EH Polyphase debugging... can anyone help?

[Morocotopo]

 Hi people, I built the EH Polyphase with Paul Nelson´s project files (Thanks, thanks for your sharing your work Paul!), and it´s debugging time, with your kind and gracious help, of course.

1.Name of the project: Electro Harmonix Polyphase

2.Link: http://topopiccione.atspace.com/PJ17EHPolyphase.html
          http://topopiccione.atspace.com/pjimages/EHPolyphase.pdf

3.Home etched PCB, as per Paul Nelson´s project PDF. Tested traces for continuity after etching and before populating.

4.Parts Substitutions/mods:
Resistors:
All 36K: 33K
    43K: 39K
    1.3K: 1.2K
    51K: 47K
All 75K: 82K
All 240K: 220K
All 620K: 680K
    24K: 22K
    160K: 150K
    22R: 10+10R(series)(only the one between the 3140 and 4013)

Caps:
          0.0082uF:0.01uF
          0.0056uF: 0.0047uF
All 15uF: 10uF

Diodes:
1N4001: 1N4007

Optos: CLM 8200/2 (not a substitute, but... just to add more info)

All Transistors oriented according to pinout, including the lowest 2N5088 at the right. By the way, the 2N5087 I used has a little blue paint strip on the round part of the case, and one of the 2N5088´s has a sort of white one. Does that matter?

6.What it does and does not do:
The Phaser works ("phases" nicely, great deep sound) in LFO mode. The rate pot works OK, the feedback pot works but from about 5% to 95% of it´s rotation the background noise increases in a way that doesn´t seem related to rotation degree (the noise "spectrum" varies kind of randomly as I rotate the pot). When I touch the pot case, the noise increases. At both extremes the noise gets quieter.  At max feedback, background noise (unboxed, wires and alligator clips all over the place, close to my computer monitor, etc.).
In env mode, when I put the Env Sens pot at one end of the rotation (at max?) I get response to picking strenght (the phase "rises and falls", the timbre changes) with humbuckers, less so with single coils, but not at all a reasonable amount. At the other end of said pot, no change in sound at all. The Env Mod and Mod rate pots do nothing. I think that after the env controlled phase change, the (2nd?) LFO is supossed to "kick in", right? The sound seems "stuck" in one phase position. When I touch the pins 2, 6, 9, 13 of the LM324 with a metal screwdriver (making contact w/ my fingers) the phase height "rises", when I stop touching it, the phase "falls" to it´s previous position.
Tried replacing each IC with another one of the same kind (except 4013): no change.
Trimpot adjusted to 6.7V on B

Power:Power supply, filtered and regulated. LM317 based.  17.9V not connected to the circuit. Also tried it with 2 9V in series. Same problems. also tried it with a 7660 charge pump, but it didn´t phase at all with that (not enough current?).

Voltages:
-------------------------------
WITH THE SWITCH IN ENV POSITION
-------------------------------

Voltage at the circuit board end of the red battery lead = 17.8V
Voltage at the circuit board end of the black battery lead = 0.02V

Since the available schematic for this is very blurry and the parts are not numbered, I took the Layout image and put reference numbers on the semiconductors.



01 - 15Volt regulator (small, transistor case-type, not heatsink type)
in: 17.9V
out: 15.09V
gnd: 0.02V

02 - RC4558P (Texas Instruments)
1: 7.12V
2: 7.14V
3: 5.81V
4: 0.02V
5: 5.25V
6: 6.43V
7: 7.11V
8: 14.36V

03 - RC4558P (Texas Instruments)
1: 7.39V
2: 7.29V
3: 6.53V
4: 0.02V
5: 6.62V
6: 7.29V
7: 7.19V
8: 14.36V

04 - RC4558P (Texas Instruments)
1: 7.37V
2: 7.27V
3: 6.54V
4: 0.01V
5: 6.43V
6: 7.30V
7: 7.11V
8: 14.28V

05 - RC4558P (Texas Instruments)
1: 7.39V
2: 7.29V
3: 6.60V
4: 0.01V
5: 6.57V
6: 7.28V
7: 7.18V
8: 14.27V

06 - RC4558P (Texas Instruments)
1: 7.37V
2: 7.30V
3: 6.60V
4: 0.01V
5: 6.54V
6: 7.26V
7: 7.17V
8: 14.27V

07 - GD4013B (Goldstar)
1: 0.01V
2: 14.24V
3: 0.01V
4: 0.54V
5: 0.01V
6: 0.09V
7: 0.01V
8: 0.01V
9: 0.01V
10: 0.01V
11: 0.00V
12: 0.01V
13: 14.23V
14: 14.24V

08 - CA3140E (Intersil)
1: 0.07V
2: 8.00V
3: 7.79V
4: 0.01V
5: 0.07V
6: 8.00V
7: 14.24V
8: 9.37V

09 - LM324N (ON semiconductors)
1: 8.00V
2: 7.81V
3: 7.94V
4: 14.29V
5: 7.96V
6: 7.86V
7: 7.92V
8: 7.99V
9: 7.85V
10: 7.99V
11: 0.01V
12: 7.92V
13: 7.99V
14: 8.00V

10 - LM1458N (National?)
1: 2.07V
2: 7.95V
3: 7.29V
4: 0.01V
5: 7.95V
6: 1.57V
7: 11.61V
8: 12.37V

11 - 2N5088
C: 8.55V
B: 1.20V
E: 0.69V

12 - 2N3904
C: 8.55V
B: 0.69V
E: 0.01V

13 - 2N5088
C: 0.08V
B: 0.58V
E: 0.01V

14 - 2N5087
C: 9.55V
B: 9.01V
E: 9.57V

15 - 2N5088
C: 0.53V
B: 1.14V
E: 0.52V

Phew! That was a long list!

Knowledge gathered from other posts on the subject:
- Main LFO: 4013/3140
- Env ¿LFO?: 324/1458
- 820K resistor: "Envelope Modulator voltage is added to the EF voltage through an 820k mixing resistor"

Understanding the schem completely is way over my head...

I hope someone can give me some help, because it ALMOST works, and I want to get it working fully.

Thanks in advance

Morocotopo

[Rectangular]

Hey morocoptopo

I built one of these about a month ago and had pretty much all the same problems you've mentioned. I'm still waiting for my reverse log pots to arrive, so I suspect that might be one of the issues with my envelope section.  you used quite a few substitutions,  I'm no electronics expert, but I think the difference between a 620k and a 680k is a lot more important than swapping say,  a 1.3k with a 1.2k resistor.

as for your feedback noise issues,  from what I've gathered (along with other polybuilders), the type of 4558's used is important in determining how this unit feeds back. I assume the 4558's EH was using back in the 70s were very cheap (japanese or korean), and thus the sound scattered, into a noise timbre, when fed back. newer 4558s may be a lot more precise,  causing them to self-oscillate into a rather harsh overdriven sinewave. Ive tried high-noise/cheaper dual opamps, and they're slightly better than the 4558s I was using, but still not perfect.

the 7660 doesn't provide enough current to drive this unit,  it was the first method I tried in powering the polyphase. a 7660 can provide around 10mA max, where as the polyphase draws around 45mA,  so I'd stick with your present power solution.

if you make any progress with getting the envelope section fully working,  please let me know, as I still can't hear any LFO in mine, and it should be there.

-Rec

[Morocotopo]

 Rectangular, thanks for the info. So I guess it needs cheaper opamps then. Will try to find BA4558´s because someone mentioned those in a thread about this circuit.
As fot the envelope section, I´ll keep trying to debug it. I don´t use rev log pots, I use log ones, but I don´t think those should influence how the thing works, only how easy it is to control it...
In the meantime, if someone has any more ideas they will be welcome.

