John, JCX:
It would be interesting to try to define, at least roughly, what does "wide open loop bandwidth" actually mean in terms of frequency.
Theoretically, in the broadest of terms, that could well be simply 20 kHz. With such an open loop bandwidth, the amplifier does not require NFB to operate within the accepted audible bandwidth.
My own personal and completely subjective limit is twice that, i.e. 40 kHz. Twice the minimum always seems preferred to me than just in the nick of time and/or limit.
If memory serves, Otala & Lohstroh managed to get it up to 100 kHz. Hence the really way out slew rate of 100 V/uS for an amp with a low power ouput, unheard of in those days. At the time, the original biggies, like SAE, declared 40 V/uS for their 300+W amps. However, taht said, I should also add that the same 40 V/uS was quoted for ALL their amps, from 50 to 300 W, which implies some sort of bandwidth limiting inside the cicruits, I see no other way how output voltages from 28.3 to 69.3V would all produce exactly the same resulting number.
From a purely practical standpoint, I have heard a few low slew rate amps, which barely satisfied the 0.5 V/uS per peak volt criteria, sound subjectively better than a few other amps with wild slew rates and about ten times the nominal speed. This taught me that slew rate is good, but is not the universal panacea many hope for. People will always tend to look for one broad sweeping method of solving everything, even if that never works as easily as that.
Obviously, when setting the upper limit of the OLB, like it or not, we are also setting the amount of NFB we will use, or if you like, have to work with. AFAIK, you can never have both, wide OL bandwidth means less gain, and that mean less NFB required.
While I am a proponent of wide OL bandwidth, I am NOT against NFB as such, I do not think of it as automatically bad - quite to the contrary. Of the few amps I have heard which boasted zero global NFB, I wouldn't run out and buy any of them. The one common point was that they all sounded just a bit out of focus, so to speak, as if not quite finished. This leads me to believe that NFB is a good thing indeed, provided it is used to iron out what is an already good amp even without it. As say when the amp has an OL full power THD of say 0.5%, 20-20.000 Hz into 4 Ohms and is quite stable.
But even this is a sort of a floating point. Some amps will be quite happy with 12 dB of global NFB (e.g. Harman/Kardon Citation 24), others will need 17 dB for the same effect (e.g. H/K 6550 integrated), and still others will need more. Point is, it's very hard to generalize something that is model dependent.
But this should not put us off from trying.
It would be interesting to try to define, at least roughly, what does "wide open loop bandwidth" actually mean in terms of frequency.
Theoretically, in the broadest of terms, that could well be simply 20 kHz. With such an open loop bandwidth, the amplifier does not require NFB to operate within the accepted audible bandwidth.
My own personal and completely subjective limit is twice that, i.e. 40 kHz. Twice the minimum always seems preferred to me than just in the nick of time and/or limit.
If memory serves, Otala & Lohstroh managed to get it up to 100 kHz. Hence the really way out slew rate of 100 V/uS for an amp with a low power ouput, unheard of in those days. At the time, the original biggies, like SAE, declared 40 V/uS for their 300+W amps. However, taht said, I should also add that the same 40 V/uS was quoted for ALL their amps, from 50 to 300 W, which implies some sort of bandwidth limiting inside the cicruits, I see no other way how output voltages from 28.3 to 69.3V would all produce exactly the same resulting number.
From a purely practical standpoint, I have heard a few low slew rate amps, which barely satisfied the 0.5 V/uS per peak volt criteria, sound subjectively better than a few other amps with wild slew rates and about ten times the nominal speed. This taught me that slew rate is good, but is not the universal panacea many hope for. People will always tend to look for one broad sweeping method of solving everything, even if that never works as easily as that.
Obviously, when setting the upper limit of the OLB, like it or not, we are also setting the amount of NFB we will use, or if you like, have to work with. AFAIK, you can never have both, wide OL bandwidth means less gain, and that mean less NFB required.
While I am a proponent of wide OL bandwidth, I am NOT against NFB as such, I do not think of it as automatically bad - quite to the contrary. Of the few amps I have heard which boasted zero global NFB, I wouldn't run out and buy any of them. The one common point was that they all sounded just a bit out of focus, so to speak, as if not quite finished. This leads me to believe that NFB is a good thing indeed, provided it is used to iron out what is an already good amp even without it. As say when the amp has an OL full power THD of say 0.5%, 20-20.000 Hz into 4 Ohms and is quite stable.
But even this is a sort of a floating point. Some amps will be quite happy with 12 dB of global NFB (e.g. Harman/Kardon Citation 24), others will need 17 dB for the same effect (e.g. H/K 6550 integrated), and still others will need more. Point is, it's very hard to generalize something that is model dependent.
But this should not put us off from trying.
Last edited:
To each his own, Jan.
