it working.  Congratulations! 

[ghiekorg]

Thanks Mark
Yeah you're right, but i wanted to try. Now i know that too
Thank you for your help