This is a library of perfboard and single-sided PCB effect layouts for guitar and bass. I'm not an electrical engineer by any stretch of the imagination, just a DIY'er who likes drawing layouts. It is meant for the hobbyist (so commercial use of any of these layout is not allowed without permission) and as a way to give back to the online DIY community.

Thursday, April 2, 2015

MXR Script Phase 90

The Script version of the Phase 90 is probably the holy grail of phase shifters and was the first effect released by MXR in 1973. It's probably most famous for helping create Eddie Van Halen's brown sound on the first 2 Van Halen albums. It was designed before dual op amps were really affordable, so it used 6 op amps, in addition to 4 JFETs and a PNP silicon transistor. I've wanted to do a layout for the Phase 90 for a while, but all those op amps intimidated me, so I held off till a request for it came in. I used the original MXR board as a starting point. I ended up with 2 jumpers, but original units had a 150k resistor spanning over half the board (going from the base of Q5 to pin 6 of IC5).


One of the weird things about this layout is how the JFETs are oriented. The drain and source pads are further apart than the source and gate pads because the ground fill has to go between those gaps, and this quirk is evident on the original board from MXR. I would imagine 2N5457s would work if you can't get ahold of the original 2N5952s, and a 2N3906 can be swapped for the 2N4125 if necessary. Also, if you're not a fan 741 chips, a TL061 or 071 should work fine.

86 comments:

  1. Hi, how is the LED wired up? Anode to PCB and Cathode to 3PDT?

    ReplyDelete
  2. Is speed 3 meant to be connected to the pad in the ground circuit above the 150k resistor at the top ofvthe pcb?

    ReplyDelete
    Replies
    1. No, Speed 3 is left unconnected, or you can jumper it to Speed 2. The Speed pot just acts as a variable resistor.

      Delete
  3. Just to avoid mistakes... which kind of condenser shall I use for these projects ? Electrolytics are clearly indicated, my doubt is for the green and yellow ones.
    Cheers
    Alessandro

    ReplyDelete
    Replies
    1. You mean capacitors? Green ones represent film and yellow ones represent ceramics. Use whatever you prefer.

      Delete
    2. Thanks, yes sorry, bad translation from Italian.. I mean capacitors!

      So, for the green ones, i can use these ?
      http://i.ebayimg.com/00/s/MTYwMFgxNjAw/z/vnoAAOxyXzxTIG~6/$_35.JPG

      Delete
  4. I didnt do it right i just get clean signal going to amp no swoosh.

    ReplyDelete
  5. That transistors I can replace the 2n5952 ?

    ReplyDelete
  6. Is the schematic available I'm trying to trace the audio path for it I know it's exactly the same more or less with the one using quad and dual op amps but I'd like to accurately plot it out I'm currently audio probing my non working build thank-you Billy

    ReplyDelete
    Replies
    1. I believe I used this schematic when I laid it out.

      http://www.generalguitargadgets.com/pdf/ggg_p90_sc.pdf

      Delete
  7. hello i have made this layout and i found a mistake
    the source of the transistors is not connectec to the 5v1 zener supply as shown in the schematic
    the phaser was weak before that and now with a little jumper it became the real thing
    tnx for the layout but pls correct the mistake
    cheers from greece

    ReplyDelete
    Replies
    1. Yep, I've been looking at the schematic. You're absolutely correct.

      Delete
    2. so which is the wrong part, did it and there is a volume drop, so how do i wired the diode?

      Delete
  8. Is this verified..? I also see the difference betwen layout & schematics. Zener 5v1 is not in schematics position..?

    ReplyDelete
  9. This comment has been removed by the author.

    ReplyDelete
  10. Do you know some layout for phase 90 with tl074?

    ReplyDelete

  11. Hi, in my city the capacitor of 15 uf is not commercial there are only 10 uf and 22 uf,
    which one should it fit better?

    ReplyDelete
    Replies
    1. Try both. Or put 2, 33uF caps in series.

      Delete
    2. I did it with the 10uf capacitor and there is massive volume drop
      I also heard that there was a problem with the layout and that the 5.1v diode is not in the schematic or is incorrectly accommodated is it true?

      Delete
    3. The 5.1v diode issue was corrected. If you've got a volume drop, increase that cap value.

      Delete
  12. This comment has been removed by the author.

