LM13700 control current appearing as voltage at output?

I've got an LM13700 in this VCA configuration, and for whatever reason the gain control input (I'm applying a triangle LFO, ranging from 2-7v) is appearing as a voltage at the output. Does anybody know why this could be happening? :

• Are you applying LFO to gain input or Vin? Commented Aug 17, 2020 at 6:36

With change the bias current (gain) of the amplifier, also changes the input bias current and the input offset current. The input offset current generates voltage at 470 Ohm resistors, which is more visible at the output of the amplifier.

It will not be perfect, but it can be improved:

-With smaller resistors at the input (470 Ohm > 47 Ohm).

-By zeroing the initial offset voltage. (Minimum triangle signal at the output without input signal.)

• Hmmmm. Lower value resistors on the inputs to Vref did not appear to do anything. However, what did help a bit was instead of having two 470 ohm resistors, I used a 1M ohm to Vref from the noninverting input, and then wired the inverting input straight to Vref. That improved things about 25%, and then the key after that was to adjust Vref down by about a volt. Perhaps in doing that I achieved what csabahu meant by zeroing the initial offset voltage? Curious for anyone's further input on this; I kinda got a solution but I don't really understand it. I'm using a single supply if that matters. Commented Aug 18, 2020 at 4:17

The output of any differential input VCA is related to the input by an equation of the form:

$$V_{OUT} = V_{GAIN} \times A \times (V_{NONINV} - V_{INV})$$

Where $$\A\$$ is some constant, so that total voltage gain is $$\A\cdot V_{GAIN}\$$.

If you have unchanging, unequal potentials at the inputs, such that

$$V_{OFS} = V_{NONINV} - V_{INV} \ne 0$$

then the equation becomes

$$V_{OUT} = V_{GAIN} \times A \times V_{OFS}$$

Since $$\V_{OFS}\$$ is constant and non-zero, $$\V_{OUT}\$$ will vary proportionally with gain control signal $$\V_{GAIN}\$$. Could this be what you are seeing?