Tim, how can I reduce hte offset of my servo?

[Category 5]

First off let me say all of my pres are sounding freaking amazing. The following is partly because I am obsessive and want everything to be PERFECT, and partly because I can hardly resist ripping the pres from the rack and messing with them some more.

One of my J99s servo amp has a bigger offset than the other. One has a rail voltage difference of like -11.7/10.8 and the other is real close (like 11.7/11.56)

The one that's close settles to 0 output offset extremely quickly, and basically stays there. The other one takes 30 second to a minute to settle, after switching gains, and then drifts a couple of mV above and below 0.

I know you're going to tell me it's a non issue. But if I get bored here, and am looking for something to tinker with, what do I need to change to get the servo voltages to be closer? The regulators themselves, or resistors? If resistors, which ones?

Thanks!
Shane

[tele_player]

That's interesting, I hadn't looked at the schematic of the J99 since I built my pair last year. Tim's changed the regulation on the voltages for the servo amps...

I'm curious, what voltage are you running your opamps at. For the 12v regulators to operate in spec, they need over 14.5v, if I'm reading the spec sheet correctly.

[tpryan]

Are you sure it never settles? The time constant of the servo is so long it may take a while for the oscillations to settle. The servo amps are connected as integrators, and since this is a "twin servo" there are two of them. Reducing the input resistors will reduce the time constants, reducing the output resistors will increase the leverage the servos have over the main outputs. The original Jensen values are quite conservative, but with good reason. If the time constants are significantly faster, the servos will attempt to null the LF content of your signal. And there's no need to apply a corrective signal much larger than the offset you're attempting to null. That's why the supply rail voltage doesn't matter that much.

[Category 5]

Thanks Tim. It pretty much settles, but changes from -1/2mv to +1/2mv averaging about 0. THe only reason I thought it should be better is that the other channel settles to 0 and stays there, but the preamps both sound fine and are pretty much matched to within a db from 15 all the way to 35k so I'll just leave it. Since I put it in the rack and wired it all up I have been reluctant to derack it and disassemble it just to tinker so i guess I'm on to my mic projects.

It's amazing the looks i get from singers regarding the N72. The initial response is "wow, this mic makes me sound incredible!". Little do they know!

Shane