John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
my upper range is generally set, as THE objective, to have the "natural" (i.e. open loop) slew rate at the point where it satisfies the 1V/us/Vpeak of the amp, i.e. say 40V/uS for a nominally 28.3 Vrms of output, or at least as near to that point as possible That implies an OL bandwidth of around 110 kHz.

See? Any of those books, if read, would tell you the definition of slew rate, and clearly state that the slew rate is not affected by the amount of feedback. It would also explain the difference between slew rate and the rise/fall time.

Now, missing such elementary knowledge (and the above is only an example), is anybody surprised that there's so little to discuss technically between engineers and the golden ear brigade?
 
Jan, SY's point is well taken. While some were indeed convincts, they were the treu minorits. It would be more accorate to say that the original settlers were social misfits, of the "wrong" religious perusuation fleeng persecution, and probably some indentured servants and a few adveturous people simply wanting a better life.

Australian colonies were intially remote prisons, people were sent there to work off their sentence, and if they were Irish, that was crime enough. With a few rebellious Scots and Welshmen thrown in.

In both cases, current politics aside, I can't help admiring both ex colonies for their overall achievements.

Just like there's a saying: God created all of the earth, only the Dutch created Holland. And it stuck, even today, top land reclamation know-how is still Dutch. But also so much more, not to ennumerate it all.
 
Just one detail regarding Christophe Esperad0's 4 point list: my upper range is generally set, as THE objective, to have the "natural" (i.e. open loop) slew rate at the point where it satisfies the 1V/us/Vpeak of the amp, i.e. say 40V/uS for a nominally 28.3 Vrms of output, or at least as near to that point as possible That implies an OL bandwidth of around 110 kHz.

I've demonstrated several times here that with a tiny amount of positive feedback of offset voltage you can take almost any amp including the Otala amp and give it an openloop BW <100Hz. Your concept of "natural" slew rate makes no sense, the implication that feedback increases some lower "natural" slew rate is just plain wrong.
 
See? Any of those books, if read, would tell you the definition of slew rate, and clearly state that the slew rate is not affected by the amount of feedback. It would also explain the difference between slew rate and the rise/fall time.

Now, missing such elementary knowledge (and the above is only an example), is anybody surprised that there's so little to discuss technically between engineers and the golden ear brigade?

Waly, you are twisting my words. I didn't qualify my view as anything but my view, certainly not as a technical requirement I never stated or implied that wide OL bandwidth is cruical, or even necessary for a good slew rate, although in the majority of cases in which the slew rate is of decisive influence by high VAS Miller compensation caps it's that necessarily high compensation is needed because the VAS is amde to work with tremendous gain rates. Larger caps will reduce slew rate, and the opposite, high local NHB will redue the required size of Millercaps and will consequntly improve slew rates.

I never ever even hinted that I consider myself an owner of "golden ears". I don't have golden ears, I only wide open ears and I do trust them even without using the 'scope or whatnot to tell me what I'm listening to. Therefore I avoid the logic "it measures well, therefore it must sound good.". THIS is another reason why engieeners and DIYers find it hard to talk to each other.

Your engineeriing blind belief in measurements as the only reference point of quality is not infrequently based on reference to the wonderful measuring gear engineers have at work, which you damn well know is bayond the financial reach ofmost DIYers, so frequent deands to "prove it" is in fact a fraud, because much of the equipement you use is not yours. And you end up with a great design which measures wonderfully, the only trouble being that such designes more ofthen than not sound about average at best. But it lets angineers impose their views on everyone, because they can "prove" with measurements all that needs proving.

This makes a DIYer so much more honest and practical. He's building something for himself, and all that matters is that he likes it because he's the one making it for himself, not to gather world wide glory as a desiger.

And, forgive me for saying so, I find it very pretentious of you to judge what I know or don't know, since you've never measured or heard anything I ever did, just as I would not dare to rebuff you by saying you're a tone or even stone deaf engineer who needs an oscilloscope to even realize he's listening to music, for exactly the same reason - I've never heard anything your did, and thus have no basis for qualitative comments.
 
