FET Preamp Cable

[ggedamed]

You could try a 2SK30A (TFT) or 2SK208 or 2SK184 (SMD) with the -Y suffix or (even better) -GR.

[bancika]

I might be able to find 2SK30A. Which bias resistor should be suitable for that?

[PRR]

> how the simple one biases with 48v and 2.2k.

The 48V never goes NEAR the FET!!

The FET is powered by 12V, Zener regulated.

[diydave]

I wouldn't bias the fet directly from the 48v.
I'd use a voltage-divider (regulated zener, or just resistors) to get a reference-voltage of 12 volts max.
By doing so, you can use your ordinairy fets (j201, 2n5485, 2n5487, or whatever) and ordinairy bias-resistors and stuff the fet in the jack.
Surely the gain will be sufficient for a piezo.

[bancika]

I breadboarded it yesterday, but used 1.8k source resistor (didn't have 2.2k). Strangely, it works even when I turn 48V off on my interface. Apparently, there's voltage there at all times. I get around 4V on the drain with 48V turned off, and 19V when it's on. It produces sound in both cases but there's a lot of low frequency noise and pop every few seconds. Maybe it's because the circuit is exposed on the breadboard and all the unshielded clip-on leads I have running from between the breadboard and two cables (XLR to interface, 1/4" to guitar).

Btw, I confirmed with the guy who designed the circuit that it works fine with 48V.

Now I don't know what to do, try to build it properly and see if the noise is still there or switch gears and built something more elaborate in a pedal. I found a drilled B size box with jacks and everything...hmm, what to do, what to do. Anyone knows some good preamp that would be a good candidate for a pedal?

[ggedamed]

I might be able to find 2SK30A. Which bias resistor should be suitable for that?

First of all, please take everything in my post with a grain of salt. I'm just posting this to see if I got it right.

That's how I'd do it: the JFET will have to be measured beforehand. Let's consider a 2SK30A-GR with I_DSS = 5mA and V_GSoff = 3V. I'll try to bias the JFET at V_GSoff / 2 = 1.5V. So, in this circuit the source will stay at 1.5V.

For maximum headrom I'd like to have half of the available voltage plus the bias voltage on the drain (48V / 2 + 1.5V). The drain resistor is the 6k8 resistor in the XLR interface, so my ideal current I_D should be 25.5V / 6k8ohm ~ 3.75mA. It looks like my transistor will fit into the specs. The drain-source voltage V_DS should be around 24V (25.5V - 1.5V), which means the power dissipation, at 24V * 3.75mA = 90mW, it's in the 100mW limit.

This means the source resistor should be around 1.5V / 3.75 mA ~ 390ohm. Of course, next step would be the breadboarding.

As you can see, JFETs with large V_GSoff and I_DSS match better this circuit. It doesn't hurt that 2SK30A has lower noise specs and maximum drain-gate voltage of 50V, also. It also has much larger headroom, which you may like or not.

[bancika]

Update: built the preamp inside the jack (did a pretty neat job I think) but noise issue I mentioned before persists. I recorded how it sounds, anyone can think of anything that can cause it? Doesn't sound like regular ground loop or interference noise. And there's a periodical pop.

https://soundcloud.com/bancika/preamp-noise

What I'm still puzzled is that it works even when I turn off phantom power on the interface. There's still around 4V on the drain. Actually, it works better that way, when I turn on 48V noise level rises, while signal stays about as loud.

[bancika]

update 2: I breadboarded the Tillman's power supply that he built in a separate box and in that configuration it works fine, so I guess it's the interface and it's phantom power that doesn't work well with the preamp.

Cheers and thanks all for help.