Quality Power - Page 19 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #541 of 637 Old 02-22-2014, 07:38 AM
AVS Special Member
 
mcnarus's Avatar
 
Join Date: Nov 2007
Posts: 6,006
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 262
Quote:
I saw my first 16 bit 44 KHz capable ADC/DAC in 1971 and the application had nothing at all to do with audio.
IIRC, AT&T was using digital sound even earlier than that.

If you can't explain how it works, you can't say it doesn't.—The High-End Creed

mcnarus is offline  
Sponsored Links
Advertisement
 
post #542 of 637 Old 02-22-2014, 07:50 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by mcnarus View Post

Quote:
I saw my first 16 bit 44 KHz capable ADC/DAC in 1971 and the application had nothing at all to do with audio.
IIRC, AT&T was using digital sound even earlier than that.

May be true. A good cite or date or some darn thing would nail down that part of the discussion.

But what about the other uses of digital coding/decoding? It is clear that most if not all that was written about digital encoding in the 1930s was coming out or near to Bell Labs. But that was before the transistor so it supports my claim, not detracts from it.
arnyk is offline  
post #543 of 637 Old 02-22-2014, 08:38 AM
AVS Special Member
 
mcnarus's Avatar
 
Join Date: Nov 2007
Posts: 6,006
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 262
Quote:
May be true. A good cite or date or some darn thing would nail down that part of the discussion.
Does "I read it on the Internet somewhere" count? smile.gif

At any rate, pulse code modulation was patented in the 30s (in Britain), and Bell Labs was synthesizing sound by the 50s. (Remember "Daisy"?) The "audio industry," as we know it today, was just getting started then. All the key work, at least post-Edison and Tesla, was being done in the telecommunications field.

If you can't explain how it works, you can't say it doesn't.—The High-End Creed

mcnarus is offline  
post #544 of 637 Old 02-22-2014, 09:53 AM
AVS Special Member
 
JHAz's Avatar
 
Join Date: Mar 2009
Posts: 3,844
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 31 Post(s)
Liked: 145
Quote:
Originally Posted by DonH50 View Post

Well, almost... High temp SC only need go down to 77 K (-196 degC, -321 degF). Low temp, 4 K (-269 degC, -452 degF). Not quite 0 K, but pretty chilly...

That's OK, everybody exaggerates. smile.gif

I knew someone would be bust my chops on that. But a) too lazy to check the correct number and b) the exaggeration seemed funnier to me. smile.gif
JHAz is offline  
post #545 of 637 Old 02-22-2014, 10:18 AM
AVS Special Member
 
DonH50's Avatar
 
Join Date: Feb 2010
Location: Monument CO
Posts: 5,900
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 81 Post(s)
Liked: 229
Sorry, wasn't meant to bust your chops, and I thought it was hilarious, except somebody will do it... smile.gif Long week and working on tax stuff makes me cranky.

Actually, given the prices for some cables, I am surprised somebody isn't already. However, I have not seen anything so far that is long enough to use for speaker cables. Maybe too costly, maybe not needed, maybe technical issues with the films and cooling, I don't know (not my field). There have been large efforts to extend SC lines for power delivery but I have not read anything recently. Funding probably got cut; global warming will make it harder to cool the lines, so we'd better focus on fixing that first, eh? wink.gif

It'd be great if we could get some of the folk together for dinner and a beer (or whatever) sometime.

"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley
DonH50 is offline  
post #546 of 637 Old 02-22-2014, 12:09 PM
AVS Special Member
 
JHAz's Avatar
 
Join Date: Mar 2009
Posts: 3,844
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 31 Post(s)
Liked: 145
No offense taken at all. I just like the mental picture of some fool freezing himself to death to improve speaker wire conductivity. Presumably the voice coils would be immobile at those temperatures with any humidity at all.

As far as global warming, didn't you see The Day After Tomorrow (hope I am remembering the title right). Grab a bottle of whisky and wait. . .
JHAz is offline  
post #547 of 637 Old 02-22-2014, 01:13 PM
AVS Special Member
 
DonH50's Avatar
 
Join Date: Feb 2010
Location: Monument CO
Posts: 5,900
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 81 Post(s)
Liked: 229
Voice coils and other things immobile, yes, but the noise floor would be incredible... smile.gif I think it was Arny who said cables make a huge difference, just try removing them. biggrin.gif

If I remember that movie rightly, tequila might be a better choice...

"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley
DonH50 is offline  
post #548 of 637 Old 02-24-2014, 02:00 PM
Member
 
spkr's Avatar
 
Join Date: Apr 2013
Posts: 190
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 45 Post(s)
Liked: 48
Apparently, Mr. Salesman has his own herd of crickets because that's all I hear for an answer.
Quote:
Originally Posted by spkr View Post

Pardon me again, which survey is this conclusion from?
spkr is offline  
post #549 of 637 Old 02-24-2014, 02:19 PM
AVS Special Member
 
mcnarus's Avatar
 
Join Date: Nov 2007
Posts: 6,006
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 262
Oh, Christmas, this thread was dead in the road. Couldn't you just leave it there?

If you can't explain how it works, you can't say it doesn't.—The High-End Creed

mcnarus is offline  
post #550 of 637 Old 02-24-2014, 06:49 PM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 17,739
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 459 Post(s)
Liked: 318
So you guys are still here??? biggrin.gif
Quote:
Originally Posted by arnyk View Post

I'll play the "Derivative Technology" card against that one. ;-)

I saw my first 16 bit 44 KHz capable ADC/DAC in 1971 and the application had nothing at all to do with audio. It also cost about a half million dollars so I knew for sure it was never going to be used for audio. BTW it was acutally capable of a true 16 bits and 200 KHz sampling and had analog multiplexors in and out. It was attached to this:



and somone snuck in one night and did use it to digitize a LP... ;-)
Someone did? Hmmm. This technology is before my time smile.gif. Your comments that it was able to digitize audio with such high resolution and the ADC/DAC cost half a million dollars got my antenna up smile.gif. So I did some research. This is what I found.

