A Phase 90 Thread... *sigh*--I mean, how exciting!!

[BadIdeas]

I was hoping to install a switch on my block P90 to make it function as an EVH, but I think I now know enough that I can't, especially w/ only a DPDT switch. Still, I was hoping to make a learning experience out of this project and want to know more about just what the differences are and do, exactly. Some things I don't understand at all, others I am just looking for verification on, so here is what I know:

1) Remove R28
I've read this one often enough. It appears to be feeding the output of all the phase stages in to an earlier phase stage, emphasizing the effect. Removing the resistor would mellow it out, therefore. I had thought previously that it was allowing some signal to bypass some phase stages, decreasing the effect, but I don't thing that's right. Do I understand this correctly?

2) Remove R32 & C11; jumper R31
I believe this turns the gain stage at the beginning in to a unity gain stage. I don't know what the gain is increased for.
If the gain is 1+(output to -in Z/ -in to ground Z) then the gain would be 1+(0/infinity), but if I leave R32 and C11 in the circuit, then it is 1+(0/2.2kOhms and some reactance), so basically the same. That said, I can easily swap the first two points with a DPDT toggle.
Again, am I right?

3) Remove C12
Not a clue

4) Swap C4 & C2 w/ .047u
Changing C2 would seem to block more bass, which does not seem desirable. Haven't a clue about C4

5) Swap R20 for a 1Meg or use a 1Meg pot as depth control
I don't know why that works, but I would love to find out.

This community is so awesome and helpful, Don't let me down! 

[Mark Hammer]

I have no idea what the little magic switch on the current issue P90s does, but suffice to say that those switches are generally DPDT at best, and the differences between the 1st issue "script" model and subsequent ones entails more differences than can be accomplished via a little pushbutton switch.

1) The script uses 22k resistors in parallel with the JFET; block uses 24k.
2) The script uses a 22k feedback resistor; block uses a 24k.  22k gives a slightly more intense sound.
3) Script uses a sextet of 741 op-amps; block uses a trio of TL062 duals.
4) Script uses a unity gain input stage; block provides a gain of a little over 5x in the input stage.
5) Script uses a .01uf output cap in series with the signal, giving a rolloff around 106hz; block uses .1uf, dropping the rolloff down to 10hz (inaudible unless you play bass)
6) Block uses a 680pf cap across the base and collector of the output transistor; script has no cap.
7) Script uses a 200k trimpot feeding a 1M fixed resistor to apply the bias voltage to the JFETs; block uses a more complex arrangement, involving two fixed resistors along with the trimpot to let the trimpot be more "dialable", and feeds a 750k resistor.

If you made a Tonepad P90, many of the Script features are already built in.

What the impact of #1 is would depend on the JFETs, the bias setting and the resistor value from the bias trimpot wiper; ie., nothing you can't fake.  #2 is a VERY modest change in resonance, and quite frankly, you're better off with a variable resonance anyway.  Stick an 18k resistor and a 25k pot in series instead of the R28.
Nothing you can do about #3,but who cares; the TL062 is a better chip anyway.
#4 is likely to lead to the JFETs distorting a little more easily.  If you want to restore the unity-gain aspect, simply find the 2k2 resistor or the cap it connects to, and lift one lead to make it open, converting the input to unity gain.
#5: like I say, you won't hear it unless you play bass.
#6: I don't know what the cap does, but it's easy to spot, and you can simply unsolder it from the block model.
#7: Provides a fixed range and positon of sweep in the spectrum.  It's a compromise setting, really.  I doubt whether the one type of arrangement is "better" than the other.

[BadIdeas]

Thanks, Mark.

I was thinking of a DPDT submini toggle, or maybe even squeezing in a 3PDT mini toggle. The 24k Resistor you mentioned in #2 would go to one pole with one throw connected to it's other connection.

On #4, did you mean unity gain would distort in the JFETs or the 5x gain? I was thinking of connecting the 10K resistor's spot to a pole w/ a jumper with the resistor itself leading from one throw, and a jumper from the other. With the switch in the jumper's position, the opamp would be set to unity gain.

If anyone has information on the transistor cap, that would be great. Maybe it filters out some of the highs?
Actually, I'm even more curious about C4. It's the .1uF cap in series with the JFET/resistor combinations and appears to lead in to the bias voltage, which seems very odd. Would the audio signal have any effect on the phasing?