A slew of H/K low NFB products through my home over the last 25+ years has clearly demonstrated to me that even with low global NFB numbers, the control the amps have over the speakers can still be rather complete.
And I refer to my AR94 speakers, NOT to my current speakers, which are a very gentle load.
http://users.ece.gatech.edu/mleach/papers/lowtim/feb76feb77articles.pdf
I didn't realize this idea dated back so long ago. Degan I think it defines your amplifiers origin role model very well. The author is starting from nowhere. We only need go back 7 years to 1969 to see the first op amp style power amps ( e.g. Sinclair Z30/50 ).
I know it is a mistake to lump slew rates with TID. Slightly different thing with overlapping outcomes.
I didn't realize this idea dated back so long ago. Degan I think it defines your amplifiers origin role model very well. The author is starting from nowhere. We only need go back 7 years to 1969 to see the first op amp style power amps ( e.g. Sinclair Z30/50 ).
I know it is a mistake to lump slew rates with TID. Slightly different thing with overlapping outcomes.
even if you had magic slewrate amp, IMHO still does not exist such speaker to faithfully reproduce square signal (inertia,mass,...)
or been invented some audiophile magic grease for speaker cones, which improves this ?
What factual basis is there that speaker cones or tiny dynamic drivers or balanced-armature drivers can not 'catch up' to the speed of high slew rates? If I understood you correctly.
So we need at least 40 V/uS slew rate op-amp's in the application of when they are driving speakers directly? Really?
Is this only taking into account the Voltage steps, or the maximum slope of the ideal sine waves as well?
I've read once that the slope needs to be considered, which makes sense to me, when discussing slew rate theoretically.
Yes. Slew rate of a sine wave is largest at the zero crossing of course.
The value, in Volt/Second, is just 2*pi*f*Vpeak.
So a 40V peak 20kHz signal has a max slewrate of 2*pi*2*10^4*4*10 = 50*10^5 volts/sec which is, of course, 5V/us.
So if your amp can handle 40V/us that's a factor of 8 safety factor.
Jan
What's f in the formula?
You seem to know the mathematics well, do we factor in settling time as well? What is an ideal settling time?
I've listened to op-amp's for years, playing around with them until the late hours of the night, which I really enjoy, I just wish I knew which exact parameters I'm actually listening to.
In my blind tests I can still hear IC's in the same way as I hear them sighted, so I've mostly set aside the illusion part but I have no idea what I'm hearing in the spec sheet, or if it's even in the spec sheet or in any dScopes at all, haha!
For example, the Cyril Bateman measurements of capacitors seem to be too err...... unique?... to exist in the standard functions of any dScope or in any spec sheet in the entire world.
That makes zero sense to me, but if assumed true it could mean there's a Cyril person for op-amp's as well? Just asking.
You seem to know the mathematics well, do we factor in settling time as well? What is an ideal settling time?
I've listened to op-amp's for years, playing around with them until the late hours of the night, which I really enjoy, I just wish I knew which exact parameters I'm actually listening to.
In my blind tests I can still hear IC's in the same way as I hear them sighted, so I've mostly set aside the illusion part but I have no idea what I'm hearing in the spec sheet, or if it's even in the spec sheet or in any dScopes at all, haha!
For example, the Cyril Bateman measurements of capacitors seem to be too err...... unique?... to exist in the standard functions of any dScope or in any spec sheet in the entire world.
That makes zero sense to me, but if assumed true it could mean there's a Cyril person for op-amp's as well? Just asking.
Scott, this is the star moment of diplomatic speech.
I'll be damned if I know which of the two you'd take.
Obviously, I'd take the 20 kHz OLB one.
That's what I thought, but I should have said I see no reason the second would be prefered over the first. I doubt anyone would put forth a design for either that would not have enough gain at 20kHz to make the distortions a don't care, the first better on this in any case.
This seperation of open loop BW from gain BW product leads to some fuzzy thinking. The Barrie Gilbert article you keep refering to clearly stated that it was the modulation of the unity gain crossover that mattered. You should also pay more attention to Ron's second paper, modern SS device speeds have made PIM a non-issue, though beloved valve amps still have it in spades.
I have demonstrated here before that via something like Vos feedback via the null pins you can make many op-amps have nearly infinite DC gain (even negative) and an extremely low OLBW. I suspect this would work with many PA's. Think about it, this amounts to very low order positive feedback (you are only nulling the offset at DC) and has no effect on the major AC characteristics.
Absolutely, this is how engineers USED to think. Sadly, now the norm is usually just "good enough."
And what is: "good enough"?
Are today's IC's "good enough" for any possible audio application? I'm sure some IC designers think so, but I have not found it so. Of course, we can't make EVERYTHING with all discrete components. It isn't practical or convenient, but IF you really care about audio quality, then you will find that discrete will beat IC, all else being equal. This includes my own designs, as I make both, successfully.
