Polyphase - Troubleshooting

[Rectangular]

hey guys, I've been putting together Nelson's polyphase clone project for the past couple weeks. finally finished wiring tonight, and when I got to fire it up I came across the following problems:

1. the lfo section doesn't seem to be sweeping at all

2. the feedback pot doesn't have any audible effect on the sound

however, this thing does indeed phase, the envelope section works (but with no modulation). I went through the audio(phaser) portion of the board with an Audio Probe and haven't run into any noticeable problems, however I'm not so sure how to debug the control section, the audioprobe is useless there...    from what I understand the control section (lfo+envelope) is completely separated from the phaser section, by means of the optocoupler, so it shouldn't matter that I'm using homebrew optocouplers, I can clearly see the LEDs responding to the envelope, and I can make the thing phase by holding my hand over the LDR's like a photo-theremin

here's the schematic:

here's the project file:  http://topopiccione.atspace.com/pjimages/EHPolyphase.pdf

Dave B, Nelson, lowstar, if any of you guys can lend some of your experience on this project I'd really appreciate it

thanks for your time
-rec

[Rectangular]

ok, with a bit more debugging I finally got the feedback knob to give some kind of response. unfortunately, that consists of it squealing around 90 degrees into  the rotation, then becoming full-on static/noise past that

[Rectangular]

another update:

apparently the LFO is working in some form,  I found that if I dial the Rate knob to around 39K (very hard on a 1M pot !) I get some very limited lfo rate control. however, the moment I try and exceed or decline from that 39k "sweet spot", the LEDs (in the optocouplers) just stay on.  I can't get back to the sweet spot at this point,  I have to 'reset' the lfo stage, either by using the lfo/env switch or unplugging it/plugging it back in.

very strange behaviour, I must say

-rec

[lowstar]

hi rectangular,
i can´t give you any dmm readings at the moment cause i have to wait till i´m in my lab again (no electronix at home), but as a shot in the dark, check the orientation of the 5088 on the lower right hand corner of the pcb...i had it soldered in backwards first cause i looked at the symbol for flat side, but you have to do that one by the cbe markings, the other ones correspond also by the symbol. and that one is in the lfo section, afaikr.

cheers,
lowstar

[Rectangular]

hey lowstar, thanks for the look

as a rule, I always socket all of my transistors, so I've been flipping that whole set back and forth to see if there are any behaviour changes. I think they're all oriented correctly. I want to say there's something up with the 4013 or ca3140,  I dont really know those chips, I'm not familiar with this design, which is making it really hard to debug.

another thing I've noticed is that when I was trying to set the trim pot for 6.7 volts, that the voltage kept slowly rising on its own, into the 7-8v area ! some capacitor somewhere is up to something ? another mystery.

let me know if you have any other ideas 

[Rectangular]

more tinkering... the LFO seems fine, I realized the reason it was so hard to dial in, was because I'm using a linear 1M pot as a replacement. Ill have to wait until my 1m rev log arrives (still waiting for it), and hopefully that should help

regarding the feedback adjustment. would you other poly builders mind telling me what the feedback sounds like on your clones ? on my original polyphase, with the feedback fully cranked, I get a kind of whistle/timbre, nothing too extreme, but still audible through the range.  on my Clone it  starts squealing about halfway through, eventually becoming static noise, then cancelling the sound out.

lastly, when in envelope mode, I don't detect any LFO modulation whatsoever, have you guys had this problem as well ?

I get the sinking feeling there are some differences between the factory schematic and the finished polyphases. I spoke with Howard Davis about these polyphases a couple years ago, and he mentioned a feedback adjustment pot, so maybe we're missing some stuff

-rec

[RaceDriver205]

I just finished mine too, and its not working. The LFO on mine is working (I have a scope now thank god), but the signal does not seem to be going through right.
Don't spose you have a scope too?

[Rectangular]

sorry, no scope.  what kind of problems are you experiencing specifically ? the phaser stage doesn't sound right ?  have you gone through it with an audio probe ? its a pretty straight forward opamp array, just follow the signal through each 4558 and make sure there's signal at the two inputs of each 4558, and one at the output. what are you using for the optocouplers ? I made my own using leds/ldrs, so I can clearly see them responding to the LFO/env signal

once you get it making sound, I'd be interested in hearing how your feedback knob is affecting the sound. I haven't heard back from Lowstar, Dave B, Nelson, or the other poly-builders, but the feedback response seems to be all wrong

I think a lot of my problems are power supply related. but I've brought that into another topic

-rec

[RaceDriver205]

Ive went through the opamp stages, and as I remember it stuffed up at the first one. I couldn't work out why. Unfortunately, Nelson etc have not been around the forum for a fair while. They may have moved on, which I hope to do sometime in the near future 

[Rectangular]

did you try swapping the 4558 with another one ? always the chance its defective. I don't want to list out all the basics, but double check the surrounding voltages, make sure there aren't any cold leads. its a really tight layout, so there's a lot of opportunity for errors. if you've ever seen the inside of an original poly, its a made up of about 5 spacious matrices, with the resistors all facing the same direction, very tidy. if I can't get mine to sound any better I'm just going to 100% duplicate the original layout. its very easy to make mistakes when the components aren't named. I can't refer to any junction off-hand

keep me up to date ! 