Anyway, if I went with a 3PDT, I guess I could either use the third pole for C2 or C4. I'm guessing C2 (the transistor cap) is the more significant of the two. Thanks for your help. I should be getting my parts next week. It's pretty silly I can get a Block P90 and convert it to near EVH specs for less than the Script unit.
DIY RULES!!

[moose23]

Cheers for that info Mark, just picked up a block Phase 90 cheap and was in need of an explanation for the popular mods, rather than just a list of them.

[Mark Hammer]

The JFETs distort in response to excessive signal level, so the 5x gain in the input stage increases the likelihood that hotter pickups will bring out the ugliness in the JFETs.  At the same time, some folks like the bit of boost fro when they kick in the pedal.  A compromise would be to change the 2k2 resistor for another 10k unit, giving a gain of 2x.

There are so many different possible regen settings.  I don't know why one would want to restrict oneself to the two R28 values in the script and block issues.  If space is an issue, though, consider use of a 3-position toggle.  Replace R28 with a 33k resistor.  The toggle would switch in either a 68k or 36k resistor in parallel, to get you values of either 33k (minimal regen), 22k (script), or 17k (stronger than script).

A second mini-toggle could be used to select between two different values of the gain-setting resistor in op-amp stage 1, or between gain and unity-gain (block/script).

That doesn't "undo" all the block changes, but addresses the most noticeable ones and gives more options than a single do-all switch.

[kymar]

7) Script uses a 200k trimpot feeding a 1M fixed resistor to apply the bias voltage to the JFETs; block uses a more complex arrangement, involving two fixed resistors along with the trimpot to let the trimpot be more "dialable", and feeds a 750k resistor.

I want to install depth control pot instead of 1M resistor.
Is it necessary to swap bias 20k trimpot to 250k one? or it does not affect on depth?

[Mark Hammer]

R.G. has generously worked all that out for you already.

[kymar]

this is more suitable for me
http://www.aronnelson.com/gallery/main.php/v/Schematics-etc/MXR_Phase_90_to_Script.gif.html

[Mark Hammer]

That will work too.  Replace the 750k resistor with a 390k and 500k pot.  That ought to provide some decent variation in phasing range.

[kymar]

but what about trimpot (R27 mod)?
has it effect on depth width?

[Mark Hammer]

The JFETs need to be biased.  Whatever bias voltage you set it to should be stable.  The two different biasing subcircuits (script and block type) both divide down a stable 5.1v source (kept rock steady by the zener).  The 250k trimpot is a simpler arrangement than the 25k+resistors, but it risks overshooting or undershooting the target by having all possible valid and invalid bias voltages (0 - 5.1) crammed into the rotation of the trimpot.  The 25k+resistors arangement makes it easier to nail down the appropriate bias voltage by having the trimpot only cover the most relevant part of the entire range of possible voltages.

The bias trimpot can actually be used to set the overall range or starting point of the sweep.  I have a Rocktek phaser, which is essentially the same circuit as a P90, with different JFETs, and I moved the bias trimpot to a chassis-mount pot to do just that.  Just like the script P90, though, the pot covers a much wider range of values than is useful, and extreme settings to one side or the other of the pot get you no phasing whatsoever.

[kymar]

Thank you Mark for the helpful information!

[BadIdeas]

A second mini-toggle could be used to select between two different values of the gain-setting resistor in op-amp stage 1, or between gain and unity-gain (block/script).

That doesn't "undo" all the block changes, but addresses the most noticeable ones and gives more options than a single do-all switch.

Yeah, the epiphitree hit me this morning and said I should use two submini toggles. Thanks.
I'm working from the Aron Nelson Schem, by the way.

That will work too.  Replace the 750k resistor with a 390k and 500k pot.  That ought to provide some decent variation in phasing range.

I don't understand LFOs that well; that's one of the things I was hoping to learn from this. They make my head hurt. How does that resistor (R20) affect the depth?

[Mark Hammer]

I don't understand LFOs that well; that's one of the things I was hoping to learn from this. They make my head hurt. How does that resistor (R20) affect the depth?

R20 sets how much current is applied to the gates, and is not part of the LFO.  The LFO is summed with that current coming through R20.

[BadIdeas]

It looks like R20 not only supplies current to the JFETS, but also the -in of the LFO. Does the function of the LFO have to do with the charging and  discharging of C7? Wait a second, R20 is connected to BOTH inputs... This is where it gets really sticky for me... 

[kymar]

one more
What is the purpose of r22 resistor? it seems to have different values not only in script and block models but in each version of device. I have 4M7.
it is part of LFO (out, i think)? how it work and how it affects the sound? what happens if change the value of it?