LM13700 OTA with Opamps instead of on-chip darlingtons

KammutierSpule · Joined: Feb 07, 2008 Posts: 40 Location: Aveiro - Portugal

I have seen some designed using the LM13700 OTA that instead of using their on-chip darlington, they use an external Opamp to the effect.

What they are trying to solve by using Opamps?

ixtern · Joined: Jun 25, 2018 Posts: 145 Location: Poland

PHOBoS

dslamnig · Posted: Sun Feb 07, 2021 10:30 am Post subject:

From my hands-on experience:

It's the classic observer/observed problem. There's no observation without interaction. In quantum physics (at least the Copenhagen interpretation Wink

observation causes the collapse of the wave function.

But back to macroscopic, it's the triangle or saw waveform that will be affected. In the usual VCO core setup a cap is fed a constant current from the OTA, so the voltage rises or decreases linearly - that's the waveform. As long as you don't touch it it's perfect (if we disregard OTA and cap non-linearities). But when you connect a resistor to the ground and the gate of the input transistor of the buffer, the current leaks in a non-linear way and the waveform is compromised.

I found, as a lot of other people before me, that the least invasive way to observe the cap voltage is to connect it (only) to the positive input of an opamp, and wire the opamp as a voltage follower. TL07x has JFET inputs, so the current will only leak to charge the JFET gate cap, and this is much less then when using a buffer.

And yes, there's the buffer output offset problem, too.

I learned all this the hard way when building my dual-core VCO:
https://electro-music.com/forum/topic-72488.html