Line amp single to bipolar suply

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

zamproject

Well-known member
Joined
May 11, 2010
Messages
1,503
Hello

I have two of this preamp that I'll like tu use in a bipolar supply setup (for practical reason)
Initially it work on 24v, regulated to 22V, I'll like +/-12 or 15

I attach the schemo and the mod I think I have to do, I'll like double check /advise,
I basically remove the onboard regulation, DC bias section, and hook both AOP to V-, as output transistor.
I have a doubt at R22 (just before output) 0V or V- ?
Also not highlighted in the document, but what about R8 (Q3 E) 0V or V-

Best
Zam

 

Attachments

  • LineAmpMOD.pdf
    190.9 KB · Views: 20
I have many many doubts. Q3 lacks bias. Q2 does nothing (allows Q4 Q5 to blow-up). C5 at "n33", 33 nanoFd? Two grounds on T2 primary windings??
 
Hello PRR

This is a reverse engineering drawing so there is some mistake despite triple check ...
as unpractical component notations as it reflect what I physically spot in the PCB layout before drawing the schemo

Q3 bias...ok I have to hook R12/C12 to V- and not remove this part.
C5 is for sure 33p regarding the function, you spot a mistake I suppose
Q2 I'm nether sure... I check so many time, but this is (for the moment) what the PCB tell me  :-\
About trafo, plz don't consider this ghost  winding, I'm not sure about it, trafo have 6 pin + can, I pick 6 pin trafo in kicad
I suspect that one is screen one is can and maybe there is a centre tap grounded (trafo probably use in other design ?)

Best
Zam
 
Ok...miss at least one connection...Q2c to R7/C9 junction
 

Attachments

  • Q2.jpeg
    Q2.jpeg
    76.8 KB · Views: 15
zamproject said:
I have two of this preamp that I'll like tu use in a bipolar supply setup (for practical reason)
Initially it work on 24v, regulated to 22V, I'll like +/-12 or 15

One opportunity was using a module as is, connecting -V to GNDA and +V to +V, gnd from bipolar power supply leaving N.C.. That's possible just because you have an input and output transformer which provide galvanic isolation. Then you should just disconnect connection between GNDA and chassis.

Right now, you have to do some modifications, IMO:
Q3 should be biased to +0,65V or so using R19 and R12,
R8 should be connected to -V,
D1 and D2 are backwards, and it would be nice that they are connected to -V, then R3 becomes 5k or so,
C15 and C8 should be bipolar,
I would remove bootstrap capacitor and R17, R14 goes to +V, new R14' goes to -V from Q5 base,
C18 should be bipolar if you do not bias whole chain little bit to +, using R19 and R12
R22 goes to GNDA,
and if Q2 is Vbe multiplier, then D5 and D6 are backwards too.
You have shortened coils at input and output transformer.
I probably forgot something.




.
 
Hello

moamps said:
One opportunity was using a module as is, connecting -V to GNDA and +V to +V, gnd from bipolar power supply leaving N.C.. That's possible just because you have an input and output transformer which provide galvanic isolation. Then you should just disconnect connection between GNDA and chassis.

Yes that would be easy, especially as audio ref (GNDA) chassis ground and trafo screen have different trace/different connector pin and various jumper to separate them or not
But I plane to insert a bipolar circuit (opto comp) in place of passive input attenuator, before C7, so I cant let GNDA float between the two trafo

moamps said:
Right now, you have to do some modifications, IMO:
Q3 should be biased to +0,65V or so using R19 and R12,
that's more or less what it is now by measure relative to 1/2 V+
moamps said:
R8 should be connected to -V,
OK
moamps said:
D1 and D2 are backwards, and it would be nice that they are connected to -V, then R3 becomes 5k or so,
mistake, well spotted.
moamps said:
C15 and C8 should be bipolar,
OK
moamps said:
I would remove bootstrap capacitor and R17, R14 goes to +V, new R14' goes to -V from Q5 base,
C18 should be bipolar if you do not bias whole chain little bit to +, using R19 and R12
R22 goes to GNDA,
and if Q2 is Vbe multiplier, then D5 and D6 are backwards too.
OK
moamps said:
You have shortened coils at input and output transformer.
Explained in previous post

Thanks for this

Now there is more than I was thinking.
I'll see if it's not simpler to make the inserted pcb floating ...

Best
Zam
 
Ok... strategy change

I think it will be way simpler to split the circuit in two with a +/-12 psu:

everything before C7 dual +/-12 supply with 0V ref to ground (for the moment its only passive)
everything after C7 floating single supply 24V (between +12 and -12) connected to actual power input

at PCB there is wire jumper between 0V psu (wich will become -12) trafo screen and trafo can/chassis
and the few component before C7 will have GNDA separated from the actual schemo and stay GNDA (including the inserted active circuit)

I just don't know what to to with the output trafo screen, to the -12V or to the 0V/GND ? (mu metal can will be at GND)

Best
Zam
 
Back
Top