    ReplyDelete
    Replies

    1. Hi, I see above that this design is not verified, someone has built it and can verify its function ????
      Greetings and thanks in advance

      Delete
  13. Hi, sorry for the previous post, I see above that this design is not verified, someone has built it and can verify its function ????
    Greetings and thanks in advance

    ReplyDelete
  14. This comment has been removed by the author.

    ReplyDelete
  15. There has got to be something wrong with my board. o-scope shows that i am phasing on the top two opamps and then 3 and 4 are dead. I will dig into it and report back.

    ReplyDelete
    Replies
    1. Hooked up the function generator and o-scope, seems the top two opamps are phasing and the bottom two are not, so i guess i have more debugging ahead of me. Will report back.

      Delete
  16. I found a tiny fleck of solder between a trace and the flood ground (I really need to invest in an inspection microscope) and a possibly smoked 741.

    The v2 layout is verified.

    AT the moment i'm using an opa604 in a couple positions but no way should that matter. i have tl061 chips on the way.

    Might have to work up a layout for tl064 / lm348 + tl072 for a smaller build. It's neat to be faithful to the original, but I feel like a 125b is bigger than this effect really needs to be.

    ReplyDelete
    Replies
    1. Glad you got it sorted out. If you shave down the sides a bit it shouldn't be hard to fit in a 1590B.

      Delete
    2. Maybe, but I already drilled holes in an orange 125b. *shrug*

      Anyway, this is incredibly frustrating, but I'm back to where I was earlier this week.

      After a successful test, I decided to see which of the two 741 chips was bad, since opa60x parts are pricy and gilded for this job.

      Both of the ones I'd replaced with opa604 were fine in the input buffer position.

      And then I discovered that the 3rd stage has heavily attenuated output again. And installed opa606 in all of those positions, and still have good signal output from 1 and 2 but not 3 or 4.

      But when I started, I had good, phased output. *shrug*

      Delete
  17. Would LM741CN be ok for this build?

    ReplyDelete
    Replies
    1. And are 50VDC capacitors and 0,4W resistors enough? Could BF256B replace 2N5952 and 2N3906 replace 2N4125? I'm really sorry, but I'm green in electronics. I have some manual abilities, but no real knowledge. Always made stuff acording to the instructions without even knowing what is supposed to do what. I'd really appreciate if you could tell me if these could be used in these builds:
      Xotic EP Boost
      Could I use a BC338-25-DIO instead of a 2N5088 and BF256B instead of 2N5457?
      Little Angel V4
      Any common replacements for 7805?
      Hollis Ultra Flanger
      What TL072 should I use? Is BZX85C3V9-TAP ok instead of 1N4730?
      Wampler Faux Spring Reverb
      Are J105 and J106 ok instead of J201? Any common replacements for 7805? What TL071 and TL072 should I use?
      I would be the happiest man alive if you've answered

      Delete
    2. The LM741CN should be fine, and 50v caps and 1/4watt resistors are fine. For most of your other questions, you'll need to do some googling and compare datasheets. If the specs aren't too far from the originals I'd guess you're probably fine.
      The 7805 can also be a 78L05. It's a 5 volt regulator, so there's not going to be a ton of alternatives.
      Use whatever TL071/72 you can source.

      Delete
  18. I just did an TL072 based one, with geda. I not verified it yet, but if anyone is interested, just send me a note.

    ReplyDelete
  19. This comment has been removed by the author.

    ReplyDelete
  20. there is an mistake on the PCB... at the fourth UA741, fourth pin must connected with ground with this change is verified.

    ReplyDelete
    Replies
    1. Hi! Has the layout shown on top been updated to correct the mistake you're talking about? Thank you.. :)

      Delete
    2. check the bottom right image on layout and you find the mistake

      Delete
  21. Hi, and thank you for your work! Unfortunately I couldn't find the Q5 PNP BJT transistor used for this project. I have a BC 177 PNP transistor and on papaer it looks like similar to Q5. I think that this transistor could do this job. Are you agree with me? Thank you!

    ReplyDelete
  22. Hi man, the picture 1 is empty or lost, please fixed :)

    ReplyDelete
  23. I found problem. Pin 4 of third 741 is not connected to ground on layout.
    On picture is ok but not on layout.
    Now is verified.
    Mine 90 work fine.

    ReplyDelete
    Replies
    1. Awesome. Thanks for pointing it out. It's fixed now.

      Delete
  24. https://zapodaj.net/223a7b86ce3e8.jpg.html

    ReplyDelete
  25. hello guys, can you tell me if this project has been verified by someone? I have followed the whole assembly phase of the components that requires the project, but it does not work, only emits a loud background noise. can you tell me something?

