Some ideas for the P90

[Kipper4]

The JC Mallet modded version mixes the dry and wet signal via the 22k pot a IMO useful feature that get that vibe or phaser feel to the pedal.
You could get this in a 1590b with 3 pots side by side (no dust covers) or triangulated.
Maybe even have room for a battery too mate.
Breadboard the mix pot and see if you think it's worth it.
What op amps are you using?

[Ben Lyman]

Thanks Rich, I will look into that mix pot, it does seem like it might be useful.
Maybe I wouldn't need a depth control if I had the mix pot to tame it a little bit.
I'm using a couple TL072's and I have a variety of FETs but none of those weird old kinds in the schematics.

I'm gonna post these links again here since we've gone on to page two:
Mark's schematic (THANKS!) from Ampage: http://www.ampage.org/schematics/mxrphase45.pdf
the GGG schematic: http://www.generalguitargadgets.com/pdf/ggg_p45_sc.pdf
Here is JCM's: http://www.lynx.bc.ca/~jc/phase45modded.gif

I'm thinking of replacing the 1M (depth resistor?) with something bigger or a pot?
Or is the mix knob a better way to go?

Other changes of mine include:
2x TL072
2x 2N5457 or 2N5458
Series 1n5817 polarity protection
Series 100r power filter and bypass caps 100uF & 100nF

[Ben Lyman]

Well, I tried the 22k mix pot, maybe I did something wrong but I don't see how, it's such a simple thing and written so clearly in JCM's schematic. Results were this:
Far CCW turn= no effect at all, just loud clear guitar
Middle= nice effect with volume drop
Far CW turn= volume returns with very subtle vibrato effect

best results seem to be from 9:00 to 3:00 ability to do some mixing but I am not 100% convinced it's worth it... nor am I 100% convinced I wired it right 

I think I will start working on a layout for a 1-knob, 2-switch version:
Speed knob, plus switch for vibe/phaser caps, and switch for deep/shallow phase.

while I am working on that, hopefully if anyone has anything to say, please speak up 
Thanks!

Edit: I just had a thought. Look at JCM's 22k mix pot. it goes straight from the IC out (buffer?) to the output cap. No wonder I got dry signal, right?
Hmm... maybe a 10K pot with a 10K resistor? 

[duck_arse]

external bias pot?

[edit :] that kipper fella, didn't he add a switch wot changed the wet phase into the out mixer? that might be a thing here.

[Ben Lyman]

Good idea Duck. JCM has the external bias pot so I started with that. I don't know if it's my choice of FETs & ICs or just the nature of the P45 but my 250K bias pot has a very narrow working range, about 10k-50k. As a result JCM's 100K bias pot (2x 47K straddling) was pointless, I pulled it out and got a better working result with a 50K pot in between a 120K & 81k, even better still with a 25k between some resistors I can't remember. Pulled it out, which I so often do when I listen to the results and weigh it against drilling a hole in an aluminum box. lol 

Here is another test I ran which I liked, just like in my previous experiment with the P90 and a variable feedback resistor. Look how the dry path goes around through a 10K in the P45 schematic. JCM uses the mix pot, some others might break the link with a switch to get straight vibrato, so I put a 100K pot in series with the 10K wired with just the wiper and one other lug for variable resistance. This seems the best way to my ears, pot fully turned offers no resistance allowing for full phasing mix, turn the other way resists the dry signal and emphasizes only vibrato. Basically a mix control without the volume drop in the center.

Again, I am still not even sure it's worth drilling a hole to be able to do this when I get the same amount of satisfaction by stomping my foot on something and playing guitar with or without switches and knobs! 

On to one more stupid question: Is there a way to get more volume out of this?
I don't really want to add a booster transistor but it seems like the pedal is just barely equal to the bypassed volume.
It would be nice to be able to adjust it up or down a tiny bit.
Why no "output mixer" like in the P90? Just cost cutting or it doesn't need it?

