Why does a Veils channel in exp. output negative voltage with no input


I’ve noticed that a Veils can output a negative voltage if in exponential (or near exponential) mode when the input level attenuator is turned up. I’m curious as to why that is. Looking at the schematic it seems like there’s an AREF_-2.5 that is input in a couple different places in the circuit, and I’m curious if this voltage is somehow “leaking” into the output signal path when the resistance of the exponential potentiometer is low. Note this is probably way off because I know nothing about circuits, hah.

A neat little trick: That voltage can be used as an offset/bias similar to some other VCAs when you mix in a CV signal into the input of a subsequent channel. You can then take this mixed signal, invert it, and then use it in the cv in of another channel of Veils.

I wonder if the circuit could be altered to provide a positive voltage instead. This would remove the need to invert the signal.


This is due to control feedthrough. The negative voltage applied to the Vc pin of the 2164, required to obtain a high gain, also leaks to the input. And this small negative current is ridiculously amplified because of the extreme gain you can obtain when the exponential response knob is turned fully CCW and the gain knob fully CW (a gain which is well outside the range at which the 2164 is well-behaved).

The 2164 can be made to operate with less control feedthrough, but this can only go away completely with a trimmer.

Streams doesn’t have this quirk because the gain range is tamed, even with the exponential response turned fully CCW (the gain never goes above 20dB).


Thanks for the explanation!