another jfet gain stage to try in a DIY preamp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Gus

Well-known member
Joined
Jun 3, 2004
Messages
5,249
Location
n
Something different than another preamp gain stage.

Note the quasi constant current resistors in both stages.

It uses 1/2 of the available power supplies for the gain, - side is used for the quasi constant current.

You should bias up without any trimming needed.

R8 value sets the gain. EDIT Lower resistance gives more gain, max gain when shorted. Try a reverse log potentiometer 20k to 25k

A circuit like this can be found in a solid state Fender Harvard amp.

You can use a Darlington in the 2nd stage or a high hfe NPN or...
You can scale up the current in the 2nd stage.

DIjfetpreamp.png
 
I was thinking more in one of the small racks that people build modules for that have +- supplies. This also has a gain control that lets you set each stage gain, not leaving the gain stage on full and using a volume control after it.

You can use a number of JFETs 2N5457, 2SK30A, 2N3918 etc. and it should bias up. The source to ground voltage will change a little. Think Vgs at the drain current that the circuit stabilizes at. So as long as the drain current is not more than IDSS the source might move a few volts referenced to ground.
Lets make up some Vgs numbers
-16VDC(plus the Vgs at ID) via a 27k to ground.  No Vgs drop added 16VDC/27k= .59mA
(16VDC +1Vgs)/27k=.63mA
(16VDC +3Vgs)/27k =.70mA
.70mA-.63mA=.07mA
.07mA x 12k= .84VDC change


You don't always need a active CC device if you have enough extra voltage you can get close enough with a resistor.

The gain stage is set up to work more in the + to ground supply range and the clipping should be interesting.

You can increase the output stage current, as shown it is about 5mA.

I first noticed this bias method years ago in the groove tube amp book with lots of schematics IIRC early 90's? used in a Fender Harvard schematic JFET preamp section. I am looking for the schematic on the web.
 
It sims inverted close to gain of one with R8 at 20K as John posted.

Make R8 a 20k or 25K potentiometer and you have a gain control.
 
What does quasi-constant current mean? R3 current is roughly constant (in the flat part of the logarithmic large signal TF), but those other ones seem pretty 'naked' to me?

Btw. was thinking maybe you can add a similar RC as the gain control across R1 and convert it to a distortion control (ie. control the percentage current swing of the FET without changing signal gain). Less useful to be sure, but the point of this circuit is color, no?
 
R2 and R4 go to -16VDC
So the change in current is less because the - supply is being used for setting the stages current R2 for the source, drain current and R4 for the emitter, collector current.
Because of the use of the -supply in this manner the currents in the stages stays a little more constant not as good as constant current devices but should be good enough for this circuit.

The nice thing is the jfet should not need selection. One can change the R1 and R2 values for the jfet used make sure there is enough drive for the follower and you don't add too much noise from the drain resistor value before the base.

The gain stage is in the +16VDC side. Input resistor value is limited by PCB leakage, think piezo.  Useful output is limited by the + supply (output pair has one Vbe drop not two of a Darlington)

I am not a fan of the Hamptone design. People seem to like it and I have noticed people have racks with +- supplies. If I wanted clean I would use opamps. 

I am surprised there are not more posts in this thread.

CJ two threads for more effect voltage range.
https://www.diystompboxes.com/smfforum/index.php?topic=103492.0
https://www.diystompboxes.com/smfforum/index.php?topic=93817.0
 
Gus said:
I am surprised there are not more posts in this thread.
For a discrete gain stage without gain?  ;D

I've used several discrete JFETs over the years, but never to intentionally make distortion.  (If I wanted more distortion I wouldn't have selected for matched Vgs in one of my old phono preamps).

JR

 
Nothing?

I post this nice gem of a circuit fragment and nothing?

people have 500 series and people seem to like the Haptone jfet circuit

Here is a better gain stage IMO

 
So I have an O16 ounces for grid input Tube.  Had about 20Db gain.  Put R8 variable 25k and I have a 40 dB mic pre with some dirt in it.  The transformer Is plus 8 headroom so I need to pad as well? Haven’t tried the Hampton this is pretty simple I’m prolly missing coupling requirements
 
This is a nice & remarkable coincidence, I was looking around for the thread to the 'Fender Harvard topology' (the discrete JFET version, as Gus has addressed before as well here & elsewhere), and by magic, there's this fresh thread! :)

Nice, should bve a trigger to finally test these.
 
Gus said:
Found a link for the Fender Harvard in the technical section.

It has an interesting solid state preamp. Note the drain resistor value and the RP of 12AX7

I recall we had a chat about this before, possibly elsewhere, yep, makes one curious how these two circuits (one in JFET style, the other tubed) compare in sound.
 
repost of the schematic
 

Attachments

  • 500rackpreamp2.png
    500rackpreamp2.png
    155.7 KB · Views: 112
> I post this nice gem of a circuit fragment and nothing?

Many folks here would not know a "gem" if it fell in their lap, and can't handle a "fragment".

But some old piece of utility junk from the 1940s, that's kewl, no matter how badly documented.
 
PRR said:
Many folks here would not know a "gem" if it fell in their lap, and can't handle a "fragment". 

Haha :D    Funny, but no doubt  true. 
Well, I appreciate the repost Gus, I missed it the 1st time round.
Cheers.
 
PRR said:
> I post this nice gem of a circuit fragment and nothing?

Many folks here would not know a "gem" if it fell in their lap, and can't handle a "fragment".

But some old piece of utility junk from the 1940s, that's kewl, no matter how badly documented.
Hey I'm old junk from the 40s....

JR
 

Latest posts

Back
Top