[Ben Lyman]

This is what I'm starting with, it will be a little easier for me to suss out, providing I've drafted it correctly.
Thinking of keeping it very simple, C6 and C7 will be replaced permanently with 10nF and 100nF respectively.
R16 (10k) will have a 100K (or higher) pot in series wired as a variable to cut out the dry signal.

Any bold faced errors?

[Ben Lyman]

Okay, you guys convinced me the bias control is pretty cool 
So to keep it simple, I am only having two knobs, speed and bias.
I couldn't see the point of having a 100K bias pot that only has an effect right in the middle, so using my DMM I was able to find that the bias pot was only working in a range of about 44K.
Here's the updated schematic including all mods, they aren't a lot.
Bias is a 50K pot with 100k fixed on each side.
note the 27k R10 for added volume, I think?
R25 is 3M3 for more deeper and R29 is 5k1 for more ultra super mega fast

[duck_arse]

Okay, you guys convinced me the bias control is pretty cool 
.....
I couldn't see the point of having a 100K bias pot that only has an effect right in the middle, so using my DMM I was able to find that the bias pot was only working in a range of about 44K.
.....
Bias is a 50K pot with 100k fixed on each side.

oh, we thought you knew, this is the way it's done. you can measure the voltage range of usable bias, then pick a pot value, like 50k, 10k, 5k, whatever is close handy, and do the resistor string and Ohm's law to work the top and bottom resistors, if you are the trainspotting type, but your way works.

[Kipper4]

And your Q1 Q2 are? please.

[Ben Lyman]

oh, we thought you knew...

You presume a lot there, Duck 
I tried a few resistors on each side of my 50K pot but settled on 100K each, because the pot bias functions about dead center anyway.
The end result is that the pot has a small spot at each side where the modulation seems to go away completely but I can live with that.

Rich, the Q1/Q2 are 2n5457 today. I didn't notice any difference from putting in 2n5458, I think either kind should work.
I also have not bothered with any matching, though I have read plenty about how important matching them is... I just don't want to 

[Kipper4]

Thanks Ben When I get the chance I'll bread it.

[Mark Hammer]

R25 sets the width of the sweep.  You'll find some issues that use 3M9, for a narrower sweep, ideally suited to faster rates, and some use 3M3, for a wider sweep.  Keep i mind that MXR aimed for a one-size-fits-all approach with that pedal/design.  One fixed sweep width, one fixed wet-dry mix, one fixed feedback setting.  When people express a preference for this issue or that, they selecting the P90 issue where the fixed feedback and width settings more closely match what they usually aim for.

By "wider", I mean that each cycle sweeps higher up before coming back down again.  Typically, on any two-knob modulation effect, whether flanger, chorus, or phaser, the faster you set it, the narrower you set the "depth" (more accurately, sweep width) control.  The usable range would seem to be between around 2M5-4M5, so a 2M2 fixed resistor with a 2M log pot would provide a usable range of sweep widths.  Alternatively, a 4M7 default value, with two other fixed resistors strapped on in parallel, via a 3-position toggle, would provide 3 different sweep widths, tailored to different LFO rates.  For a higher-feedback slow sweep, you'd probably want somewhere in the 3M-3M3 range.  Of course, with a 2-stager like you show, there IS no feedback, but still one would generally like a less boing-ey sound from faster sweeps so a narrower sweep width helps.

[Ben Lyman]

Thanks again Mark! It's amazing how much I've learned since starting this thread.
And yes, I have changed the topic to that of a P45 with some of the vibe mods from JCM.
I think for now I am going to stick with what I have already started assembling but in the future I will mess around some more with the width mods.
This one might get a spst for one deep mode but for now just the bias pot and speed.
It might take a few more days but I will make a quick sample vid to show how good it sounds.
I'm very impressed at the sound this thing gets compared to the headache of building the P90.

