LM13700 as a voltage controlled R, lots of disadvantages

[ExpAnonColin]

I thought I had run out of H11F3s (but I didn't! phew) so I started doing calculations to use LM13700s as VCRs.

According to the datasheet, and to my calculations, the resistance that is voltage controlled can be modelled by the following equation:

Rx=(R+Ra)/((19.2)(Vc/Rabc)(Ra))

And Vc/Rabc can't exceed .001 A to prevent the thing from blowing up.

So if we wanted say 15k...  At the max, using 1-9v as a Cv, 1v would provide the max resistance as the denominator would be lower.

15000=(R+Ra)/((19.2)(1/Rabc)(Ra))

Let's simplify a bit

15000=(R+Ra)(Rabc)/((19.2)(1)(Ra))

Ideally we would want Ra to be extremely small so that the Vc would have a greater effect on the output resistance.  I am going to assume that hardwiring it is 1 ohm of resistance.  This is a bad assumption to make unfortunately (disadvantage #1)

15000=(R+1)(Rabc)/((19.2)(1)(1))

15000=(R+1)(Rabc)/19.2

To keep Rabc at .001...

V=IR

9=.001R

R=9000 or let's just use 10000 because I like standard values.

15000=(R+1)(10000)/19.2

Simplifying...

288000=10000R+10000
278000=10000R
R=27.2
Disadvantage #2: 27.2 is really small.
Let's just use a 100ohm trimpot, eh?

At 9v...

Rx=(27.2+1)(10000)/((19.2)(9)(1))
282000/172.8=1631.9
Disadvantage #3: 1.6k isn't exactly close to 0, though it's not a bad ratio.

Disadvantage #4 is that the resistance must be tied from ground.  Not always desireable.

How can we fix any of these?  Because if we figure out how, we have the ultimate expression pedal system.

-Colin

[StephenGiles]

:roll:  :roll:
Stephen

[gez]

Couldn't be bothered to go through all your calculations, but most circuits that incorporate variable resistance into the design have either a stop resistor or 'built-in stop resistance', so to speak.  This circuit would fall into the latter category and you'd factor it into the design.  

Why is 0 ohms resistance so desirable?  You'd be hard pressed to get it from any semiconductor.

[Paul Perry (Frostwave)]

You can (according to the data sheets) make a fully floating resistor from a couple of LM13700s. But I can't think of any cases, where you wouldn't be better off redesigining the circuit containing the resistor, to incorporate the required external control in a more natural way.

[ExpAnonColin]

Couldn't be bothered to go through all your calculations, but most circuits that incorporate variable resistance into the design have either a stop resistor or 'built-in stop resistance', so to speak.  This circuit would fall into the latter category and you'd factor it into the design.  

Why is 0 ohms resistance so desirable?  You'd be hard pressed to get it from any semiconductor.

I wasn't clear-I was looking at this to simulate a pot, in which case it seemed best to try to get as close as possible to 0ohm.  The H11F3 can do ~300ohm-50k without too much trouble.  If it was being used to cycle through resistances to control the gain of an OA or something, then yeah, it would be more desireable.  THanks for your input.

-Colin

[ExpAnonColin]

You can (according to the data sheets) make a fully floating resistor from a couple of LM13700s. But I can't think of any cases, where you wouldn't be better off redesigining the circuit containing the resistor, to incorporate the required external control in a more natural way.

I had seen the floating resistor example, but it was still connected to ground... pardon my ignorance, but why is that?

You're probably right (about redesigning the circuit).

-Colin

[ExpAnonColin]

:roll:  :roll:
Stephen

:? Is something the matter here?

-Colin

[gez]

I wasn't clear-I was looking at this to simulate a pot, in which case it seemed best to try to get as close as possible to 0ohm.

You made yourself clear Colin, the point I was making is that pots which decreases in resistance as you turn them are usually used in conjunction with a stop resistor.  When semiconductors are subbed for a pot there'll be some minimum resistance from a LDR/FET channel/whatever which acts as a stop resistor (if one isn't used) and gets factored into the equation one way or the other.  Leave out the stop resistor that would be used with a mechanical pot and factor it into the OTA design, or, as Paul said, find a more elegant way of doing whatever it is you're trying to do.

[StephenGiles]

No my friend, I just don't do maths - not since I left school in 1964!!!!
Stephen

[ExpAnonColin]

I wasn't clear-I was looking at this to simulate a pot, in which case it seemed best to try to get as close as possible to 0ohm.

You made yourself clear Colin, the point I was making is that pots which decreases in resistance as you turn them are usually used in conjunction with a stop resistor.  When semiconductors are subbed for a pot there'll be some minimum resistance from a LDR/FET channel/whatever which acts as a stop resistor (if one isn't used) and gets factored into the equation one way or the other.  Leave out the stop resistor that would be used with a mechanical pot and factor it into the OTA design, or, as Paul said, find a more elegant way of doing whatever it is you're trying to do.

I see, I see your point.

-Colin

[Transmogrifox]

I can't see a reason with reference to guitar FX to use an OTA as a VR.

For Trem/volume control--It's a a voltage controlled variable gain amplifier already.

For filters:  It's easier to make filters using the OTA as an OTA instead of a an extra elaborate resistor simulator, as in lower noise and less distortion.

For ring modulators and so on, there may be some value experimenting here.  Might be possible to make an arrangement with the simulated VR that better suppresses the carrier  :?:  :?:  :?:

Just my 2 cents (I might have to stick my foot in my mouth if you make a real interesting sounding filter that depends on an element of distortion or sweep transfer characteristic of the CV that can't be easily achieved w/ a standard OTA configuration)

[ExpAnonColin]

Hey Transmografix, I was thinking you'd chime in.

Like I said, my application was as a sort of universal expression pedal system-so that your 100k expression pedal could be used to control any pot.  And hopefully use the same board layout for all of these little active systems.  Otherwise, you're right.

-Colin

[gez]

I can't see a reason with reference to guitar FX to use an OTA as a VR.

RG came up with a good idea for Stephen's circuit:

http://diystompboxes.com/sboxforum/viewtopic.php?p=256034&highlight=#256034

I use LM13700s all the time and sometimes I only need one, so it's useful to know what you can do with the 'spare'.