Perhaps most of us have heard/listened to high quality GNFB and OL amplifiers.
My takeaway from such listening is that OL amplifiers although perhaps having 'loose' FR, can sound 'friendlier' and more 'peaceful' than GNFB amplifiers.
To me this is one key....we have all heard amplifiers that are 'steely', 'wiry', 'on edge' etc.....ie the typical modern AV receiver.
How an amplifier stage handles it's own self generated excess noise imo is one subjective arbiter........OL amplifiers do not 'recycle' self noise.
Dan.
Both CAN be built to have no perceptible noise at the output. OL amplifiers are actually more susceptible to signal modulated noise, hi GNFB amps generally have the active devices deviate from their quiescent operating points far less except directly at the output. The best case are old tube amps made with carbon comp resistors which could get to the point where the noise sidebands are above accepted thresholds of audibility. Recycling of self noise is definitely a loose and ill-defined concept.
I'm sure it's easy to find a cheap AV amp with under designed power supplies and inadequate bias to ease thermal considerations. In my opinion conventional design of quality audio (I mean commercial) equipment is grossly under-represented here in these discussions. There are many hi-end pre-amps and amps full of 8-legs and piles of GNFB (even sold for stupid money to audiophiles), the principles involved simply don't bother with tweak nonsense.
Last edited:
That points to series stages excess noise....any dirt at the beginning gets chaotically and exponentially exaggerated.most definitely. And the closer you get to the source, the more realistic it sounds. fewer stages, steps, processes.
Get the replay system self excess noise way down and really old typically ear bleeding recordings like for example Robert Johnson/Jimmy Rodgers etc become great, really great.
Dan.
Such easily substantiated claims.
What is more unsubstancied in my saying that in your argumentless answer.?..
Or was it ironic.?.
For you old geezers, yes. Not for Bruno. The tube amp 'flight of fancy' had 60dB at 20k. The name nCore in the nCore amps comes from the fact that he wrapped n-th order feedback around it, including the output filter. Last time I looked n stood at 6.
Once you can let go of your emotional attachment as if audio is special, you realise that it's just a control problem and the sky becomes the limit.
Did you see the spec for the Mola Mola DAC? ' Distortion: unmeasurable (estimated -150dB)' . A DAC.
And especially for John C: see my sig line. RH was way ahead of us but nobody realised it, not even you ;-)
Jan
Last edited:
What is more unsubstancied in my saying that in your argumentless answer.?..
Or was it ironic.?.
Known by whom? Where is there even a context around which to have an argument (I prefer discussion)? This "fact" stands outside of amplifier topology or device technology? So if someone develops high current output devices with GHz ft's there is still this 35dB wall?
I m talking of audio amps, preferably VFAs, and of course of GNFB if you had noticed, because i m about sure that you did read "NFB" instead.
(Generally. There are examples of GMNFB amps, which even have temp compensated outputs. In some cases, thanks to 8 legs)
What is your signal source? How do know what the amplifier is doing in isolation from everything else?
Regarding typical modern AV receivers, most of them built to a price point and yeah, they sound like it.
I have a tube guitar amp that sounds great with my electric guitar. My hifi amp doesn't sound nearly so great for that source. Should I assume that excess noise in my hifi amp is the problem?
Fine, citation please where this 35dB max was derived. Preferably not someone's VFA thread with a given bag of parts, sighted listening, etc.
Yes, in this case I was referring to an output tuple that say runs at 200mA with no signal but still puts out 10A peak. At the input an LTP running at 1mA a side might only need to commute a few 10's of uA for full output while resistively loading the VAS to get "low open-loop GBW" will require it to do a lot more work.
IMNSHO opinion this is one of the most bogus "design" tricks going.
Last edited:
AFAIK that s the most possible without eating too much in stability margins, now if you are content with an amp that need a huge inductance to be stable you can use much more GNFB at this frequency, eventualy you can count on the cables to help a little, i guess that their characteristics could be actualy very critical for some randomely baked designs.
There is a tacit assumption here about device limitations/technology at a particular point in time or at someone's budget limit not any universal principle.
That s roughly 35 years that we are inside this particular point of time, certainly that for anything digital related nodes shrinks brought some progress but in the area of analog class AB/A amps components from 1980 are no worse than current ones, and sometimes were even better like Hitachi T03 lateral fets that are currently sold in TOP3 cases...
- Status
- Not open for further replies.