LSK pre - BAF 2013

At the BAF2013 Mr. Pass presented a nice JFET preamp. Today I cobbled up a quick prototype of one channel to test its behaviour and it's very nice. I have a BL rated Toshiba JFETs so some values had to be changed and some parts added (P1 to set the DC offset at the output and harmonic content, caps in cascode voltage divider to improve HF extension a bit). Serious build with buffer added at the output will follow. For the time being here is the schematic and the PCB layout of the preamp (copper side view, 45x50mm) :

EDIT (Oct. 27, 2013): 'final' schematics and pic of built preamp are in the post #66
 

Attachments

  • LSKpre-sch.gif
    LSKpre-sch.gif
    6.7 KB · Views: 11,029
  • LSK-BAF2013-PCB.GIF
    LSK-BAF2013-PCB.GIF
    33.4 KB · Views: 9,683
Last edited:
Member
Joined 2006
Paid Member
...i believe the original has LSK and LSJ of 15mA ( C grade ) Idss JFETS....this one has 2SJ and 2SK of BL grade so in principle a little lower...this should help to calculate voltage drop on R19 and R25 resistors and the resulting absolute max voltage swing.
 
Last edited:
Doesn't this design look like 1/2 a Blowtorch? Just sayin... And if you were to go balanced?
Or, even the Vendetta MC Pre-Amp?
Or, is it maybe the BA-4 front-end? Although, then, as already mentioned, it would need a little tweaking to get to the fully useable output voltage swing of the typical BA output stage (Or F4).
But I like it!
 
Official Court Jester
Joined 2003
Paid Member
output voltage swing is determined by cascode voltage potential

so , roughly (asymptotic to ) +/-18V ........ meaning 36Vpp

so , not exactly blowing F4 from pants territory , but almost

cascoding da thing everywhere and rising PSU voltage accordingly.... will bring ya (as FE) in PA realm .... (think about including amplified diode in output drain vertical , as inputing part of OS ) :devilr:

and , yes ........ Soroboruo ......... :clown:
 
Last edited:
I see the 45v rails, but I don't really know how to calculate output voltage swing. :p

Output swing is happening in realm of voltage difference between the PS rail voltage (45V) and cascode voltage (18V). How much it will be depends on current available to output pair (supplied by 1K/3W resistors to both pairs) and Id of JFETs on one side and the loading resistor (1k5 - the one between output pair drains and the ground) on the other side.
The output swing will be the product of these 2 values, so with the lower value of loading resistor comes the lower gain and the clipping comes sooner (less available swing). For example, with 4k7 loading resistor I got 19V_peak before clipping.
Basically, the thing here is about current swing that gets converted to voltage on the loading resistor (transimpedance amp, I/V converter, that kind of stuff...)
 
Last edited:
At the BAF2013 Mr. Pass presented a nice JFET preamp. Today I cobbled up a quick prototype of one channel to test its behaviour and it's very nice. I have a BL rated Toshiba JFETs so some values had to be changed and some parts added (P1 to set the DC offset at the output and harmonic content, caps in cascode voltage divider to improve HF extension a bit). Serious build with buffer added at the output will follow. For the time being here is the schematic and the PCB layout of the preamp (copper side view, 45x50mm) :

Can't find gate stoppers R21 and R24 on your PCB.
 
Can't find gate stoppers R21 and R24 on your PCB.

Looks like I forgot them...
There are few options to get out of predicament:
to interrupt the gate trace and solder them on the bottom side of the PCB,
or to shorten the gate pin of the JFET and solder the resistor directly to it. Third option is to use BJTs (BC550c/BC560c) as the output pair (works great) so you don't need the gate stoppers.
The last option would be to draw the PCB again with R21,R24 on it... ;)