EH Bad Stone (CD4009) - theory behind this fault?

[pringe]

I have just finished repairing a Bad Stone (CD4009 version) that came to me non-working, having had what looked like previous attempt(s) to fix it. Initially repairing loose and broken wires got it working, I also fixed an elec cap C3 which had been replaced backwards between GND and Vbias and some suspect looking lifted trace repairs. There was one fault remaining that was eluding me - the LFO would start up OK on power on but would gradually develop a spike on the positive zero-crossing which over a few minutes would become bigger, the LFO amplitude would reduce and eventually collapse. One clue was the CD4009 being very sensitive to a hand being placed near it, causing oscillation and all kinds of distorted LFO waveform if touched (it made for some interesting gnarly sounds though!) and needing a power cycle to reset. I suspected this IC might need replacing but doing so (and replacing the LM324) didn't improve things. The board is a bit of a mess from previous work so next thing was to try and clean up and make good some problem areas. Tracing the circuit through alongside schematics and testing components in the LFO area I eventually discovered that pin 14 of the CD4009 was floating and not connected to the node that connects all inverter inputs with C10 / R25 / R26! Restoring this connection fixed the problem and the LFO is now happily stable and the pedal fully functional again.

My question is regarding the theory behind this - in the circuit what is the role of the inverter at pins 14/15/16 of CD4009? What was the reason behind the gradual distortion and collapse of the LFO with this inverter input unconnected? I've attached a schematic that I found online, hope it's OK to post here.

Thanks!

[StephenGiles]

I was told that this circuit, although sounding very good when working, was subject to a "see what you can take away for minimum parts count and still make it work" exercise in order to cut costs. How true that is I'll never know, and still don't understand fully how the 4009 works here.

The Oakley Phaser which also uses the 4009 gives this explanation - Each of the five all pass filter stages uses one of the MOSFET transistors inside a CMOS invertor chip, the 4009UBE, to control the centre frequency of the stage. By using a single integrated circuit for all transistors means that no matching of devices is required as would be the case if we were using discrete components.

[pringe]

Ahh I see, so in essence it was used as a cheap way of getting a set of matched MOSFETs! Interesting, thanks for the info.

I've been reading up on the general theory on phaser circuits and think I'm getting to grips with it a bit...I'm still not sure what the 6th inverter is doing (pins 14/15/16 - connecting to Vbias via R44 6k8) but am I right in saying the other 5 inverters/MOSFETs are used to provide in essence a variable resistance for each all pass filter stage based on the LFO voltage? In this way the amount of phase shift for each stage changes, and so do the resulting notches? The 4558s also act as buffers allowing each stage to not load / be loaded by the others.

I'm also wondering why the 6th 4558 stage is different to the other five?