True CV from a pot?

[ExpAnonColin]

I'm having a bit of trouble here.  In the big scheme of things I'm trying to use 2 H11F3s as variable resistances in opposite to simulate a pot, each being controlled by the same expression pedal.  Unfortunately, LEDs are not super friendly to simple potential dividers-and I can't, of course, just wire the pot to ground for one H11F3 and to 9v for the other (in opposite).  And simply switching the pot's center lug from cathode to anode won't help either-LEDs aren't super friendly to taht, like I said.  They need a better ground, not a ground with a 10k resistance to 9v.

But I know that in many cases you can easily switch, say, an LFO from anode to cathode to invert the LED's flashing, for example.  Which is because, I assume, that when the LED cycles low, it is actually low-there is no connection to any other voltage in there.  It's just low, it's what the LED is looking for,so it can work either way.  My question is, can I still use the 10k pot (I have not experimented with larger values simply because I am being stubborn and want it to work with my 10k expression pedal) to create a clean, 0v-9v (or 4.5v would be acceptable) control voltage based on the position of the pot, using something like a variable voltage regulator?  Or... something?

-Colin

[Paul Perry (Frostwave)]

If the subject really means, "true pot from CV", then a way to get close, as I think has been pointed out somewhere on this board before , is:

Suppose the cv goes from 0 to 9.
Then run the cv to two resistor /led series combos, one goes to ground and the other to +9.
Imagine what happens when the cv is 0, one led is off & the other at max.
Opposite, when the cv is 9.
And at cv = 4.5v, the diodes match.
So, you have what you want! (less a bit of dead spot at each end, get around that with a slightly narrower range of cv input, say 1.5 to 8.5v).

[ExpAnonColin]

But the problem with such a system is that I DON'T have a CV as such.  I have a CV where at either end of the pot's travel, I have the 9v or 0v, but with the stipulation of a 10k resistor from the other end.  So when the LED that's connected to V+ and needs a groudn to light wants to light, it doesn't go all the way because there is still a 10k resistance from 9v on the cathode.  That's why I want a true CV here-I don't have one yet.

...I think.

-Colin

[Paul Perry (Frostwave)]

I can't quite see what you are trying to do.
I assume you are trying to make a pot from a pair of H11F3s, and that you are trying to control them from a CV.

But.. is this other CV being generated from a pot?? with a resistor in series?? I don't get it at all.

"Two nations, divided by a common language." - Winston Churchill

[R.G.]

You can do a much better job of simulating a pot from a control voltage by doing a two-capacitor simulation of two resistors in series and controlling the duty cycle of the two caps inversely. It's effective and linear - which photoFETS and LDRs are not - if complicated.

But Paul's right. Until we have better clues to what you're trying to do, we're just shotgunning bats in the dark at midnight.

[ExpAnonColin]

OK, I tried to explain it as best I could.

I'm trying to use a 10k pot to control 2 H11F3s to create a potential divider which functions as a pot of any of various values.

The 10k pot is wired as a potential divider with one lug to 9v, one lug as the "output", and the other to ground, creating a CV as Paul was describing.

However, it only works when the CV is going to the cathode of the H11F3.  When it goes to the anode, it does not light up properly (try it with an LED, just put the cathode to 9v and the anode to this CV system.)  I assume this is because I am using a 10k pot and the anode doesn't like seeing 9v through a 10k resistance.

But I am set on a 10k pot.  So I want to create a completely seperate CV, with no resistance to either ground or 9v, in some way, with this 10k pot-and this is waht I'm asking how to do.  Is there some way to use the 10k pot to control a variable voltage regulator to create a regular DC voltage that will control an H11F3 both on it's anode and cathode, much like a normal old LFO does?

RG, can you point me to more info on the method you described?  I am using H11F3s simply because they are able to have a resistance range of anywhere from 20k to 500k if tweaked properly.