Thanks

Morocotopo

[RaceDriver205]

Yeah, I finished mine and it didn't work. Signal does not go through the first opamp stage (it appears). Haven't had enough interest in it to return to fixing it yet. Shame Nelson has cleared off.

[Morocotopo]

I´m guessing there are a few people with non (fully) working polyphases, for what I´ve read here and in other threads...

Since some resistor values asked for in the schem are non standard, at least in my country, I substituted with the closest one. Supposedly, a plus/minus 10 % sub is OK with analog circuits, right? Does anyone think problems could be related to that? Any of you guys used resistor subs like me?

Race, what do you mean, Nelson no longer is a member of the forum? Too bad...

Wouldn´t it be nice for some of the succesful builders to chime in if they have any ideas?

Thanks

Morocotopo

[RaceDriver205]

No, he's just not been by for like 1/2 a year.

[lowstar]

mine works perfectly, and i have chimed in before trying to help the other polyphase builders.
i´m going on a 3 week vacation and have to leave the house in 2 hours, i must do other things today, so i can´t take voltages of mine or the such, i´ll try to when i´m back.
i can say that i did not sub any resistor values, the only subs i made were that i used 10µ electro´s instead of the hard to find 15µ.
keep in mind that it is a big build, i had some short here, some thingie there, it took a long time to debug. i also had one transistor turned around, i mentioned the whole thingie in one of my earlier threads.
i´m the one with the ba4558´s also, the normal rc4558p gave me a lot of noise and weird feedback in some pot rotations. banzai carries them.
but keep in mind that mine worked with the normal rc4558p´s also, so for the builders that cannot get any phasing out of it, the other op-amps will not be the ticket IMO.

gotta go now !

cheers,
lowstar

[Morocotopo]

 Thanks lowstar. So you can get those value resistors? I couldn´t find them, at least in 5% ones. Maybe they exist in 1% metal film ones, I never used those. Will investigate.

Morocotopo

[Morocotopo]

Last bump/call for help.
Just to see if anyone has any idea.

Thanks for reading

Morocotopo

[analog kid]

 

Wouldn´t it be nice for some of the succesful builders to chime in if they have any ideas?

very late response I know but sorry I missed out on helping with you potential EH Polyphase clone builders ; ) I was the first one to build one of these and worked with Paul through the the beta testing  process when unfortanately we discovered it needed a complete redesign of the pcb to adress lfo noise ( extensive decoupling /filtering!, which he did a very nice job on the pcb changes) anyway this has been long ago but I've built 3 of these units ( wink! I have a rather nice stock of original NOS Clairex 8200/2's) as well as refurbed 2 original EH units.  Obvioulsy I've been through this design far more than I care to have during troubleshooting and Hopefuilly I could have/can help with issues you might have had.   SORRY but PM me if you may any longer have need for assistance
MAN ! Looking through the couple questions that popped up earlier this year I'm really P'd that I missed that since I was so absorbed w/ this crkt for quite awhile I'd have loved to have discussed it with someone else   

... what do you mean, Nelson no longer is a member of the forum? Too bad...

afaik Paul has never built this effect , or hadn't when I was building and corresponding w/ him. NOT that he wouldn't have been a huge help as he knows the crkt like back of his hand obvioulsy. he helped me many times with thoughts

 I also cloned the Echoflanger which was a freakin huge accomplishment imho and it's just an amazing device I tell ya. and just as with this unit , I was very alone in it and suprised that I could find no sign of anyone else ever cloning one    but I guess same goes as with the A/DA Flanger which when I built mine from M.Irwin's notes , the only folks I could find who'd ever built IT were Irwin himself , S.Giles and B.Bergman and none on a PCB (APPLAUSE)
so I guess I've made a habit out of that
analog kid

[Morocotopo]

Hi analog kid, thanks for responding. My build ended up like this: I replaced all 4558´s for BA4558´s and the background noise problem went away. As for the envelope controlled part of the FX, I just got fed up of trying to debug with my limited knowledge and disabled it: took away the pots and the IC´s (to lower current consumption), leaving me with a VERY, I repeat, VERY nice sounding phaser with rate and feedback pots. I added a volume control at the output because, in true bypass use, the FX has a quite higher volume than the straight signal. I also added a R at one of the rate pot  lugs (don´t remember wich one), to limit max speed, because max speed was unusably fast, not even perceptible, either that or the thing locked up.