    ReplyDelete
  26. The version 3 (v3) is correct! VERIFIED!

    ReplyDelete
  27. I built this one, but sound is really, really low. Some tip about that?

    ReplyDelete
  28. I believe Speed 1 should be Speed 3, in case if you want low speed at 1 and high speed at 10.
    BTW I have built this with carefully paired 2n5457 JFETs, but I can hear the effect only at high speed. At low speed nothing is going on. The trim pot is not helping. Maybe I need another set of matched JFets.

    ReplyDelete
  29. Verified with some comments. It is a difficult build. Consider to choose an OTA phaser, they are easier.
    1. I used BC 327 for q1
    2. I used matched 2m5457 FET transistors
    3. The speed1 should be speed 3 if you want full speed at full clockwise
    4. Adding a feedback resistor (18-33K) between the fourth and the second stage will make it wetter, however it is no more script phase. Too low value resistor could cause distortion. The feedback resistor connects the jumper on the left side with the leg 2 of the second stage IC.
    5. I used TL071 for the frst five IC's and NE 5534 for the last which does the mixing.
    6. Maybe it is better to feed this pedal with 12 volts for more headroom at the mixing.

    ReplyDelete
  30. This comment has been removed by the author.

    ReplyDelete
  31. This comment has been removed by the author.

    ReplyDelete
  32. Thank you for great job! Works at first start: 071chips,russian Кп303а(2n5952),2n3906(2n4125).But effect really works with 50k pot, not 500k:)

    ReplyDelete
  33. As a suggestion to others who can't easily find 2N5952's, I found on another site that the Phase 90 works best with fets over 2v of Vgs. I built one of these with some matched 2N5457's (within 3mv) that were around 0.4v Vgs, and the effect sweep was completely wrong (went from 100hz to 10khz, rather than 100hz to 1khz ) and distorted quite badly. Swapped for some nearly matched J211's (within 30mv) that were close to 3v Vgs and it fixed both problems.

    ReplyDelete
    Replies
    1. Very good. 2v of Vgs would be 2v of Vgs off? And what does would be matching within 3mv or 30mV? Would be the maximum difference about each Fet? Thank You!!

      Delete
  34. Ola. Teria como explicar a ligacao da chave 3pdt

    ReplyDelete
  35. I created the GGG mxr phaser phase 90. When I set the trimpot stage, I set the sweetest point to the more prominent phase point, but I get a distorted sound scratching distortion sound how can I get rid of this situation I want to get a clean phaser

    ReplyDelete
  36. I used j113 and 2n5457 fets but the problem didn't go away

    ReplyDelete
  37. I'm trying to build an MXR Phase 90 but using J113 as the JFETs instead of the original 2N5952. All the J113 are matched, but the output is slightly distorted. It's not as noticeable when lightly strumming some chords, but picking strings with some amount of force will produce some ugly bass-y distortion. I've also tried different FET configurations according to these designs but none of them seemed to help. I noticed on the data sheets that the J113 I'm using has a forward gate current of 50mA while the original 2N5952 has only 10mA. Would it be worth it to get another set of JFETs more in line with the 2N5952, or does that spec not matter?

    ReplyDelete
  38. Do you have the original vector file for better print quality?

    ReplyDelete
  39. Can you please make a layout with quad or dual op amp..
    :)

    ReplyDelete
  40. Is there a way to add additional knos of depth or something else?

    ReplyDelete
  41. This comment has been removed by the author.

    ReplyDelete
  42. Its verified!! Mine with 2n5485 trans works fine! Tnx and more power! Regards fr Philippines. Jun.

    ReplyDelete
  43. Hi.
    I have problem with trimmer and potentiometer. I cant find 250k trimmer and C500k pot in here. Just 200k 500k and A and B type potentiomer. I have tried A500 and B100. I also tried 200k and 500k trimmer with 220 resistor solder to leg 1 and 3 but no phasing. The pedal bypass when turn off but when on, i cant here anything. Any suggestion? Btw, im using 2n4125 and 2n5457 trnsistor and Lm741 opamp

    ReplyDelete
    Replies
    1. You can series a 47k resistor on lug3 of the 200k trimmer.regarding on the phasing try check the bias voltage and that zener diode

      Delete
  44. For trim por u can use also 220k trim
    I think its written 224

    I have made few phasers with 220ktrim

    And for pot
    Use a B1M pot and solder 1megaohm resistor from lug 1 and 3 that sud simulate c500k pot
    U can also just keep B1M with out resistor

    I have done both..

