Hamptone JFET - mic preamp with parts on hand

[L´Andratté]

Oh I just built the ad797 from Samuel Groner´s site (I used 1:1.73 Sowters which I think are the optimum ratio) and I´m blown away by it. Nice.

 

[user 37518]

Oh I just built the ad797 from Samuel Groner´s site (I used 1:1.73 Sowters which I think are the optimum ratio) and I´m blown away by it. Nice.

Do you mean this one https://www.nanovolt.ch/resources/microphone_preamplifiers/pdf/G_r1.pdf? its a very clever design, he uses the 797s as the first stage of an instrumentation amplifier and by ommiting the differential stage you get a differential balanced output, very simple yet effective.

 

[L´Andratté]

Exactly that!
Recommended!
(I think all of the circuits from this site are serious stuff...)

 

[Scodiddly]

The breadboarding has begun.

I really did have all the parts for the basic design, even some rather odd-valued resistors that go with the voltage regulator.  Can't believe I never got around to building the damn thing.

 

[user 37518]

The breadboarding has begun.

I really did have all the parts for the basic design, even some rather odd-valued resistors that go with the voltage regulator.  Can't believe I never got around to building the damn thing.

Pics or it didn't happen...

 

[L´Andratté]

The breadboarding has begun.

I really did have all the parts for the basic design, even some rather odd-valued resistors that go with the voltage regulator.  Can't believe I never got around to building the damn thing.

The Hamptone?
There´s a lot of good advice around from PRR and others about that circuit...

If you are going for "colour", also have a look at Ricardo´s two transistor circuit I linked to on page 1, it sounds nice and is really simple and noise-optimized for 1:5 to 1:7 transformers
Have fun!

 

[Scodiddly]

Pics or it didn't happen...

Got it passing audio, but my DC voltages and obviously therefore my bias stuff is all out of whack. Which shouldn't be a surprise to anybody, really.

 

[user 37518]

Got it passing audio, but my DC voltages and obviously therefore my bias stuff is all out of whack. Which shouldn't be a surprise to anybody, really.

Its funny to see hi-end parts such as the cinemag trannie in such a low-end setup. Keep us posted!

 

[L´Andratté]

FWIW
Since every jfet needs different biasing, it´s usually a good idea to use a trimmer for
source resistor and adjust for linearity- people also use stock resistor values and go through a lot of jfets to find a good one, but idk it seems wasteful since jfets are not ubiquitous anymore...

When I once built this I tried it without the dubiously biased current source of the output
transistor replaced it with a 2k2 like some of the forum titans recommended and heard no difference (with bridging load like 5k)

 

[Scodiddly]

Of course now I'm in the position of wanting a scope at home.  My work one lives in my travel tools, and I'd rather leave it there and do something clever instead.  Probably find some generic audio interface that will work with Linux and do a soft scope.

Or just bring my work one home for a couple days. 

 

[Scodiddly]

Two important steps.

1) Testing FETs.  All of the remaining nine or so produced higher voltage at the drain than recommended, though there was at least a reasonable group of 4-5 once I adjusted the supply resistor upward.

Changing the lower bipolar to a 2.2k resistor helped simplify the circuit while I worked on the FET bias.

2) Changing supply resistors.  V+ to FET drain ended up being a fair bit higher than the schematic, as did the V+ to base resistor on the lower bipolar. However, I now have less heat from that bipolar and overall better sound.

Time to perfboard two of them to create the mic preamp topology.

 

[Scodiddly]

Wow, this thing has some serious gain.  And directly drives a pair of headphones on the output.

 

[user 37518]

Wow, this thing has some serious gain.  And directly drives a pair of headphones on the output.

Do you have any instrument to measure its distortion?

 

[Scodiddly]

Do you have any instrument to measure its distortion?

Yes, SMAART.  You can generate a sine wave and look at the harmonics and (if I can remember the key combination) have it calculate THD. 

Already used it, the frequency response is pretty flat and yes there's some distortion present.  Haven't brought my scope home yet to look at the waveform.

 

[Scodiddly]

Changed the thread title to reflect the chosen project.

Over the weekend I committed two channels (total four "modules") to perfboard.  Just need to package everything together now, and (parts permitting) add a phantom power supply at some point.

I've got some old Yamaha PM1000 switches lying around, rectangular versions of the old "eyeball" pushbutton switches.  Wonder if I can make those fit into the box for pad, polarity, etc.

 

[Gus]

something that might help
schematics posted again later in the thread
https://groupdiy.com/index.php?topic=70837.msg903106#msg903106

 

[user 37518]

Yes, SMAART.  You can generate a sine wave and look at the harmonics and (if I can remember the key combination) have it calculate THD. 

Already used it, the frequency response is pretty flat and yes there's some distortion present.  Haven't brought my scope home yet to look at the waveform.

What are the THD and THD+N specs of the amp? you can also try and download ARTA which is free and can measure THD.

You should measure THD at different frequencies, usually the most revealing is measuring THD at 20KHz.

 

[Matador]

I wonder how much more adaptable the circuit would be to different JFET's by substituting the 1.8K source resistor with a CCS set to 0.8mA (to drop 12V across the 15K drain resistor), using a BJT with a diode on the base (like what is done on the MP1, at the bottom of this image):

 

[Scodiddly]

I wonder how much more adaptable the circuit would be to different JFET's by substituting the 1.8K source resistor with a CCS set to 0.8mA (to drop 12V across the 15K drain resistor), using a BJT with a diode on the base (like what is done on the MP1, at the bottom of this image):

Forgive my ignorance, but where's the output?

 

[Scodiddly]

Finally done, or done enough to put on the lid and play with it for a few days.  Successfully built with only parts on-hand, which only really shows up with a couple of mismatched XLR jacks.