Hello everybody,

I'm back with an oscilloscope and a signal generator.
Now i'm able to do some 'real' testing.
I have worked the two last weeks with it and here are my 'conclusions'.

First, i put my reference amplifiers, namely Quad II to see how they
are.
I looks like good for me in the LF region, not much sag at 30hz.
I can also see some overshoot maybe 10% at 10khz.
I gently roll off over 20khz.

I didn't know how to set up the GNFB.

Thank's to Patrick Turner i know : GFB in db = 20 log (V1/V2). V1=Vout
with GNFB and V2=Vout without.
I adjust my GNFB to 15db.

It is heavily mototboating at 14hz.

I then changed the signal capacitors from 0.1mfd to 0.47mfd, as adviced
somewhere here and also by a colleague.
Miracle i now have very good stability.

I also included a 470pFd in the feedback loop to reduce the overshoot
at 10khz.

I tried and abandoned :
- stabilising networks with the output transformers (no improvement for
me) optionally adviced in the GE book (amplifiers from 5 to 1000w)
- low pass filter in the first stage to cut the very high frequencies
as adviced omptionally in the GEC book. Not an improvement for me

Now, it measures 'flat' from 20hz to 40khz and the square signal is
very clean at the output. Very small sag and small ringing and good
stability.

Maybe i have built something better than a quad II ? who knows...

I will continue my testings and keep you informed. Also with real
figures as i only gave you here impressions.

I must say that now i'm quite happy with the sound i can hear.

See you,
Luc D.

said:

Maybe i have built something better than a quad II ? who knows...


Perhaps measurably better.
The only criterium of "good" or "better" is that *you* like the sound.


I will continue my testings and keep you informed. Also with real
figures as i only gave you here impressions.

I must say that now i'm quite happy with the sound i can hear.


Good news! Enjoy your amp!

--

"Audio as a serious hobby is going down the tubes."
- Howard Ferstler, 25/4/2005

a écrit dans le message de news:
...
Hello everybody,

I'm back with an oscilloscope and a signal generator.
Now i'm able to do some 'real' testing.
I have worked the two last weeks with it and here are my 'conclusions'.

First, i put my reference amplifiers, namely Quad II to see how they
are.
I looks like good for me in the LF region, not much sag at 30hz.
I can also see some overshoot maybe 10% at 10khz.
I gently roll off over 20khz.

I didn't know how to set up the GNFB.

Thank's to Patrick Turner i know : GFB in db = 20 log (V1/V2). V1=Vout
with GNFB and V2=Vout without.
I adjust my GNFB to 15db.

It is heavily mototboating at 14hz.

I then changed the signal capacitors from 0.1mfd to 0.47mfd, as adviced
somewhere here and also by a colleague.
Miracle i now have very good stability.

I also included a 470pFd in the feedback loop to reduce the overshoot
at 10khz.

I tried and abandoned :
- stabilising networks with the output transformers (no improvement for
me) optionally adviced in the GE book (amplifiers from 5 to 1000w)
- low pass filter in the first stage to cut the very high frequencies
as adviced omptionally in the GEC book. Not an improvement for me

Now, it measures 'flat' from 20hz to 40khz and the square signal is
very clean at the output. Very small sag and small ringing and good
stability.

Maybe i have built something better than a quad II ? who knows...

I will continue my testings and keep you informed. Also with real
figures as i only gave you here impressions.

I must say that now i'm quite happy with the sound i can hear.

See you,
Luc D.

Great, so you adapted feed back to your trannies.
Today trannies are probably better that thoose available when the LeakII was
designed !

Yves.

"Yves" wrote

Great, so you adapted feed back to your trannies.
Today trannies are probably better that thoose available when the
LeakII was
designed !

Can be, but commonly aren't. Many available valve amps now have OPTs
that look too small to me. Cheapness, as ever.

cheers, Ian

"Ian Iveson" a écrit dans le message de
news: ...
"Yves" wrote

Great, so you adapted feed back to your trannies.
Today trannies are probably better that thoose available when the
LeakII was
designed !

Can be, but commonly aren't. Many available valve amps now have OPTs
that look too small to me. Cheapness, as ever.

cheers, Ian

Wind them yourself !
Yves.

I plan to do that one day...

If you could see the one i have bought from Ian Wuesten you wouldn't
say that they are too small.
In fact they are at leaste 2 times bigger than the quad II opt.
Luc.

wrote:

Hello everybody,

I'm back with an oscilloscope and a signal generator.
Now i'm able to do some 'real' testing.
I have worked the two last weeks with it and here are my 'conclusions'.

First, i put my reference amplifiers, namely Quad II to see how they
are.
I looks like good for me in the LF region, not much sag at 30hz.
I can also see some overshoot maybe 10% at 10khz.
I gently roll off over 20khz.

I didn't know how to set up the GNFB.

Thank's to Patrick Turner i know : GFB in db = 20 log (V1/V2). V1=Vout
with GNFB and V2=Vout without.
I adjust my GNFB to 15db.

It is heavily mototboating at 14hz.

I then changed the signal capacitors from 0.1mfd to 0.47mfd, as adviced
somewhere here and also by a colleague.
Miracle i now have very good stability.

It still may not be as good as you would wish.
A gain step network will make it OK, even with 0.1 uF caps,
and stop the motor boating, although large rail caps help.



I also included a 470pFd in the feedback loop to reduce the overshoot
at 10khz.

I tried and abandoned :
- stabilising networks with the output transformers (no improvement for
me) optionally adviced in the GE book (amplifiers from 5 to 1000w)
- low pass filter in the first stage to cut the very high frequencies
as adviced omptionally in the GEC book. Not an improvement for me

It's tricky to get right, sometimes you need to set the
HF step network with a pot and tuning gang for the least overshot
at low levels into 0.22 uF load.
The the 470 pF might be able to be reduced a bit.




Now, it measures 'flat' from 20hz to 40khz and the square signal is
very clean at the output. Very small sag and small ringing and good
stability.

What happens with a 0.22 uF load?



Maybe i have built something better than a quad II ? who knows...

I will continue my testings and keep you informed. Also with real
figures as i only gave you here impressions.

I must say that now i'm quite happy with the sound i can hear.

keep soldrin'

Patrick Turner.



See you,
Luc D.