Maybe you can post all the mods, tweaks, or whatever you did to make it work (or didn´t do, such as subs) for the benefit of someone else? Maybe even encourage me to build another, fully working one...

I can get 8200/2´s here for around U$S6 each, I could sell some to people at U$S, let´s say, 20 each... (he, he, he, he... just kidding)

Echoflanger? Sounds interesting... did I hear project file for the rest of us? 

A/DA flanger on perfboard?
(crowd chanting and hitting their knives and forks and dishes on the table) We want pictures, we want pictures!!!...

Morocotopo



Morocotopo

[Morocotopo]

Forgot to say thanks for your help offer analog kid.

And, also:

Next thing to do: a 18V charge pump to power this thing from a 9V power supply! Should supply upwards of 50 mA, or it won´t phase. Hard one to do. I investigated the LT1054, but no dice, as a voltage doubler the chip supplies (I think) around 50 mA. Maybe two of those in parallell? Hmmm....

Morocotopo

[StephenGiles]

You asked for it!

[Morocotopo]

(shaking head in disgust)
See kids, where an addiction will take you? These people are INSANE!! 

Morocotopo

[bleubleu]

Hi!

I've just completed building the Polyphase, the unit works but I have some problems. I used all exact resistor, capacitor and ICs, except the CLM8000s, I used VTL5C3s.

1) I cannot get the 6.7V at point B. Even if I play with the trimpot, I cannot get higher that 5.5V there.

2) My unit works very well in Envelope mode, but not so well in LFO mode. When I switch back from Envelope mode, the  LFO "goes up" at the specified rate speed and never comes back down. Seems like its only doing half of its cycle. Probably related to problem #1.

3) I also used JRC4558 amp and yes, they are noisy... Playing with the feedback pot sometimes picks up radio stations and strange things happen. I have not used IC sockets, shame on me. grrrr.

Any help ?

I will never try such big DIY projects, they are simply not worth it. I will focus on smaller/simpler projects, biggest bang for the buck.

In the mean time, here is a picture of the monster....


Bigger picture here : http://www-etud.iro.umontreal.ca/~gauthmat/stuff/poly.JPG

Mat

[bleubleu]

Well I seem to have fixed my problem.

I simply was feeding it with almost-dead battery... Notice on the picture, that I had temporarily removed the voltage regulator. Now everything seems to work. The feedback pot still give some strange sounds (picks up AM radio stations, hiss, etc.) when not properly grounded. The final test will be when I put it in a metal box with everything grounded and shielded from radio interference. I also had to add a volume pot because the output was way too loud for my amp, causing clipping.

In the mean time, I recorded some sound samples using my MIM strat and a electric piano. The feedback and rate was set relatively high so the phaser sound is very audible. You can get a more subtle sound if you like. There is some clipping in the MP3s, but it is simply because I recorded too loud, the phase is not the problem.

LFO files use the LFO mode (The phaser oscillate at the specified rate)
ENV files use the envelope mode (The phaser is controlled by the input volume).

• http://www-etud.iro.umontreal.ca/~gauthmat/stuff/guitar_lfo.mp3
• http://www-etud.iro.umontreal.ca/~gauthmat/stuff/guitar_env.mp3
• http://www-etud.iro.umontreal.ca/~gauthmat/stuff/epiano_lfo.mp3
• http://www-etud.iro.umontreal.ca/~gauthmat/stuff/epiano_env.mp3

Mat