Lowering the noise on my MPF-1

[Sage]

Hey, guys.  I've built a copy of the Maestro Parametric Filter (MPF-1), an old 70s pedal that basically creates a bandpass filter that you can boost or cut in your guitar signal.  It was used by Alex Lifeson back in the 70s, probably to enhance the upper midrange of his guitars when soloing.  I've successfully built a clone using this strip board layout.

The pedal sounds great, and can go from a subtle bump in the mids to a pronounced "c0cked wah" effect.  Unfortunately, it adds some noise when engaged -- there's a bit of hiss that gets added in, particularly when playing through an overdriven amp.  I'm sure part of this is just because it's boosting the already-present noise, but I'm wondering if it's not adding its own, and if I could clean it up somehow.

The schematic of the original is here (note the strip board layout removes the rotary switch and controller and replaces them with a couple of toggle switches, but this doesn't affect the effect circuit):


I'll be honest: I successfully built it, but with my limited experience/knowledge, I don't really understand the schematic.  So I'm wondering if it's possible to improve the s2n on this in any way.  In particular, I'm wondering if upgrading the TL022 chips in here to something else would improve the s2n ratio.  Any thoughts?

[Mark Hammer]

Probably.  Those are low-power chips, but not low-noise.  You'll often see them used for LFOs because their low current draw results in less audible ticking.  Keep in mind the pedal stems from an era when most pedals were battery-powered, so current-draw was an issue.

Now that you made me think of it, I picked up an old MXR 6-band EQ today, and it also uses TL022.  Looks like I have some chip changing ahead of me.   

[Sage]

What do you recommend I replace them with? I usually see the TL062 recommended as an upgrade to the TL022, but the TL072 is lower noise, and there are more hi-fi options out there. Can I use any op amp, like an NE5532?

[Sage]

I tried swapping in some TL072 chips but it didn't make a lot of difference. I'm wondering if the CD4007UBE is the source of the noise now. Could that be it? Is there a better part I can replace it with?

[Mark Hammer]

Sadly, I am unaware of any pin-for-pin substitute for the 4007.  On the other hand, maybe you can try plugging in different 4007 chips and seeing if one is quieter than the others.

Finally, I am given to understand that Bi-Fet op-amps show their best noise characteristics with higher input resistances than the MPF-1 uses.  Perhaps NE5532 might yield better results than TL072.  But this is speculation on my part.

[Sage]

Well, thanks for the advice.  I tried an NE5532, but it didn't audibly lower the noise level.  I'll note that the hiss is far less noticeable on clean amps than overdriven ones, and adjusting the "Height" knob (increasing/decreasing the volume of the bandpass filter) has no effect, either.  At this point I'm convinced the CD4007 is the culprit, as it's meant to be used as a digital logic chip, but seems to be configured here as a sort of inverting op-amp.  When I read other instances of this chip being used in this particular way, it always seems to be regarded as sort of a hack.

I'll try swapping in another 4007 to see if it makes a difference, but I doubt it will.  I can live with it as is (it's really not much louder than my vintage-reissue script Phase 90).  Was just hoping to find a quick and easy improvement that didn't require changing the layout.

[pinkjimiphoton]

try a smallish cap from about R12 area to ground. start with 100pF, somewhere between there and 4.7n you should be able to nuke some of the hiss without @#$%ing up the signal too badly.

or try and put a small cap in series with R17, you wanna just roll off the hi's enough to nuke the hiss. but it MAY adversely affect the sound of the circuit. maybe try Mark's StupidlyWonderfulToneControl there... just so ya can shave off the hiss.

i don't think the 4007 will make much diff here.