I checked that audio show link and looks interesting , considering going.
How would you rate it against for example east coast show?
-
My uncle Lloyd was named after Lloyd George but I don't think we are related to the former British prime minister. (Although my maternal grandmother's maiden name was Warner and I am told she was a distant relative of the Warner Bros Warners.)
I haven't been to many audiophile shows so I wouldn't be the one to ask. The L.A. Audio show will be a bit of a wild card since the LAOCAS parted ways from T.H.E. Show over the summer.
I would hope that the reason that my amplifiers 'sound good' according to the reviewers is because of both what I contribute and also avoid in my designs. The original design is over 40 years old, but it would still hold up well today. That was the JC-3 (1974) It was designed before I went to servos, but it was very fast for the lowish f(t) 4MHz that I was forced to use at the time.
Attachments
Last edited:
Sorry, Bonsai, please not be 'clueless'. There is much more to it, I actually measured phono worst case rise-times. 10us AFTER EQ is observable.
Good luck with the .5ppm Class B power amp! '-)
The added related factor that often accompanies high slew rate is 'differential phase'. Check it out.
Good luck with the .5ppm Class B power amp! '-)
The added related factor that often accompanies high slew rate is 'differential phase'. Check it out.
Last edited:
Use it -- or any opamp for that matter -- in a composite to boost the effective slew rate to any level required. Because we don't ever actually slew our amps it does not matter that the "true slew" rate won't be higher when actually overpacing the amp (going open-loop).
You can make the 1st list but the 2nd will be short I'm afraid.
In science and engineering, progress goes in steps of generations, generally. New kids come up with new inventions and improvements, and the 'sitting' generation will fight them tooth and nail hanging on to their position, so you need to wait for one generations to disappear and then you make a quantum step in progress.
Happens in all walks of society, even economics and politics.
Jan
My view of slew rates is they show that most modern amplifers are mostly current amplifiers except when a TIS ( trans-impedance stage or VAS ). When this is made optimum the byproduct is a high slew rate. This can even be coupled with simplicity. Rarely is as most off the shelf op amps need to suit many needs. For example where current flow to external circuits is critical.
As far as I can see op amp design would be greatly simplified if a MOS FET output stage is added. Whilst they add significant problems of their own they do not ask much of the driver that can be just a simple TIS/VAS. This is where descrete op amps are interesting as they offer big sized parts solutions.
One falacy of MOS FET problems is the Cgs at perhaps as much as 1n5 will require very high driving current. It won't as it is mostly bootstrapped to the load. The Cgd will be important as it is anchored to the voltage rail ( usually far lower capacitance ). Sadly I don't think an ideal MOS FET exists for this if wanting class AB use. IRF530/9530 might be OK if 20 mA class A. At +/- 45 V rails. I suspect a simple double LTP would be OK as the op amp and >3 mA VAS could work. If wanting to be very fancy a JFET input LTP. BF720/721 ( BD139/140 BC639/640 ) for LTP 2. BC337/327-40 as input? Some think a JFET square law is an advantage. I suspect it is merely that the transconductance is less and forces the correct design choices. I have tried Darlington and Cfbp TIS. On paper it's a great choice. Perhaps it is. It sounded like something was lost by doing it. The Cfbp ( Sziklai ) worked very well when using the usual Cdom, not a hint of instablility. The idea being to have more LTP 1 tail current choices.
As far as I can see op amp design would be greatly simplified if a MOS FET output stage is added. Whilst they add significant problems of their own they do not ask much of the driver that can be just a simple TIS/VAS. This is where descrete op amps are interesting as they offer big sized parts solutions.
One falacy of MOS FET problems is the Cgs at perhaps as much as 1n5 will require very high driving current. It won't as it is mostly bootstrapped to the load. The Cgd will be important as it is anchored to the voltage rail ( usually far lower capacitance ). Sadly I don't think an ideal MOS FET exists for this if wanting class AB use. IRF530/9530 might be OK if 20 mA class A. At +/- 45 V rails. I suspect a simple double LTP would be OK as the op amp and >3 mA VAS could work. If wanting to be very fancy a JFET input LTP. BF720/721 ( BD139/140 BC639/640 ) for LTP 2. BC337/327-40 as input? Some think a JFET square law is an advantage. I suspect it is merely that the transconductance is less and forces the correct design choices. I have tried Darlington and Cfbp TIS. On paper it's a great choice. Perhaps it is. It sounded like something was lost by doing it. The Cfbp ( Sziklai ) worked very well when using the usual Cdom, not a hint of instablility. The idea being to have more LTP 1 tail current choices.
One way to loose the highter harmonics is to hide them is hiss. This can be done by using higher gain. Some circuits do not get ragged if doing this. If I had a gain of 50 and 0.1 THD I would not run away from it. I suspect like dither hiss can make it nicer. When negative feedback is added we shift the distortion spectrum towards a non musical result. What we see is a lack of loop gain at HF. Hence my interest in video op amps.
This could be a simple reason why LP sounds nice. Also why MC is liked as it usually takes hiss to the higher levels we can tollerate. Julian Vereker asked me one day " can you hear hiss with my preamps on LP when an LP playing " I said no. He beamed and said " That's what thought " and walked away!
I once recorded CD ( Teac I think ) onto a Nakamichi Dragon on Maxell metal ( cheapest called MR I think ). It was Schubert's the Trout. This was always something I prized which my wife sent to the charity shop. It was pushed as far into the red as I could and no Dolby. If someone was to say this was why we like LP's I suspect it could be true. It's these two simple things. Compression and hiss almost where we can hear it.
This could be a simple reason why LP sounds nice. Also why MC is liked as it usually takes hiss to the higher levels we can tollerate. Julian Vereker asked me one day " can you hear hiss with my preamps on LP when an LP playing " I said no. He beamed and said " That's what thought " and walked away!
I once recorded CD ( Teac I think ) onto a Nakamichi Dragon on Maxell metal ( cheapest called MR I think ). It was Schubert's the Trout. This was always something I prized which my wife sent to the charity shop. It was pushed as far into the red as I could and no Dolby. If someone was to say this was why we like LP's I suspect it could be true. It's these two simple things. Compression and hiss almost where we can hear it.
Hang on. Your posted 'evaluation' process has no design in it. You roll and glom and listen. You have posted nothing about checking the device will work in that location, or re-optimising anything for that part. It's almost as if you consider data sheets to be used to stop your bench wobbling.
I may be selling you short, but your posts are 'buzzword', 'glom', 'listen', 'ship'.
Jan, tends to happen very much as you say, and the human behavior involved is very consistent with our understanding of modern cognitive psychology!
Interestingly, much of this old book is still quite insightful: https://www.amazon.com/Structure-Scientific-Revolutions-Thomas-Kuhn/dp/0226458083
Yes it is, it isn't some sort of conspiracy, it's a natural human reaction.
I agree with Bill it is not an iron law, his examples may be the exceptions that confirm the rule.
You see similar things in the area of electric cars (it took Musk to get the big three or four or five to appreciably move to electric propulsion); it too Branson to get NASA to move to reusable space transportation concepts.
Admittedly audio is a fringe area but because this behaviour is governed by human traits, you see the same things here.
Jan
Hi Jan,
Disagree ref electric cars and the Branson effect (given both space-x and orbital have supplied ISS and Branson has yet to reach orbit). But its a beer discussion. The Prius to me is far more significant than the Tesla. Its just fugly and slow.
Oh and the shuttle was supposed to be a reusable concept. It was just 30 years too early and hamstrung by being a military platform.
Bill
Disagree ref electric cars and the Branson effect (given both space-x and orbital have supplied ISS and Branson has yet to reach orbit). But its a beer discussion. The Prius to me is far more significant than the Tesla. Its just fugly and slow.
Oh and the shuttle was supposed to be a reusable concept. It was just 30 years too early and hamstrung by being a military platform.
Bill
Hi Bill, as you point out, not all developments should be thought of as scientific revolutions. However, such revolutions do occur from time to time, and when they do there is often much resistance.
For more intermediate cases, some people may accept new ideas sooner than others, and some will be the last to admit that a new model should replace an old one.
In some cases, people can never accept change and eventually they get old and fade away. The new generation isn't so wedded to prior models, and can more easily move on to new ways of thinking.
The book I mentioned describes some of the more dramatic and interesting cases, and makes observations about some common generalities.
For more intermediate cases, some people may accept new ideas sooner than others, and some will be the last to admit that a new model should replace an old one.
In some cases, people can never accept change and eventually they get old and fade away. The new generation isn't so wedded to prior models, and can more easily move on to new ways of thinking.
The book I mentioned describes some of the more dramatic and interesting cases, and makes observations about some common generalities.
Last edited:
John, 20us slew rates or ~50ns small signal rise times are absolutely fine for phono. The cartridge cantilever can't track those transients but, that's besides the point.
I can understand some customers wanting discrete circuits. To use an auto analogy 'Some like Porches, others prefer a Morgan'. The customer is always right and a good marketer will fill those needs honestly and for the right price. But, to claim opamps are 2nd rate . . . I don't think so.
Quick Experiment: Feed a 10 kHz square wave via an inverse RIAA network into an opamp RIAA (or a discrete one). The signal out of the IRN is a series of very fast rise time spikes that then quick decay exponentially to 0V. A 20V/us opamp reproduces the square wave perfectly.
If its good enough for LIGO and Einstein's Gravity Waves, I am sure its a perfect device for any high end audio stage. DBT and Measurements of course - no peeking ;-)
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.