[sdiy] The TL074 has changed!

[mskala at northcoastsynthesis.com]

On Tue, 5 Aug 2025, Mike Beauchamp wrote:

> I've modified a PCB with the old TL074 so I can measure the current between
> Q14 and the LM13700. The saturation (or as I call it my intentional
> "squish") of the circuit is noticeable. With the knob at 100%, the current
> is 32mA. If I add a 5V CV to resonance-CV the current only goes up to 37mA.

I hope you mean 0.32mA and 0.37mA.  The absolute maximum for the LM13700
is 2mA.

> With the new TL074H, the current drops immediately to 0mA as soon as the
> resonance-CV goes over 1V with the knob at 100%. So I don't think it's
> clipping, I think it's doing something entirely different.

I think part of the issue is that the voltage at the bias current input of
the LM13700 needs to be at least a volt or two above the negative rail (it
is the input of a current mirror).  If the op amp's output goes below that
voltage, then it's hard or impossible for the transistor to pass any
current at all; the current through the emitter goes through the base into
the op amp instead.  And this is something that could be different between
the 074 and the 074H, if the 074H is able to drive its output closer to
the negative rail.

My suggestion would be to measure the voltage at the op amp's output and
see if that's different between the two chips.

-- 
Matthew Skala
North Coast Synthesis Ltd.