UGS adventures

You're both right, gentlemen.

Will try to remove Rout on one set of modules, and use my volume pot
to as both Rout and volume, as Phillipe said... And will try on my old
perf board prototype with 1K volume pots, in the same configuration
(no Rout on board).

Arghhh, wires everywhere...
 
Some news...

I spent some time testing different configurations :
- Remove Rout from the modules and let the 10K output volume pot
play the Rout role
- Same thing, but with a quad 1K pot - So Rout is back to 1K
- 10K volume pot at input, with a 10K Rout
- 10K volume pot at input, with a 1K Rout

I have taken care before each listening to adjust output levels to the
same value with an AC voltmeter, to avoid to be fooled by false impressions
of better of worse dynamics...

Well... Quite difficult to say, but I didn't feel differences... :xeye: Some times I
was inclined towards one version, then another one, back to the first,
no, it is the third, finally no, the second, and so on... Endless game...
As a whole, all the versions sound the same to me, but it doesn't mean
you should feel the same... So test and choose the one your prefer ;)

Cheers
 
Hi Cheff,

Thank you for reporting these results. They are very helpful. I appreciate your careful methodology and thoroughness.

I am still working on my version of the circuit. I recall that you had asked me about that earlier in the thread. At this point I have deleted the followers. I am still trying to find a balance of Rout and common mode feedback that produces an optimum result. Again, I appreciate your work in this area.

Cheers,
Graeme
 
Hi Cheff

I also have more gain with the amp 10x.
Last week we did the same test with a few listeners and the difference is very easy to hear.
So I don't understand your findings.
The difference in my amp are feedback and R out and a 50k pot in.
The only thing i'm going to change is the potmeter.

Rob
 
Rob Dingen said:
So I don't understand your findings.

Just assume I'm deaf :)

Seriously, I didn't mean the differences don't exist. Just that I can't
hear them on my system, which is different and surely inferior to yours.

And the most important thing is that it's the configuration YOU prefer
and feel comfortable with. So keep it ;)
I'm just glad you like this design, whatever it may differ on peculiar
points from one version to another.
And if others can test, and prefer a particular configuration, no problemo.

Nobody's wrong, nobody's right, everybody's different :)

Cheers,
 
Hi Cheff

I don’t assume your deaf maybe small ears :D
I also think you don’t have to bash your own system because I think you can hear the difference on every system.
Maybe you can tell something of your own system.
And where do you live maybe we can meet.
I soldered 10k in the output and during listening test put a parallel resister of 1,2k.
That’s quick and easy.
I can tell this difference is not a tiny difference when you went to the toilet and come back you still hear de difference so you can’t fool people with this.
In a few weeks my AX amp is ready and maybe that has something to do with it.
And I don’t want to talk about right or wrong that doesn’t exist in audio.


Rob
 
Control Board

The control board is also my design, but the final pcb was a bit
beyond my local shop possibilities, so I had it made by a pro shop,
which explains the different color
The µC is an Atmel AVR ATmega64. It's remote controlled (Philips
RC5 code), with a learn function. It is specifically designed to meet
my needs, but if needed, I will post the schematics, Gerber files and
software.


Hi Cheef !


Anyone can me send he schematics, Gerber Files and Software for the control board.

My e-mail: michbeh2000@yahoo.de

Thanks Michael.
 
Late - but hopefully not too late - I want to make my contributions to Cheff´s UGS.
This module looks really fine for preamps; but what´s about to use it as a basic module for power amps? Just like Pass Labs does it successfully with their UGS-modules!

My suggestions are:
Substitue all bipolars with MJE340/350 instead of ZTX450/550; it´s easy to fix these (rather old - does anyone know better ones?) devices isolated with srews on a U-profile heatsink.
Incorporate a variable voltage source Vbias between Q9 and Q11 (Q10 and Q12).
Cascode biasing resistors are 33k-15k-33k.
Increase Rout to 11k (or twice 22k on the top and bottom of Vbias)
With Rfb is 330k and Rin is 11k all this results into a total gain of approx. 28

I could not hear this variation in comparison with Cheff´s module, but simulation results are very promising.
With a supply voltage of +/- 70V all semiconductors dissipate approx. 4 watts - that means you have to put an extra heatsink onto the U-profile!

Regards

Hans
 
Hi Cheff,
it's quiet here....:mad: Nobody else has built this pre ?????:bigeyes: I am waiting for my "main board" to plug on the 2 UGS boards. Hope to receive it soon and to test it....;)
No more news for the control volume and the "fioritures"....:cool: As i can see and guess, most of the componentns are in SMD (CMS in French !:D ) and some DIYers will be disappointed quickly coze it seems it's hard to solder. Luckely i solder not often but sometimes SMD componants.
Pleaaaaaaase Cheff !!!!:D
Au moins j'aurai essayé même ici !
 
hm4nine said:
...
My suggestions are:
Substitue all bipolars with MJE340/350 instead of ZTX450/550; it´s easy to fix these (rather old - does anyone know better ones?) devices isolated with srews on a U-profile heatsink.
Incorporate a variable voltage source Vbias between Q9 and Q11 (Q10 and Q12).
...
Regards

Hans
Hi, Why not use MPSA42/MPSA92... and use a file to adjust thickness? :)