VCA with low CV feedthrough again

As said in my last post it is to much effort to use active CV feedthrough compensation for the result and to use the second LM13700 OTA for a second VCA. Meanwhile the PCB and the faceplate for my http://www.haraldswerk.de/NGF/NGF_VCA_DC_due/NGF_VCA_DC_due.html arrived.

20141112_01

After stuffing the PCB I selected an LM13700 OTA for best DC performance. As expected the DC performance was good. About 2mV CV feedthrough over the whole 5V range with a THD of 0,2%. But wait….. there is an second OTA as well. My assumption was both OTA behave the same because they are on the same die. Nope. I was wrong! The second VCA shows a CV feedthrough of 7mV over the 5V range. Not that good.

20141112_02

It is hard enough to find an LM13700 OTA with low CV feedthrough but it will be a lot harder to find an OTA where both sides behaves the same. This explains some trouble I have had with the active feedthrough compensation as well.

So I skip the idea to use both OTA in the LM13700 for VCA with low CV feedthrough. The way to go is to select the chip for one “good” site and leave the other site alone. Not a big deal, they are not that costly. I’ll do it with my next VCA the NGF VCA DC one.

If I find the time I will put up a test circuit to find an LM13700 with identical behavior on both sites and use it in my active feedthrough compensation design to see if I can do better then 1mV feedthrough.

20141112_03