However, a JC-3 will NEVER beat a Vendetta Phono stage, and I doubt that any IC based design from anyone else, will either. No, only other discrete designs, especially tubes, can meet or beat a Vendetta SCP-2. Why is this so?
It is NOT that technology has not improved over the last 50 years that linear IC's first came into being, of course it has.
The first linear IC's, like the uA709 were wondrous when first introduced, but for audio, they were flawed. They suffered both from limited gain bandwidth and very nonlinear open loop linearity. Still, they were wonderful for lots of applications, and are still are today.
Now, it might be easily forgotten, but even the military was not always happy with the uA709, LM101, or the uA741, all developed in the '60's. Therefore they were willing to pay extra for better quality op amps made by Harris (formerly Radiation Inc) with true complementary bipolar parts, when they needed the extra performance. Many instrumentation companies also used Harris parts in critical locations in their test equipment as well, in the 1970's. Later, the major IC companies developed their own truly complementary processes and that is what we have available to us today, with the 'better' linear IC's.
This has moved up slew rate 10-50 times (not a factor, today, unless quality is totally disregarded), so TIM is virtually eliminated in preamps today. Power amps are still within range of having TIM, and that is what I am trying to point out to Nigel.
Now, if IC's have little or no TIM, and very good THD, aren't they virtually perfect? Why bother with anything else? Listening will still bring out the differences, so we still work on the inherent problems with today's IC designs. Could they be eliminated if a special AUDIO version of a particular IC was produced? Probably.
Are today's IC's "good enough" for any possible audio application? I'm sure some IC designers think so, but I have not found it so. Of course, we can't make EVERYTHING with all discrete components. It isn't practical or convenient, but IF you really care about audio quality, then you will find that discrete will beat IC, all else being equal. This includes my own designs, as I make both, successfully.
However, a JC-3 will NEVER beat a Vendetta Phono stage, and I doubt that any IC based design from anyone else, will either. No, only other discrete designs, especially tubes, can meet or beat a Vendetta SCP-2. Why is this so?
It is NOT that technology has not improved over the last 50 years that linear IC's first came into being, of course it has.
The first linear IC's, like the uA709 were wondrous when first introduced, but for audio, they were flawed. They suffered both from limited gain bandwidth and very nonlinear open loop linearity. Still, they were wonderful for lots of applications, and are still are today.
Now, it might be easily forgotten, but even the military was not always happy with the uA709, LM101, or the uA741, all developed in the '60's. Therefore they were willing to pay extra for better quality op amps made by Harris (formerly Radiation Inc) with true complementary bipolar parts, when they needed the extra performance. Many instrumentation companies also used Harris parts in critical locations in their test equipment as well, in the 1970's. Later, the major IC companies developed their own truly complementary processes and that is what we have available to us today, with the 'better' linear IC's.
This has moved up slew rate 10-50 times (not a factor, today, unless quality is totally disregarded), so TIM is virtually eliminated in preamps today. Power amps are still within range of having TIM, and that is what I am trying to point out to Nigel.
Now, if IC's have little or no TIM, and very good THD, aren't they virtually perfect? Why bother with anything else? Listening will still bring out the differences, so we still work on the inherent problems with today's IC designs. Could they be eliminated if a special AUDIO version of a particular IC was produced? Probably.
Last edited:
Think this was in the wrong thread..
well looking at this,
Testing audiophile claims and myths
I think we should all pack up and go home, in blind tests you can't tell any amp from any amp any system "reliably" from another and if its Valve /transistor/ chip/ or a combination of the two..
So there you go..buy a Midfi system and or a DAB radio and your in audio nirvana.
The question could also be asked if you were blindfolded could you tell what kind of car you were in on a short trip down a road..at town speeds!
9 out of 10 cats prefer steak but who knows whats in the cat food! (Steak on the tin must be steak)
If this is really true then measurements mean nothing, listening tests mean nothing, and circuit types don't mean a great deal either..
Regards
M. Gregg
well looking at this,
Testing audiophile claims and myths
I think we should all pack up and go home, in blind tests you can't tell any amp from any amp any system "reliably" from another and if its Valve /transistor/ chip/ or a combination of the two..
So there you go..buy a Midfi system and or a DAB radio and your in audio nirvana.
The question could also be asked if you were blindfolded could you tell what kind of car you were in on a short trip down a road..at town speeds!
9 out of 10 cats prefer steak but who knows whats in the cat food! (Steak on the tin must be steak)
If this is really true then measurements mean nothing, listening tests mean nothing, and circuit types don't mean a great deal either..
Regards
M. Gregg
"Sigh", fair enough if we just make it up like Geoff Kait, and Peter Belt?
- Status
- Not open for further replies.