-rec

[lowstar]

would you other poly builders mind telling me what the feedback sounds like on your clones ?

just like the feedback on the bad stone (sorry if that sounds funny  ). with no feeback, the phasing starts kinda lean and crisp, and the more feedback is introduced, the more spatial, broader and oyeiouyeiou-like the sound gets.

did you try swapping the 4558 with another one ?

if you search for my debugging posts, you will see that i had a problem with the feedback pot that was related to the type of 4558-type opamps used...rc4558p did not work for me, tl072 totally failed, ba4558 (a cheap generic copy of the 4558) worked. may sound funny, but that´s how it really happened.  so try different opamps maybe.
my main problem though was a short on the lower-right-hand-corner 5088 though, so i guess you´d be best of searching for shorts, checking trace continuity and double check resistor values. you´ll prolly say "yeah, that´s what i did already 10 times" but in the end, there´s always one little s****y thing we overlook (at least in my cases, it always was).

cheers,
lowstar

[Shepherd]

Can it still handle "hot" signals without the thermistor in it?

[RaceDriver205]

Er, none of us know what the thermistor is for.

[Rectangular]

lowstar. so you noticed a significant difference in the feedback response, when you swapped the your rc4558s with a ba4558 ? I don't know if I can get ba4558s locally, but Ill give it a look.  maybe the ba4558's worked because the 4558's in 1978 were very cheap, high noise, especially if EH was buying them bulk from japan or somewhere else (usa ?).  a lot of my troubles so far are linked with underpowering this unit. I tried running it on a pump charge circuit (to get it up to 18v) but it drains over four times the current than the pump charge can provide, so I need another solution

Race and Shep: I may be wrong, but looking into my poly,  the thermistor is just a tiny little black nub. its located in the LFO/env control section of the pedal, so the sound isn't being affected, I think it just keeps the LFO freq in check, or maintains the right rectification for the envelope, neither are *too* crucial

[lowstar]

UPDATE on my noise issue w/feedback pot:
went through each stage with an audioprobe, and, sure enough, one of the rc4558p´s was defective. replaced it with a ba4558 and most of the noise in general in the feedback path and the shhh on the first 10% pot rotation was gone.  icon_mrgreen
at the end of the pot travel, there was still a spot where it would come back again kind of abrupt, so i thought "ok, let´s switch them all for tl072 to bring down the noise floor". did it, but at the spot where the 4558p´s would bring back the noise, it now sounded like a mobile telephone trying to make a call... Shocked

so i swapped all the 5 opamps for ba4558 (actually cheap generic opamps), and...bingo ! now it even behaves in the last 10% of the feedback pot rotation, and the overall noise floor is also better.
a funny finding, i always thought those rc4558p´s were way better than ba4558...

that´s from my original post about the polyphase and the feedback-pot issues.

i think you should try running the thingie from ~20 volts and see what happens.

btw, could you inspect the thermistor closer and try to read what´s printed on there ? i´d be really interested to put that finishing touch on mine, even if it doesn´t have much impact probably.

cheers,
lowstar

[Rectangular]

hey lowstar

your feedback problem seems identical to what I've got, so I'll look into finding some ba4558s. another potential problem, is that I'm not hearing any LFO modulation in the ENV mode, not sure why that is. I presume yours has audible  LFO in env mode ? I've ditched the pump charge and am now just running it off a dedicated power supply, so I know the power issue is out of the way

unfortunately the thermistor is soldered underneath one of the pots, but I couldn't see anything labeled on it, its just a black tiny/short nub of a thing, maybe the size of a 1n4148 diode, with very short leads. even if it could be identified, its probably just as obsolete as the CLM8200
-rec

[lowstar]

so I'll look into finding some ba4558s.

or some other dual opamps, or at least swap the ones you have in there, maybe you have a defective one...the ba´s worked for me, but i´m sure there must be other dual opamps out there that do the trick...i just happened to have the ba´s at home as an alternative...

I presume yours has audible  LFO in env mode ?

yes, it´s impossible to overlook it (should i say overhear  ), once the envelope goes down, the modulation sets in, controllable by the mod depth and mod speed pots.

cheers,
lowstar