I've demonstrated several times here that with a tiny amount of positive feedback of offset voltage you can take almost any amp including the Otala amp and give it an openloop BW <100Hz. Your concept of "natural" slew rate makes no sense, the implication that feedback increases some lower "natural" slew rate is just plain wrong.

Completely wrong interpretaion of what I was trying to say, obviously unseccefully.

Think back about two years, we had a discussion here on how big should our slew rate be. In literarture, I found several definition stating that if the slew rate of an amp was 0.5V/uS/peak volt of output (eg. Audio/radio Handbook, from National Semiconductor, ISBN 1-882580-35-4, now out of print), and John simply said, well, why not 40V/uS/peak Volt, and why stop there when we can elatively easily achieve 100 V/uS/Vpeak with modern devices? Tat souded reasonable to me, so I agreed.

But why not go for the aforementioned 40V/uS/Vpeak under open loop conditions? Frankly, I don't know whether it would make an amp sound better still or just the as before, but it can't hurt if it's stable enough. I would consider it a success, fullfilling the accepted slew rate even under open loop conditions. What's the harm? It can only be beneficial for DIM, perhaps not in eliminating it, but at least reducing the chances of it getting out of hand. I am not claiming it must be so, I'm just saying - why not?

So far, nobody here has shown why not.
 
Think back about two years, we had a discussion here on how big should our slew rate be. In literarture, I found several definition stating that if the slew rate of an amp was 0.5V/uS/peak volt of output (eg. Audio/radio Handbook, from National Semiconductor, ISBN 1-882580-35-4, now out of print),

There are lots of examples of amplifiers with >100V/us slewrate and <100Hz open loop BW. We have moved beyond the old school way of looking at only the slewrate/BW relationship of an undegenerated bipolar LTP (i.e. Solomon's paper).
 
Waly, you are twisting my words. I didn't qualify my view as anything but my view, certainly not as a technical requirement I never stated or implied that wide OL bandwidth is cruical, or even necessary for a good slew rate, although in the majority of cases in which the slew rate is of decisive influence by high VAS Miller compensation caps it's that necessarily high compensation is needed because the VAS is amde to work with tremendous gain rates. Larger caps will reduce slew rate, and the opposite, high local NHB will redue the required size of Millercaps and will consequntly improve slew rates.

I never ever even hinted that I consider myself an owner of "golden ears". I don't have golden ears, I only wide open ears and I do trust them even without using the 'scope or whatnot to tell me what I'm listening to. Therefore I avoid the logic "it measures well, therefore it must sound good.". THIS is another reason why engieeners and DIYers find it hard to talk to each other.

Your engineeriing blind belief in measurements as the only reference point of quality is not infrequently based on reference to the wonderful measuring gear engineers have at work, which you damn well know is bayond the financial reach ofmost DIYers, so frequent deands to "prove it" is in fact a fraud, because much of the equipement you use is not yours. And you end up with a great design which measures wonderfully, the only trouble being that such designes more ofthen than not sound about average at best. But it lets angineers impose their views on everyone, because they can "prove" with measurements all that needs proving.

This makes a DIYer so much more honest and practical. He's building something for himself, and all that matters is that he likes it because he's the one making it for himself, not to gather world wide glory as a desiger.

And, forgive me for saying so, I find it very pretentious of you to judge what I know or don't know, since you've never measured or heard anything I ever did, just as I would not dare to rebuff you by saying you're a tone or even stone deaf engineer who needs an oscilloscope to even realize he's listening to music, for exactly the same reason - I've never heard anything your did, and thus have no basis for qualitative comments.

Completely wrong interpretaion of what I was trying to say, obviously unseccefully.

Think back about two years, we had a discussion here on how big should our slew rate be. In literarture, I found several definition stating that if the slew rate of an amp was 0.5V/uS/peak volt of output (eg. Audio/radio Handbook, from National Semiconductor, ISBN 1-882580-35-4, now out of print), and John simply said, well, why not 40V/uS/peak Volt, and why stop there when we can elatively easily achieve 100 V/uS/Vpeak with modern devices? Tat souded reasonable to me, so I agreed.

But why not go for the aforementioned 40V/uS/Vpeak under open loop conditions? Frankly, I don't know whether it would make an amp sound better still or just the as before, but it can't hurt if it's stable enough. I would consider it a success, fullfilling the accepted slew rate even under open loop conditions. What's the harm? It can only be beneficial for DIM, perhaps not in eliminating it, but at least reducing the chances of it getting out of hand. I am not claiming it must be so, I'm just saying - why not?

So far, nobody here has shown why not.

One clear conclusion from my side: when it comes to basic electronics, you have no idea what you are talking about, and you are hiding the lack of technical knowledge behind a wall of meaningless words and the usual "sounds better" subjective stance.
 
There is a nightclub here in Los Angeles somewhere were four of my bass horns ended up. Before that they had some typical subwoofer enclosures making bass. What I was told was that they were not happy with the muddy bass and had a major complaint from one of the neighbors down the block where the bass from those subs was the only thing they heard, it was traveling through the ground and ended up in their work space, drove them nuts. They install my horns and now had the low frequency punch they wanted without the excess low frequencies which nobody seemed to miss. I guess this goes with some of the earlier statements that cutting the sub-sonic bass cleaned up the lower register of the bass. I can tell you that two of those horns stacked would have a fundamental loading at 1/4 wavelength of only 75HZ, four would have been tuned to 50hz. These things would hit you so hard in the belly you would think someone was physically touching you. Sure there would be harmonics below this cutoff but the natural rolloff of the horns and perhaps a simple low cut filter was all that was needed to make the neighbors happy and the owners and customers of that club more than impressed with some rock solid bass. Those enclosures were loaded with dual 18" JBL drivers by the way.
 
One clear conclusion from my side: when it comes to basic electronics, you have no idea what you are talking about, and you are hiding the lack of technical knowledge behind a wall of meaningless words and the usual "sounds better" subjective stance.

That, Waly, comes from the fact that I can actuallly hear the differences without having to measure anything, in complete contrast with you, who hides behind his degree that in fact that you can't. I still do measure, without fail every time, because I'm curios in general regarding what it takes to her the differences, if any.

You have not refuted any one of my beliefs, right or wrong. Must be because you can't, even if you have a degree in electronics, and I indeed do not. Nor do I need it to hear or not hear this or that, that seems to have no meaning to you. You have no idea what you're missing. I pity you. Over and out.
 
Poor models do not last long.

Who said this guys' name 3x?

Beetlejuice-best-movie-ghost.jpg


A mathematician once taught me that our models represent the closest approximation of the reality we perceive; but they are only needed for their utility. Once those models are superseded by more accurate or elegant solutions they do not need be cherished. But they are useful for context.

I'm not making any conclusions from that observation... just sharing.

Back to the gig. Have a great day.
 
That, Waly, comes from the fact that I can actuallly hear the differences without having to measure anything, in complete contrast with you, who hides behind his degree that in fact that you can't. I still do measure, without fail every time, because I'm curios in general regarding what it takes to her the differences, if any.

You have not refuted any one of my beliefs, right or wrong. Must be because you can't, even if you have a degree in electronics, and I indeed do not. Nor do I need it to hear or not hear this or that, that seems to have no meaning to you. You have no idea what you're missing. I pity you. Over and out.

A perfect summary of the "discussion", I don't have anything to add. You keep designing audio "by the ear", meantime electronics engineers are busy working on the new space telescope and the LHC improvements. Unfortunately, we can't hear the hadrons colliding.
 
Waly,
While on a technical side of things I agree with you that if you don't understand the basic concepts of electronics to go and design an amplifier or any other audio equipment would be fairly stupid. On the other hand you have those who's only objective is to listen to the best sounding music system, at least to their personal tastes, and then those people really could care less how you created the electronics. I think this is where all this back and forth comes from, the problem rears it's ugly head when those who don't understand the underlying electronics try to come up with fanciful reasons for why they think this or that amp or whatever sound better than another. Feedback seems to trip up so many, the concept that feedback can work at the speed it does and not cause problems that inherently seem to exist is beyond simple comprehension. Physics is physics is physics, and no matter how much we learn that never changes, just or understanding.
 
Concerning power amplifier slew rate requirement and vinyl. Cartridge output is stylus velocity sensitive, eg 5.6cm/s = 1V, therefore cartridge output slew rate represents stylus acceleration. Cutting limits acceleration to c 1000G peak, ie 1000*981 cm/s2, ie a slew rate of 0.12V/uS (divide by 5.6cm/s and substitute for 1V).

If a power amplifier output of 1W into 8R represents 0dB (5.6cm/s), amplifier output slew rate corresponding to 1000G is then 0.12V/uS * 2.82 ie 0.34V/uS

That is to say programme material for vinyl playback won't demand slew rate more than c 0.34V/uS per watt at 0dB level ???

Then it seems moot to consider power amplifier slew rate to be limiting factor. it will seldom be tested even allowing for significant headroom.......

Is this correct ?
 
Sy,
I will have to ask my ex brother-in-law which model JBL those were. They may have been the 2241 but I don't think so. I remember they had something like a 3/4" p to p xmax off the top of my head. I did that about 13 years ago. Most likely the 2242 model just looking at the xmax numbers.

By the way I still have the molded sections for the throat section before the horn flare if anyone actually had a use for more of them? I still remember how to build them and could build some more if anyone is interested, enough for eight of these double 18" enclosures. Four on a side would be enough for most any outdoor festival or anywhere were high bass spl is needed.
 
Last edited:
There is a nightclub here in Los Angeles somewhere were four of my bass horns ended up. Before that they had some typical subwoofer enclosures making bass. What I was told was that they were not happy with the muddy bass and had a major complaint from one of the neighbors down the block where the bass from those subs was the only thing they heard, it was traveling through the ground and ended up in their work space, drove them nuts. They install my horns and now had the low frequency punch they wanted without the excess low frequencies which nobody seemed to miss. I guess this goes with some of the earlier statements that cutting the sub-sonic bass cleaned up the lower register of the bass. I can tell you that two of those horns stacked would have a fundamental loading at 1/4 wavelength of only 75HZ, four would have been tuned to 50hz. These things would hit you so hard in the belly you would think someone was physically touching you. Sure there would be harmonics below this cutoff but the natural rolloff of the horns and perhaps a simple low cut filter was all that was needed to make the neighbors happy and the owners and customers of that club more than impressed with some rock solid bass. Those enclosures were loaded with dual 18" JBL drivers by the way.

RF of human torso ~ 8 -10 Hz so any multiple close by should give lots of slam. Say, ~ 80 - 100 Hz? Lots of drivers can produce significant SPL at those f .

Human Induced Vibrations
The constituents of the standing human body can be excited into resonance at frequencies between 3 and 100Hz. At the lowest of these frequencies the abdominal viscera can be set into (uncomfortable) oscillation. The torso and pelvis have excitable modes of oscillation in the 5-7 Hz range. The upper torso and spine begin to oscillate between 10-14 Hz while the head and shoulders can be excited between 20-30 Hz. Finally there is an eyeball resonance in the 60-90 Hz range. Because of the relative displacements of body parts resonances below 10Hz are particularly uncomfortable and can lead to serious damage if sustained.
I'm in the 8th decade of life and don't need slam but it's nice, I find, if the system can give reasonable musical performance below 40 Hz without infrasonic junk.
 
Kindhorman, the objection being made is in the "peaking under the hood", if I'm understanding it correctly. It truly begs the question whether those who hold a certain philosophy about this amplifier architecture or that would actually be able to hear, definitively, the difference, or if a series of well-designed and generally load intolerant amplifiers would all, more or less, sound the same. The more we black box our electronics, the better (from an auditory perspective).

I'm all for folks trying to find their most preferred system, and they would do best to ignore how the electronics were created. The problem is, those that shout the loudest about some sort of subjective experience are the also the group most likely shriek and run away when someone asks them to do a blinded test. You'd think they, of all people, would want such a test, as it allows their ears to do the talking. You know, to actually achieve the "best" sound experience.

It'd be nice if the subjective crowd would just be honest enough to say, "you know, I just like this". That's awesome--I hope most folks are into audio for their holistic enjoyment, so if the glow of tubes, or a thick alloy front plate with some blue glowy VU meters, or a low impedance feedback network makes you happy, why not just be honest about that?

Just don't try to rationalize it, because that's where one gets into trouble.
 
Status
Not open for further replies.