[Kipper4]

After a few hiccoughs I managed to get it working on the breadboard.
No 4v7zeners in stock, so used a 5v1zener seem the mxr used the same.
I subbed in some 2n5458 it sounded and worked better.

All I can say Is you must have got really lucky with the fets Ben. (Jammy dodger).
I could hardly get the 5457 to work and then they where thumping.
I had to make R32=122k to get them to even go on.

Something strange happens though. When I play hard It stops phasing.
And the speed range is a bit too much IMO.
Too fast, Too slow. I guess its a suits your needs thing. YMMV.

I'll mess with it some more later
Cheers

Edit.
"Something strange happens though. When I play hard It stops phasing."
My fault. turns out its my breadboard connections shorting,

I tried the Hammer mod.
swapped R25 for 2M2 and series (varistor) B1M pot. That works ok actually nice for the faster speeds to be able to make it smoother.
Is it worth an external pot. I dunno maybe a descreet one.
It has a dead spot at one end too.


I tried it with a 3M3 and A1M pot (not sure if the pots dodgy) but it wouldnt work.

[Kipper4]

"All I can say Is you must have got really lucky with the fets Ben. (Jammy dodger).
I could hardly get the 5457 to work and then they where thumping."


wait wait.
I take that back now too.
Since sorting out the shorting they work ok.
Same Vb as above 100k 120k
I still think the the 5458 sound better somehow though.

[Mark Hammer]

I tried the Hammer mod.
swapped R25 for 2M2 and series (varistor) B1M pot. That works ok actually nice for the faster speeds to be able to make it smoother.
Is it worth an external pot. I dunno maybe a descreet one.
It has a dead spot at one end too.

I tried it with a 3M3 and A1M pot (not sure if the pots dodgy) but it wouldnt work.

Think of it like carrying a 2x4 vertically.  If it's long enough, you need to carry it low for it to fit under the ceiling.  If it's short, you have the option to carry it high or low.
By analogy, the sweep width option really needs to be complemented by the offset option.  And the offset is provided by two means, one of them being the trimpot, and the other being the 1M resistance that feeds the FET gates.  Varying the value of that resistance shifts the sweep range upwards or downwards.  In tandem with the sweep width, you can arrange for the sweep to stay low and gurgly (by having a narrower sweep width), high and swirly (also a narrow sweep width but starting higher up), or wide and majestic (start low, sweep high up).  Having it set for too wide a sweep, given the offset and bias, will bump the sweep against the ceiling.

[Kipper4]

Gottya. I'll experiment some more thanks Mark.

[Ben Lyman]

Overall, whadya think Rich? I like it a lot. Did you use the JCM caps for C6 & C7? I would make a vid but my BB components are already being transferred to the perf.
I did put sockets in for the FETs so later I can try the 2n5458s again.
Here's what I'm working with, not verified yet but if it all works out I will have the two pots mounted on the board and it should fit in a 125B, maybe even a 1590B.

[Kipper4]

Overall It's great mate.
I really enjoyed not having to match fets. I guess one could carry it on with a further dual op amp and make it a p90 style, even p180.
I'm gonna try the jc Mallet mix pot too while it's on the breadboard.
Good work mate.
I still want to investigate the bias pot further as It seems on my breadboard right now theres still only a very narrow part of the rotation that gets results.
Thanks

[Kipper4]

At some setting there appears to be a bit of fet distortion. Bear in mind it could just be my breadboard.
I swapped to the 2x 100k Vb resistors you used and the full rotation of the 50k bias pot works ok now (still using 2N5458s)
The bias pot makes some nice tonal changes especially when you have the Hammers Mod pot too. Much easier to dial in the sweet spot.

I did make a few changes to the lfo
4u7 charging cap (C12)
slow enough to make a cup of Tea between cycles at the slowest speed.

Make R29 (your 5k1) to a 22k still plenty of speed there for me. 15k would likely work well too. 10k was a bit too fast for my taste.

YMMV