Transformerless JFet mic preamp

[bcarso]

A resistor between the source and common at that point in the circuit for instance---with zero R you drive the gate and get the transconductance times the gate voltage change as a current change at the drain. With the resistor in series the new equivalent transconductance if you will is

1/(1/gm + R);

So if gm (transconductance in amps per volt) is 1mA/V, you get a drain current change of delta V in times that number. Put 1k in series now and yu get half that change for the same gate voltage change, approximately. We assume here that we are operating with the drain voltage well above the pinchoff voltage---say 5V relative to the source, and that the transistor is operating below its ~max current, Idss (the drain current at zero volts gate-source for a JFET, which are always depletion-mode devices unlike your familiar power DMOS devices).

This is an example of local feedback, and usually works to linearize the stage to which it is applied, although there are exceptions ;-).

Brad

PS I will be out of internet access for a bit hence unable to respond for a while.

 

[Svart]

that's great! thanks!

 

[FredForssell]

I just took a quick look at the Borbely mic preamp link. Correct me if I am wrong, but for Erno's JFET mic preamp to be used without an input transformer, you need to use two of his cards (and an output transformer). My JMP-1 card requires only gain control and phantom power components... and of course a power supply.

That said, I cannot tell you how much I respect Mr Borbely and his work. IMO, he along with John Curl and a few others, are living treasures in the audio design world. I think that anyone using pretty much any of his designs will be quite happy doing so. I know I would be.

Cheers,

 

[PRR]

> Correct me if I am wrong, but for Erno's JFET mic preamp to be used without an input transformer, you need to use two of his cards (and an output transformer). My JMP-1 card requires only gain control and phantom power components... and of course a power supply.

http://www.borbelyaudio.com/eb103213.asp

> EB-103/213 is a high quality, high-gain, low-noise microphone preamp in ALL-FET technology. It is designed to work with input/output transformers, however, due to its extremely low input noise; it can also be used without transformers. ...... Input/output transformer(s), application dependent components and gain control are not included in the kit.

It had me fooled at first too. But the "second input" is the feedback net, not the Gates.

> Balanced transformerless mic pre using two 213 amps.
Fig. 3 shows two 213 connected as a balanced, transformerless mic pre. Using the 54dB maximum gain option for the 213 will give 60dB gain in balanced mode.

Two of the 213 modules. And "transformerless" means "input"; the ouputput is transformered.

In my experience, in many-many situations, condensers and dynamics (and even hot ribbons) can be worked well un-balanced. So a single 213 is potentially a super-sweet iron-free mike preamp. But what a LOT of silicon!

 

[pstamler]

It looks to me like you could eliminate the output transformer from the two-unit Borbely circuit if you knew you're always be going into balanced loads. Of course, then your common-mode rejection would depend on the load's rejection, and you'd lose the possibility of a floating output.

Peace,
Paul

 

[tk@halmi]

Chris,

How about a FET based dual stage current feedback amp similar to what PRR has published with BJTs a couple of months back?
It could be done with in complementary fashion so no capacitors would be required between stages.
You would need a dual revlog pot to control gain, but it is possible to get them.
The transistor count could be kept under 10 plus the follower stage on the output.
Another way would be to cannibalize the Forssell version of the Cohen (AES 2106) circuit for the discrete stage and double it up. Then add a follower stage.

Tamas

 

[analag]

Is there a difference in sound quality between the Toshiba 2SK170 and the Linear Tech LSK170?

Analag

 

[tk@halmi]

[quote author="analag"]Is there a difference in sound quality between the Toshiba 2SK170 and the Linear Tech LSK170?
Analag[/quote]

I doubt that I would hear the difference. Although I am sure that you could read a conversation on www.diyaudio.com about claims on how the harder to get and more expensive parts sound better.
Looking at the spec, Linear has replicated the Toshiba process pretty well. Based on my experience, if your circuit uses global feedback swapping different brands of the same type of small signal transistor should make no difference whatsoever. Local feedback only circuits are inherently fussier.

 

[chrissugar]

I supose Linear's fets are similar because their goal was to reproduce the Toshiba FETs. Unfortunately I have no experience with them.

chrissugar

 

[bcarso]

They sent me a couple of samples but I haven't had the time to play with them yet.

Their planned (maybe out now) dual to emulate the 2SK389 dual should be interesting---the guy claimed that it would be quieter than Toshiba's discontinued part because of the interdigitated construction. So it may well be different in other ways.

 

[analag]

During the time when I was building the Fet/tube cascode input mic pre (Fetcoded) I ordered the LSK170B samples from Linear Tech and I also bought a bunch of 2SK170B from Ampslab which I'm still waiting for to arrive, in that time I got the LSK170 from Linear Tech and I used them with great results, so my experience is with the LSK170B. My thought was if the 2SK170 is better I probably would swap them out on a day when I was bored out of my mine.

Rowan