THAT 1570 1510 and 1512 Input/Output Circuits

[mediatechnology]

What about using a FET buffer (LSK170?) with say 2K source and drain loads and also use those loads as the 1570's bias resistors? The +/- inputs of the 1570 can be treated as independant "halves." I'm wondering if you AC couple the Rgain line (or 1570 outputs) if you can overcome the Vgs that becomes a 1570 output offset. Look at a Schoeps capsule buffer and see if you can use that without T2 and T3:



I haven't had to interface electrets to the 151X/1570 before, but it seems like the JFET ahead of the 15XX may be what you need. Otherwise the impedances seen by the 15XX are too high.

[MIQ]

Thanks Wayne,

The Schoeps CMC circuit is classic, and seems to be used inside a lot of condenser mic bodies even today. I suppose I could bias the LSK170 (or other low noise FET) gate with considerably lower impedances since it is being fed from another FET (and not a bare capsule).

I like the idea of doing the impedance conversion and making a balanced drive (phase splitter) for the 1512's input in one stage. I've got the Rgain line AC coupled but I am worried about what the output offset is going to look like since the output of the 1512 is directly coupled to a 5532 as a 0 to +6dB trim stage and that's directly coupled to a 1606 line driver.

I'll start playing with it. Thanks.

-Mike

[mediatechnology]

Thanks MIQ.

Yes, the 1G gate resistor should be lowered by many orders of magnitude.

In your original post you mentioned the 1512 and I responded with a 1570. Sorry wasn't paying attention.

Since you're using a 1512, you will definitely want to AC-couple the FET source and drain loads to the 1512 inputs and bias resistors. If you DC couple into the inputs, making the FET's loads also act as bias resistors as I originally suggested, there's going to be a large differential voltage develop. If Rgain is AC-coupled, the differential offset will not be amplified by the gain, but it will be passed directly through the inputs, appear across Cgain, and finally appear as a large output offset. (With the 1570 you could AC couple the outputs since the output is differential.)

Instead, with the 1512, look at using the 2K loads and consider AC-coupling into 2-10K bias resistors. I would recommend the low end of 2-10K. Making the input Cs between the FET and 1512 large will also keep the LF source impedance down. If the reactance of the inputs Cs are sufficiently low, then the FETs loads and Rbias appear in parallel at AC. This keeps the 1512's current noise managable.

Have you considered AC-coupling the 5532 trim stage? You might want to do that to keep DC out of the pot if there's going to be one.