-Colin

P.S.  It's funny because back in the day I remember I asked a question about what a transformer does, with no context at all, while I was trying to figure out what it did in a passive ring mod, and you tried to help RG, but I should have just been clearer in the first place as your answer would have been much more specific, so this time I tried to do so and still failed miserably

[puretube]

the 2 optofet`s LEDs antiparallell in the feedbackpath of an opamp might help the fwd voltage drop problem you`re facing...

[R.G.]

OK. I understand what you're doing right now. I still have some questions about where this is going.

Let's dispose of that. As I understand what you're doing now, you want to use a 10K pot to create a control voltage that you then apply to some collection of stuff that ends up in two H11F3 LED-to-photoFET modules working as an isolated pot. The obvious question is - how come you don't just use the "10K" pot as what you want to do? One possible answer is that the 10K pot is just taking the place of some control voltage for testing purposes only.  Perhaps there are other reasons.

Albert Einstein said that everthing should be as simple as possible, but no simpler. I think you're trying to do things too simply.

If you want a truly voltage controlled setup, you want the LEDs' current to be controlled by the voltage. As it is, you're using the 10K to modulate their current directly, and that's tough, especially if you're trying to get no current to full current from the LEDs.

One possible way to get this is to use the 10K as a current divider - you hook the LEDs' anodes to +9, the cathodes to the outer lugs of the 10K, and the wiper of the pot to ground through a current limiting resistor of maybe 470 to 1K. The current limiting resistor sets the max LED current when the 10K pot is at one or the other end of travel. In this setup, the LED currents alternate from LED to LED as you turn the pot. Note that 10K is probably not the correct value to make a smooth transition from on to off, and there's going to be a big region in the middle where nothing much happens. Worse yet, this is not a control voltage setup.

Another way is to sub in a differential amplifier for the 10K pot. You tie the two emitters of a pair of NPN transistors together, and run this to either a current source (NPN, diode, resistors) in the joined emitters or a resistor to some negative voltage. The 10K pot supplies a voltage to a voltage divider to the base of one of the transistors, the other transistor base is tied slightly above ground, perhaps by a diode drop. Now the 10K resistor causes current to be transferred through the most "on" transistor, and that varies fairly linearly with the 10K resistor setting as long as your voltage divider makes the variation at the two bases be about 25mv peak to peak.

A simpler to understand way is to note that in any linear opamp amplifier, the feedback element is being run in constant current by the feedback of the amplifier. The current is equal to the current in the inverting input resistor. So you can hook up two opamps, one following, one inverting. The LEDs are in the feedback loop. The biasing arrangement is such that one opamp has its output feed its LED 0-max current as the 10K goes from ground to 9V, the other inverts so that its LED goes from max to 0 as the 10k covers the same rotation. Now you have a truly voltage controlled setup where a single CV makes the current transition happen.

Good luck with the secret project.

[Paul Perry (Frostwave)]

If there is a low impedance 1/2 Vcc available, then combining RG & puretube, you only need one op amp. I think.. :roll:

incidentally, all this begs to be combined with the PWM thread!

[ExpAnonColin]

the 2 optofet`s LEDs antiparallell in the feedbackpath of an opamp might help the fwd voltage drop problem you`re facing...

Then what would I use to control the voltage?  Noninverting input?

-Colin

[ExpAnonColin]

OK. I understand what you're doing right now. I still have some questions about where this is going....

Good luck with the secret project.

It doesn't have to be a control voltage per se.  And there is no secret project, I am simply trying to create the end-all-be-all 10k to any-value-pot active system for expression pedals.

I'm going to need to reread all your posts... lots of good ideas.

-Colin

[puretube]

on 2nd thought, the antiparallell won`t work like you need it...
(it`s either LED gets brighter/darker, depending on which side the opamp output goes, compared to Vbias...the other stays dark  :oops: ).

but with a 2nd opamp as cv-inverter, the LED in FB-path should work fine.