• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

I m confuse ! expert help

jeapel

Member
2003-09-16 7:04 pm
quebec
hi sorry for my english

I m confuse about my recent test on my tube driver:

www.loginnovation.com/jeapel

1-6sn7-1 gain=8
2-6sn7-2 gain=18
3-6as7 follower gain=.5

total gain=72

vout max.=26 volt peak to peak maybe 10 volt p-t-p for
normal volume

for 6 months now gain=35 feedback=6db and the sound
is good

my tests:

1-get up the feedback(stable) to 9db gain=25 the sound is good but i lose something and i can t say what ?

2-remove the feedback and suprise my first impression is to get
sound one step better it s no sense to me

my questions:

1- is it possible and normal to get better sound without feedback
or i get ears problems :)

2-is it possible to destroy good tube harmonics with feedback
or the harmonics are always there whatever the feedback

thanks for help
 
It's not entirely clear what you've got there, which unit on ur pages... but in general the addition of feedback does change both the frequency response and the ratios of harmonics as well as their amplitudes. So, the short answer is that the sound *does* change.

If you have a transformer in the loop - like an output xfmr then the sound may change even more than with just tubes since the feedback tries to overcome the "natural" response of the xfmr, but it can't really, so the result is what I call as "forced response" where the frequency response is flatter for steady state tests, but the transient response may not be as good (overshoot, etc.) and the response into real world reactive loads may also be quite different than the zero feedback condition. The DF also changes with more feedback.

_-_-bear

:Pawprint:
 
ahhh... I guess this is ur OTL??

Suggest paralleling up a bunch of those 6AS7s unless ur driving some horns or some other very sensitive speakers?

But, yes increasing the feedback will still change the spectrum of the distortion products - how much is open to measurement and discussion.

You could think of it this way: a tube or series of tubes (as in an amp) has a transfer curve which is flat only in the middle, and then not perfectly flat, curving off at different rates on either extreme (the "off" end and the "on" end). The application of feedback tries to straighten this curve until there is simply no place to go reaching to the ends of the curve. The flattened curve is different than the unflattened curve. Which is better may actually depend on the specific application and use (for example: is it running out of the linear portion - and how much?)

Consider one of the bridged OTL outputs too... :D
They can be DC coupled with a little effort... and may be more linear even when AC coupled. But then they are not SE, which some may prefer - in which case linearize each stage independently and get rid of the loop feedback entirely? :rolleyes: (thinking out loud...)

_-_-bear :Pawprint:
 

jeapel

Member
2003-09-16 7:04 pm
quebec
thanks bear very interesting explications

i think i get a true tube sound now maybe

a 2 position switch 0db and 6db feedback

would be a good idea to compare the sound

maybe it s not clear on web site but my OTL

is my tube driver for my source follower mosfet

in this way i got good overall damping factor

and it s more linear that transfo

bye ( have you a web site)
 

jeapel

Member
2003-09-16 7:04 pm
quebec
hi

I just visit your web site wow! great amp.

"dc coupling to tube" you r reading my mind all my
cap are good quality solen polypropylene except
the output cap but i don t find the way to dc coupling
to mosfet

I m thinking about that maybe the no feedback is
possible in my case because i need just 10 to 20
volt peak to peak with 300 and 150volt supply
contrary to 200 to 300 volt pp swing for a normal amp
with tranfo output and need feedback for low distortion
if i m right it s a good point for hybrid design
 
Cool hybrid amp Jeapel.
222µF will pass all the current the MOSFET ever needs, so no need to DC couple unless you want to skip the capacitor for other reasons.
I tend to prefer amps and preamps with no loop feedback. I've tried several different circuits with and without loop NFB and I keep coming back to the amps without. Matter of taste I guess. Perhaps the amps seem more dynamic and alive because the gain goes up as the NFB is reduced, but I try my best to compare at equal spl.
If you do not want to use feedback, the stock circuit may have a bit too much gain. In that case I would remove the first stage and the feedback loop and bypass the 6SN7's cathode resistor. That should give you enough gain if you use CD for source.
If you want to use all the tube sections, you can always parallel the 6SN7 triodes and their cathode resistors and increase CCS to 10mA. That will let the VAS drive the 6AS7 a little better.
This is at least what I'd do before bothering with DC coupling the MOSFET.
 
hi SemperFi

Cool hybrid amp Jeapel.
.......thanks i don t count hours it s a fun hobby

222µF will pass all the current the MOSFET ever needs, so no need to DC couple unless you want to skip the capacitor for other reasons.
........maybe just a audiograde cap would be a good step

I tend to prefer amps and preamps with no loop feedback. I've tried several different circuits with and without loop NFB and I keep coming back to the amps without. Matter of taste I guess. Perhaps the amps seem more dynamic and alive because the gain goes up as the NFB is reduced, but I try my best to compare at equal spl.
........ok you tried several different circuit and you find
better sound like me , more dynamic yes and maybe
stereo image

If you do not want to use feedback, the stock circuit may have a bit too much gain. In that case I would remove the first stage and the feedback loop and bypass the 6SN7's cathode resistor. That should give you enough gain if you use CD for source.
If you want to use all the tube sections, you can always parallel the 6SN7 triodes and their cathode resistors and increase CCS to 10mA. That will let the VAS drive the 6AS7 a little better.
.........thanks very good advise but before change design i
will put a switch with 0db and 6db position and will
do long listen test
bye can i see your amp on internet photo gallery or ?