Question about the opamp biasing in this schematic (DOD 490 Phaser)

[Yonatan]

Hi All,

I'm cross-posting this question from the other site, in the hope that someone over here might know: How does the power and opamp biasing work in this circuit?  As far as I can tell it is a singly supply setup (i.e. opamps get hooked up to 9v and 0v, right?), but it's not shown explicitly, and what's throwing me off is that there is not the "typical" 4.5v virtual ground (voltage divided from 9v) used to bias the opamps (to IN+), which I've come to expect to see wherever there opamps together with a single power supply.  Here, it's much different, with the input somehow being connected to both IN- and IN+.  It's just that I don't want to order the CA3094 chips before I'm confident that I'll know how to wire this up.

I started breadboarding it and I think that the input/output buffer section is working, since audio goes through it, though I'm not entirely certain because it looks like there's a passive path directly from input to output (e.g. without going through any opamp), but it seems noticeably louder when I power it on.  And I have the LFO working, confirmed by connecting a LED to the output of the LFO's second opamp.   



Thanks,
Yonatan

[Digital Larry]

That IS interesting.

At first, I thought the voltage divider for the reference was formed by the 18k with a cap across it plus the 22k that comes from the jack that completes the connection to +9 when the plug is inserted.  But that's acting as a low pass filter on the LFO output.  I still haven't figured out the voltage reference scheme.  This is a known working schematic?

The vertical resistor way at the left of the input op-amp is not marked.    

The input is only connected to the + input of that first op amp stage, which supplies the dry out signal.

[slacker]

At first, I thought the voltage divider for the reference was formed by the 18k with a cap across it plus the 22k that comes from the jack that completes the connection to +9 when the plug is inserted.

I think you're right and that is supposed to be the reference for the opamp and OTAs, so it should be connected to the line from the + inputs of the OTAs not where it is drawn. Either that or the schematic is incomplete and something is missing that is supposed to provide the reference voltage.

[Yonatan]

I don't know if it is a known working schematic (I'm going to try to determine that  ), but it was reportedly provided by Digitech themselves, so I tend to trust it. 

[samhay]

To me it looks like the op-amp/OTA supply is +9 V and ground, which are implied and not shown - a few choice filter caps on these would probably not go a miss.

There are separate Vbias for the audio and LFO sections - both 22k/18k. This makes sense from an LFO noise perspective.

[Yonatan]

There are separate Vbias for the audio and LFO sections - both 22k/18k.

Just curious, what establishes the Vbias for the audio section?

[samhay]

Actually, having taken a closer look, it is a little more weird than I thought.
The op-amp and OTAs are DC coupled. The opamp is biased via a 470k R to the (+) input of the OTAs, which should be at ~ 1/2 V+. The OTA's current control pin (5) is biased up to 22k/18k of 9V, so perhaps this is setting the bias point?

[slacker]

I don't think you can assume the OTAs will automatically bias up to half supply. Even if they did you've got some whacky feedback paths going on with all those nodes connected together.
I'm pretty sure the schematic is wrong, I just don't know how.

[Yonatan]

And the question is not just about the OTAs, but also about the dual op-amp in the audio path (buffer?)