bipolar or fet buffer?

[featherpillow]

Okay, here's where things stand:

I tested the buffer running into the transformer with a square wave swept from 20hz on up to 20khz, and didn't observe any ringing. The tranformer was running into a 10k load, which I figured was probably an average load for this type of circuit to see. Output cap is a 220uf.

Reflecting later, if the d.c. on the output is 5V then there is something wrong. The FET should always have a zero-to-somewhat-negative gate-source voltage. With a +15V supply your voltage divider as described sets the FET gate at 7.1V. The output voltage therefore should be more than this.

The output voltage I measured was actually closer to 6vdc, about 5.96 or so, and it was measured after the 33 ohm resistor. The input is tied to ground via the 2M resistor, and generator was ref'ed to ground as well.

When I get home from work today, I'll post a revised schem of the buffer circuit.

 

[bcarso]

"The output voltage I measured was actually closer to 6vdc, about 5.96 or so, and it was measured after the 33 ohm resistor. The input is tied to ground via the 2M resistor, and generator was ref'ed to ground as well."

So, my point is, the generator was "grounding" the input. The 2M was doing essentially nothing. You want a coupling cap at the input so the source doesn't change your d.c. bias. Then I recommend that you adjust that voltage divider with the 2M to where the d.c. output is about 7.5V, or for maximum output swing. It will be close to ground but a little bit positive. If your source is referenced to ground you could get along with the 6V output though it is not optimal, and anytime you disconnect the source the buffer output soars to something like 6V + your voltage divider voltage, or about 13V.

Sounds as if you are in pretty good shape on transient response. What you really want to do in the long run is drive the beast from the source you will be using it with, as the source impedance will interact. But these effects tend to be seen at pretty high frequencies, well above audio, with some peaking for sources of a few kohms, and this won't likely be passed by the transformer anyway.

 

[featherpillow]

Here's a revised schematic:

I dropped the value of both R1 and R2, and added a .1uf input capacitor (c3, even though there is no C2). DC output of the buffer is roughly 7.5 volts, and I found that I needed to add R6 in order to discharge C1.

 

[PRR]

> I needed to add R6 in order to discharge C1.

Yeah-but.... 100Ω???

In fact, the transformer winding ought to discharge C1.

 

[featherpillow]

Yeah-but.... 100Ω???

Ha ha! I knew the second that I saw PRR's name, that I had made an error... :grin:

Okay, let me explain my reason (even if I'm wrong...)

I'm a little puzzled by how and when one should use an RC coupling network with a transformer. I'm under the impression that we should block all DC from the transformer, that's why I added the 100 ohm resistor to discharge C1--I thought I wasn't supposed to allow the transformer primary do it...

So that brings up another question--why the need to capacitor-couple this circuit going into the transformer, but not a phantom power stage with a transformer input?

 

[bcarso]

Phantom power is balanced---the d.c. potential is the same on both ends of the winding so no current flows.

With your transformer now, the capacitor charges---once. When you power down, it discharges---once. Generally no biggie.

No need for the resistor.

 

[featherpillow]

I see--that makes perfect sense!

Thanks so much for your help!