I’ve been playing with the LM13700 and want to make a VCA with it. But first, I like to understand better how it works. I’ve breadboarded the VCA schematic from the datasheet.
I have modified some values to make it work better with 12V rails since the datasheet uses 15V rails. As the gain control, I use a voltage divider between my 12V and -12V rails.
At this point, everything seems to work well, but I cannot get a gain higher than say 0.9. I’ve looked at the formulas I found on wikipedia:
I mainly do not understand this Gm term. It must be the current that controls the gain, but according to my multimeter, this is already going over 1mA where maximum rating is 2mA. The second part that is important are the differential inputs that have the 1K pot across them. I don’t really understand what is going on with the ground in the middle there though. If I put my scope across inverting/non-inverting inputs, I get a signal of about 35mV p/p while my input is 11v p/p and output 10v p/p so roughly gain of ~0.9.
Can anybody give me some directions to allow me to get higher gains out of the LM13700 ? Or is it not meant to give >1 gains and do people just add an opamp afterwards?
What R_load are you using? As seen in the formulas you quoted the gain depends on the load resistance.
Hah! This is the simplest fix indeed. I changed the RL (22k) resistor for a 33k one and now my gain is slightly above 1. Unfortunately, my cheap scope is kindof jumpy so I cannot exactly see what the gain is, but its clear that tweaking this resistor works.
RL is a fixed load to get a stable voltage out of the LM13700. The buffer is used to send this voltage as the output. Pretty clever how you can reliably convert a current back into a voltage I didn’t know that.
It’s Ohm’s law, so shouldn’t come as a surprise (well, ok, people thought Georg Ohm was crazy when he came up with that, but these days it’s pretty accepted.)
(see Dud’s link for how to wire this up with a proper opamp instead of the built-in darlington).
See the Electric Druid link for an active I-to-V converter, too.
In synths I think it’s usual for a VCA to have unity gain at maximum, by the way. Oscillators are already 10 Vpp after all!
Speaking of which, the datasheet circuit you started from apparently is intended for a 1 Vpp input signal. RIN has nothing to do with the supply voltage; it and the pot form a pair of voltage dividers to bring the input signal down below 60 mV. If you want to use this with a synth signal, you need to reduce that voltage about another factor of ten.
The 30k at upper right also has nothing to do with the supply voltage but converts the gain control voltage to a current. So it also depends on the size of the CV.
And as you’ve seen RL has to do with setting the gain, not the supply voltage.
Again, see the discussion at the Electric Druid lilnk.
The pot is just there to trim the 0 V bias, btw. Most DIY designs tend to use two identical resistors instead (e.g. the two R2s in my post that Dud linked to), with one of them being part of the input voltage divider.
The thing to note about OTAs is that they output current - not voltage. Resistor RL in the schematic is there to convert the output current to a voltage for driving the darlington pair buffer.
Here’s my take on an lm13700 based VCA that accepts control voltages up to about 10V:
U1A and U1C do the current to voltage in this design, I don’t use the lm13700 buffers.
The bias current (I_ABC) that does the volume control is provided by the op-amps driving the pnp resistors which act as voltage controlled current sources.
Yeah I know. Ohm’s law appears in so many shapes and forms everywhere that sometimes it is a surprise. I just wasn’t aware of how you would build an I-to-V converter using a resistor and a buffer, but it seems super obvious now.
In synths I think it’s usual for a VCA to have unity gain at maximum, by the way. Oscillators are already 10 Vpp after all!
I didn’t know that. I do not own a single synth and I’m trying to design modules now so knowing things like this is definitely helpful for the end-product to make sense. I just didn’t want a VCA that would have a gain <1 if you turn it all the way up so I rather have 1.1 as a maximum or something like that.
it and the pot form a pair of voltage dividers to bring the input signal down below 60 mV.
Thanks. Makes sense! The amplitude I was feeding into it was pretty hot though and it seems to work now, so I think I’m ready to commit this one to protoboard
I didn’t know that.
(well, ok, people thought Georg Ohm was crazy when he came up with that, but these days it’s pretty accepted.)
The amplitude I was feeding into it was pretty hot though and it seems to work now, so I think I’m ready to commit this one to protoboard 