CF/CCS, can it be made simpler?

[tricomp]

Hi,

I'm about to finish a 2C34-PP stereo amp. project.

To pull some usefull Watts from the 2C34 dual-triode I has to be allowed to pull grid-current, about 10mA, hence the CF-drive.
Having so far only put the amp on a scope and a tone-generator and measuring output-power, I have to say I'm pleased with the measured 12W/8-Ohm and 16Watt/4-Ohm. As I don't have a distortion analyzer I used the scope to determine the onset of distortion; not the most precise way, I admit.
When the amp's connected to my stereo and a pair of B&W 685 Speakers (88dB spl/1m, 8 Ohm.) it performs very much as hoped and expected. Rich and controlled bass (NFB applied) clean highs and nice vocals.

Now, my question is if I can safely simplify the CCS-circuit? The Cathode Follower is supplied from a Constant Current Source as per schematic (one channel shown, CF, CCS and output only).
I simply want to reduce the use of 4x6.2Vz and 4x1N4148 in each reference-string for the CCS to one of each. Then to connect each of the four BC549-bases to the 1N4148 Anode. Also could the MJE340 bases be connected to the common point between just one pair of 31K/3K3 resistors? Possibly the current in the reference chain 31K/3K3/1N4148/6.2Vz should be higher if transistors are connected as suggested above?
What you can't see from the schematic is that the -82V supply for the CCS is stabilized and comming from a TL783.

As the amp. is already built and pretty much finished I don't want to rip it apart to experiment needlessly. That's the reason for my question as I know someone here will definately have the right ansver

If anyone's interested I'll of course upload the rest of the schematic and pictures.

rgds and thanks,

/tri-comp

[wrenchone]

Why not just use a DN2540N5 depletion mode fet (Supertex)? THat's the approach I use with all my CF circuits. The current source is then reduced to the fet and two resistors. The output impedance is also pretty high without the need to cascode.

[Fenris]

And cascoding with another device is only nominally harder. One extra device, LM317 or LM134 (depending on current) and one resistor. Probably less than half a euro in cost. I cranked out a lifetime supply of PCBs for these (90 PCBs for about $40), so they're very easy to implement.

[Wavebourn]

It is already simple enough, what else do you want?
If to simplify, I would avoid stabilizing of a negative voltage if CCS is stable enough (as drawn on your picture)

[aardvarkash10]

Originally Posted by tricomp

Hi,

I'm about to finish a 2C34-PP stereo amp. project.


If anyone's interested I'll of course upload the rest of the schematic and pictures.

rgds and thanks,

/tri-comp

Yes please on the complete schematics!

Thanks

[tricomp]

DN2540N5 depletion mode fet
The reason for not using that is I'm not familiar with it ...until now.

LM317 / LM134
Use of these chips with an I-O potential of more than 100V. I don't think so!
The 2C34 g1-to-g1 swing is around 150Vpp

Simple enough
You're right, it isn't all that complicated AND it's already put together and it's working. I was just wondering and possibly I'll make another amp just like this as I already have the laser-cut and bended chassis for it. If I do I'll want to implement any improvements/simplifications, if any.

rgds,

/tri-comp

[Fenris]

Originally Posted by tricomp

LM317 / LM134
Use of these chips with an I-O potential of more than 100V. I don't think so!
The 2C34 g1-to-g1 swing is around 150Vpp
/tri-comp

http://www.audioxpress.com/magsdirx/...a/jung2779.pdf

Fig. 13C

The LM317 is cascoded using a DN2540N5 or IXCP10M45S, giving it the capability to handle 400+ volts.

[tricomp]

The pre-amp/concertina circuit.

The power-supply circuit isn't quite ready for publication as I haven't yet decided if I'm going to run the 6AN8A fil's on DC.
One channel is fully finished and I'm transfering the other pre-amp/concertina from a rats-nest to final point-to-point assembly.
After that and installing the bottom-cover I'll know about hum-level and need for DC-filament.

rgds,

/tri-comp

[tricomp]

Originally Posted by Fenris

http://www.audioxpress.com/magsdirx/...a/jung2779.pdf

Fig. 13C

The LM317 is cascoded using a DN2540N5 or IXCP10M45S, giving it the capability to handle 400+ volts.

I've been trying to get my hands on IXCP10M45S but it's simply not available from any of my suppliers including RS-Components.
I would REALLY love to work with that!

/tri-comp

[tricomp]

Originally Posted by Wavebourn

If to simplify, I would avoid stabilizing of a negative voltage if CCS is stable enough (as drawn on your picture)

That is something I've considered already as I've come across this suggestion in an old (50'ies I think) description of a cathode-driven PP-stage with a pentode-CCS. It was specifically stated that the stability of the negative supply isn't all that important.
I can easily remove the chip and see...
Not tonight, however but more likely over the week-end.

rgds,

/tri-comp