Antilog converter conversion calamity

[armdnrdy]

I've been working on a Tycobrahe Pedalflanger build.

I would like to convert the circuit to accept the "standard" single rail, negative ground power.

I've worked through most of the circuit but have hit a dead end (of my current ability) at the antilog converter.

The original circuit's power section is positive ground. The converter's Q5 base is negative in relationship to it's source.

This configuration works out similar to standard antilog converter examples that I've seen. In a bipolar power scenario, the base is grounded and the source is +V.

The problem that I'm encountering is when converting it to a positive single supply, (negative ground) Q5's base would not be negative because +VR (the reference) is not true zero volts.

If I were to use actual ground for Q5's base I'm afraid that this configuration would interfere with the op amp's bias.

Am I overthinking a solution that I've already found but don't recognize?  

I can't find any positive voltage single supply antilog converter examples after two days of searching the net.

Any suggestions?

Here is the original positive ground circuit:




This is one of my proposed "fixes"




Here is a typical antilog converter for reference:

[R.G.]

Quite possibly the simplest thing to do is to put in a charge pump to make -9V and use a bipolar supply.

[armdnrdy]

Quite possibly the simplest thing to do is to put in a charge pump to make -9V and use a bipolar supply.

That was definitely a thought. I just wasn't sure if there was a simpler way to do this.
From my limited knowledge, it looks like an antilog converter doesn't work with a positive single supply.   

[armdnrdy]

I've noticed that in single supply Ibanez flangers 9 volts is used for the op amps, a voltage divider off of the 9 for the op amp bias, then a 5 volt regulator is used with a voltage divider for the BBD, LFO, and current conversion.

I suspect that something like that would work as well.

[armdnrdy]

Before I commit to adding another power rail, I'll bread board the original negative ground antilog converter, take note of how it works, then bread board a negative ground converter to try and achieve the same results.

So..the way that I understand the workings of this circuit is: If a 1 volt pp 400HZ sine wave is applied to the input, the frequency at the output of the VCO should be 400HZ. If the sine wave is increased to 2 volts, the frequency should be one octave above - 800HZ.

1 octave per volt. Correct?

I'll post the results.

Any suggestions are welcome.

[PRR]

> looks like an antilog converter doesn't work with a positive single supply.

It is much easier to design with *bipolar* supplies.

The original has a single supply, but then uses a mid-point reference so that the log parts can swing both ways from there. i.e. conceptual "zero" is really 3.6V from either rail.

And the final output is BIG CMOS-buffered digital. (*)

I *think* you can just re-name the 7.3V power pins. "Ground" is then the negative lead of your 7V power source. The bypass caps in the rail-splitter should now probably go to the other rail. No, I have not gone through node-by-node to check this, or looked for places where one might get confused.

(*) I smell a BBD chip as the destination of this digi-signal. And some of them were conceived as negative-supply devices. Is that what you are doing? Do you know how you will do it?

> 1 octave per volt. Correct?

That's a standard for big *patchable* synths. Here there's no (user) way to get in or out of the loggy stuff directly, no need for compatability with anything else in the world. Also if you are going to specify some "exact" V/Oct function, you may want pitch-accuracy, which will force you to use a tempco resistor; OTOH a trem-type effect does not need such accuracy.

[armdnrdy]

Hey Paul,

Thanks for taking a look.

I've reworked the circuit back and forth a few times because I'm unsure how the positive reference voltage is going to work with some of the circuit sections.

A few areas of note are the Gate and Limiter side chains. I've tried switching the negative reference with a positive reference but that didn't seem correct on paper.
Q1 and Q2 drain need to be more negative than their source.
I tried replacing them with N channel JFETs but then there were other issues.

Since those areas are side chains, I think that negative ground will work in place of the reference voltage. I'll bread board those sections to be sure.

Here are drawings of the whole circuit. One original and one my work in progress.

Original positive ground:




Negative ground: