Sorting 2N3819s for KM84 circuit duty

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

tk@halmi

Well-known member
Joined
Jun 3, 2004
Messages
999
Location
Oregon, USA
Hi,

I am trying convert an ST51 to a different circuit and I am not having luck using Fairchild parts.
Out of 200 units, not one of the bias to the recommended Vs=10V and Vd=.95V.
I even got the two 100 batches about a year apart.

Many FETs hover around 10V on the source using a 10K source resistor, but then the drain will sit between 2 and 2.5 volts. I still get about x45 gain with no load so I guess they are usable in the circuit.

Has anyone had better luck with Central Semi or Vishay parts?
They tend to cost a good bit more though.

Thanks,
Tamas
 
sorry i can't offer any help, but i am interested in this circuit too. what are you using for a transformer?
 
Right now I have a CM-2480 to try. Gus recommends the LL1636 which may be more even at low frequencies due to the amorphous core.
 
IIRC .95V is a bias point that Gus found was optimal and he selected quite a number of FETs until he found one that gave .95V. But it's not a bias point that many 2N3819 will give. 1.5V is more likely. You can play around with the various transistors including the two drain resistors to get closer to your target voltages and gain. If you take a look at the various Neumann single FET circuits (U87 for instance) you'll find that they used different drain resistor values for what essentially is a very similar circuit, so the whole thing is not written in stone. You can also try a Schoeps-type bias trim pot to "force" your FET into the correct voltages.

The KM84 circuit still works at source voltages around 2V, but may not give optimal performance. Perhaps Gus can say more about that.

 
Thanks Rossi. I have already ordered trim pots to replace the source resistor and the 10K filter resistor.
I think the concern is Vgs versus Vp.
 
Interesting topic

is there a simple bench style test bed schematic for plugging in these 2N3819's
as part of the selection process ?
 
I ended up using a Fairchild J305 with about 1V on the source and 10.5 volt on the drain and a 3.92K source resistor. I assembled the mic using the original capsule and the CM-2480 because my LL1636 purchase fell through (see hazardharke conjob details in the Black Market).

As it is the mic sounds very clean. It is just a little crunchy on the top, noticable on hi-hats, but I think that is an attribute of the cheap ST51 capsule. The mic output level is so very hot that we needed very little gain on everything tried (voice, acoustic guitar, hi-hats, saxophone). The 10.5:1 output ratio does not seem to be an impediment at all. However, I am not using the 4 to 5pF cap between the gate and drain. I have some micas with that value and I will try that next.

I am also curious about using a Jensen JT-DB-E on the output, and thinking of converting an SP C1 using this circuit and the Jensen transformer.

 
The first reason I used 3.9K was because I noted the PDF "MICROPHONES" km8x circuit fragment had the source resistor value marked 3.9K.  I then asked myself why would that be marked 3.9K?  You can test things if you buy 1000 2n3819 fets. 

 
tk@halmi said:
As it is the mic sounds very clean. It is just a little crunchy on the top, noticable on hi-hats, but I think that is an attribute of the cheap ST51 capsule. The mic output level is so very hot that we needed very little gain on everything tried (voice, acoustic guitar, hi-hats, saxophone). The 10.5:1 output ratio does not seem to be an impediment at all. However, I am not using the 4 to 5pF cap between the gate and drain. I have some micas with that value and I will try that next.

Oh, I see, you omitted the small feedback cap (which takes off about 5-10 dB). In my experience, the circuit sounds better with that cap. You may also have some bass roll off without that cap. That crunchiness, too, is most likely a result of not having a small feedback cap at the FET.
 

Latest posts

Back
Top