    I do t think 2n4125 is Jfet. U need jfet for this 2n5457 is a jfet.by googling i found 2n4125 to be pnp bjt
    Ull need jfet for this
    And all 4 jfet sud be matched u can finemd how to match on youtube.


    Lm741 is good

    ReplyDelete
  45. Hi everyone can I use bf256b instead of 2n5952 and bc558b instead of 2n4125?

    ReplyDelete
  46. I can verify this also. I did find,as mentioned earlier, that speed 1 should be speed 3 in order to increase by turning clockwise. Thanks for this layout!!

    ReplyDelete
  47. I succeeded it worked using 2n5952 and 2n4125 ic ua741cp thanks again from thailand

    ReplyDelete
  48. This comment has been removed by the author.

    ReplyDelete
  49. If you use 2n5457's, just remember that the Drain/Source/Gate is set up opposite from the 2n5952's; they face the opposite direction. I pulled up schematics for the 2n5952's and had my 2n5457's facing the same way that the 5952's would face (opposite of the 2n4125 at the top), and spent hours trying to diagnose what the problem was. I fixed it, and now the pedal sounds awesome.

    ReplyDelete
  50. Verified with TL061's , 3 x 2SK30A Y, 1 x 2N5457. 2SA1015 .Used a 220K trimmer as tht was all I had. Nice sounding phaser,thank you!

    ReplyDelete
  51. The Phase 90 is a one-knob phaser. As such it was designed to provide a number of compromises. The width of the sweep was a compromise for something that worked reasonable well for fastest and slowest speeds (one usually wants a narrower sweep for faster speeds). The resonance/feedback was also a compromise setting, as was the range of where the sweep occurred.
    So here are some options for achieving a broader range of sounds out of the basic design, and making more like what you get with a 3 or 4-knob phaser.
    1) There is a 1M resistor coming off the trimpot, feeding the gates of all the FETs. Replace that with a 510k fixed resistor in series with a 500 pot, using as a variable resistance. This will provide variable offset to move where the phasing occurs in the spectrum, from lower and gurgly to higher and swirly.
    2) Right beside that 1M resistor is a 3M9 resistor connecting the 15uf cap to the same "blue" trace where the 1M resistor goes. That resistor sets the amount of LFO current fed to the FETs, and is summed with what comes off the trimpot. In other words, what the FET gates see is a combination of the default bias setting from the trimmer, PLUS whatever the LFO adds on top of that. Increasing the value of the 3M9 resistance will shrink the width of the sweep, and reducing it will make the sweep wider/farther. which is ideal for very slow sweeps. I would not go much higher than 4M3 for narrow bubbly tones, and much lower than 2M7 for ultra-wide slow sweeps. Worth noting that some issues of the P90 used 3M3, which is a little better for slower sweeps but a bit too jarring for the fast bubbly stuff. Note that pushing TOO much current at the JFET gates can cause some irritating glitches at one end of the sweep. So don't use a lower resistance coming off the LFO and trimpot at the same time. It won't damage anything, but won't sound pretty.
    Finally, note that the need for "matching" of JFETs is greatest when one aims for wider sweep. If the sweep is narrow, then the odds are good all JFETs will be similarly altered by the LFO. If the sweep is wide, then the risk increases that one or more of the JFETs will not change its drain/source resistance as you approach the peak or trough of the sweep. So if you don't know if your JFETs are reasonably matched, it may be best to avoid ultra-wide sweeps.
    3) The first issue omits any feedback/resonance, it would seem. If you wish to add some, then you would connect a resistance between pin 6 of that last op-amp in the lower left of the drawing, to pin 2 of the *second* op-amp from the top in that column of chips to the left. In other words, it goes from the output of stage 4 to the input of stage 2. Over various issues, you will find values of 22k or 24k. You can probably go as low as 20k or even 18k, but I would caution against going any further. Increasing the resistance value will reduce the amount of feedback. So, an 18-20k fixed resistor, in series with a 20k variable resistance/pot would yield a reasonable range of resonance/feedback settings.

    ReplyDelete
  52. Hi...is there a sub for the 15uf besides tants? Damn its pretty hard to find this kind of values

    ReplyDelete
    Replies
    1. you can add 2 electrolytic cap in series

      10uf + 4.7uf

      Delete
    2. Caps in parallel not in series sir.

      Delete