jFET mirrors?

Help Support GroupDIY:

hildy

Member
Joined
Sep 22, 2011
Messages
13
Hey all. Attached is a schematic from the line input section of a mixer from a "big-name" brand of audio gear. Was wondering if anyone could share some info on what's up with the jfet pair (Q4208) being fed by the bipolar current mirror setup (Q4204), as well as the bipolar pair (Q4206) in this design.

The audio signal is coming into pin 4 on the 2SK3320. And why couple the op amp inputs to the power rail through the 1k resistors? bias for the BJTs on the 4580s?

interested in your analyses.

-hild
 

Attachments

  • Screen Shot 2022-01-27 at 1.46.18 PM.png
    Screen Shot 2022-01-27 at 1.46.18 PM.png
    375.1 KB · Views: 102

JohnRoberts

Well-known member
Staff member
GDIY Supporter
Moderator
Joined
Nov 30, 2006
Messages
22,933
Location
Hickory, MS
the top of the schematic is cut off...

q4206 is cascode topology for high bandwidth.
q4208 is the actual differential long tail pair.

q4204 has half the circuit missing so hard to speculate.

JR
 

hildy

Member
Joined
Sep 22, 2011
Messages
13
the top of the schematic is cut off...

q4206 is cascode topology for high bandwidth.
q4208 is the actual differential long tail pair.

q4204 has half the circuit missing so hard to speculate.

JR
brilliant. Thank you. The missing top half is just the other side of the stereo pair, identical to the shown diagram. Assumption is that Q4204 is an active current source set up with 4201 as a mirror. (I'm a little out of my depth here; feel free to chortle.)
 

ccaudle

Well-known member
Joined
Mar 18, 2010
Messages
286
Location
Houston
Look up the term "active load," you see that type of cascode/current mirror design in discussion of op-amp circuits a lot. It has better gain than just using collector resistors. Seems unusual to have an extra gain stage in front of an op-amp like that for a line stage rather than mic amp.
 

abbey road d enfer

Well-known member
GDIY Supporter
Moderator
Joined
Jan 22, 2008
Messages
14,337
Location
Marcelland
Isnt'it a variation on the SSL "VHD" thingy? There was a recent discussion of it here, with the patent text and schemo.
Of course here sans Variable.
 

Bo Deadly

Well-known member
Joined
Dec 22, 2015
Messages
2,922
Location
New Jersey, USA
The NPNs are not a current mirror. The JFETs provide high Z but the NPNs provide low noise open loop gain. The op amp is just servo-ing the + input to match -, diffing to unity and providing a low-Z out. Not obvious what the advantage of being cascode is here.

At least that's my immediate guestimation. If I look at it again tomorrow morning I'll probably change my mind.
 

hildy

Member
Joined
Sep 22, 2011
Messages
13
The NPNs are not a current mirror. The JFETs provide high Z but the NPNs provide low noise open loop gain. The op amp is just servo-ing the + input to match -, diffing to unity and providing a low-Z out. Not obvious what the advantage of being cascode is here.

At least that's my immediate guestimation. If I look at it again tomorrow morning I'll probably change my mind.
(again, out of my depth here) - but after JohnRoberts enlightened us here on the cascode topology, my first thought was, "why exactly?". I suppose there is always a reason to take something as commonplace as a line input and make it more hi-tech or hi-fidelity, but at the end of the day, it seems like diminishing returns. But then again, it lets you write "new and improved" on the retail box. That and blue LEDs.

I was thinking the NPN pairs over on the left hand (Q4201, 4204) of the schematic were doing some sort of current-regulation magic.
 
Top