The picture is that of Electronic Association Inc (EAI)'s 680. This was a hybrid computer that was analog but had a set of analog to digital and digital to analog converters that would let it interface with a separate digital computer. I assume this is the interface you are talking about. The interface was called the EAI 693 Hybrid Linkage Interface. I found a CMU 1971 document that was offering the 680 and 693 for sale. In there, they provided the original costs for each. The analog computer was $161,000 and the Linkage interface was $81,000. In the above you say that the cost of the ADC/DAC was $500,000. Even I throw the cost of the analog computer in there, we are at half the dollar amount you state. So that math doesn't add up, especially if one considers the cost of the ADC/DAC module along.

You next say that it was capable of 16 bits and 200 Khz sampling. That is remarkable resolution for its time. I researched this further and found a couple of university papers that had built FORTRAN libraries to access the 680/693 combo. One reference there set a limit of 30 Khz for the ADC the other said this:

"the A/D conversion system consists of a fifteen channel multiplexer and a single high-speed analog-to-digital converter. The settling time of the multiplexer is 10 usec. The ADC converts signals between +- 10 volts to 13 bits (plus sign bit) resolution; conversion time is 20 usec."

So clearly this was not a 16 bit converter. 13 is a far cry from 16. 20 usec conversion times translates to 50 Khz. Adding the MUX time of 10 usec, we get down to 33 Khz -- pretty close to the other reference (not sure if the MUX had to be programmed every time or not). Either way a far cry from 200 Khz that you state in your post.

Then there is the issue of voltages and feasibility of hooking up a turntable to this system. You would have to build a level translator to drive an ADC that wants +- 10 volts. Hard to imagine you building such a thing just for this overnight adventure smile.gif. Then there is a matter of getting access to a computer system of this cost and screwing around with it. Even if you had, this is not a stand-alone ADC. It simply generates data that must be captured by the digital system. You would have had to write a program to capture that data. But where would you have put it? I am sure no one was giving you permanent storage to put the bits there for the data rates you are talking about. The "Winchester" IBM hard disk of 30 megabytes was not to be invented until a few years later after your experiment.

And what would you have done if you did record the digital samples anyway? The DAC would not have had the proper anti-aliasing filter if your 200 Khz spec is right.

Maybe you say that this was some other ADC/DAC. So I went searching for evidence of people having such capabilities. This is the reference I found on AES web site: http://www.aes.org/aeshc/docs/recording.technology.history/digital.html

"1976 - The first 16-bit digital recording in the US was made at the Santa Fe Opera on a handmade Soundstream digital tape recorder developed by Dr. Thomas G. Stockham."

So what you are saying is that this bit of history on AES web site is wrong and that you had done this five years earlier in 1971. Quite amazing if true.

So.... how do you reconcile this data with your recollection? Is it possible you were mistaken about most of this? It is a long time ago so it would be understandable if so.

Amir
Founder, Madrona Digital
"Insist on Quality Engineering"

amirm is offline  
post #551 of 637 Old 02-24-2014, 08:49 PM
AVS Special Member
 
DonH50's Avatar
 
Join Date: Feb 2010
Location: Monument CO
Posts: 5,900
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 81 Post(s)
Liked: 229
...And that is why I don't post pictures of memories here...

Arny didn't say he built it, he said he saw it.

13 + sign = 14-bit equivalent (a lot of early ADCs were treated as sign-magnitude instead of the 2's complement most processors use today).

The mux settling time should not be included when calculating the sampling rate. It probably switched, then sat whilst conversion took place. It represents a delay before conversion can start, then conversion proceeds at whatever rate the ADC can do.

I hooked my turntable up to a similar system in college, using a (tube) amplifier to get up to +/-10 V. Not much different in gain than the chain from cartridge to speaker today.

I'd have to look and it ain't worth it to me, but while the first official 16-bit audio recording may have been made on the Soundstream, there were tons of other IEEE papers on various digitizers well before then.

Storage on analog computers is something I know little about despite having worked on one in college. The one I worked on did not have storage in the way we think of it today; rather, it stored charge on capacitors for intermediate results. The answer was immediately used (it was part of an industrial power plant) and recorded on tape (paper tape!) There was also a strip chart recorder we could hook up for tracking key variables and debugging. Seemed like something in the middle of the beast was always out of whack. It was one of the most frustrating and yet fun things I have worked on. Worked on a (very) few since but the first one sticks with you...

Good for you finding the info on that old beast, that is really cool! I wouldn't beat anyone over the head with a 40'ish year old memory, however. I have many times said "I may be wrong, but I'm not uncertain." And I have been wrong!

It'd be interesting to have an antique audio forum someplace where stuff like that could be posted and discussed. A lot of history is lost every day, maybe we could capture some of it. I have found sites with pix of old components, but rarely (can't really recall any) in a forum environment.

"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley
DonH50 is offline  
post #552 of 637 Old 02-25-2014, 02:41 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by amirm View Post

So you guys are still here??? biggrin.gif
Quote:
Originally Posted by arnyk View Post

I'll play the "Derivative Technology" card against that one. ;-)

I saw my first 16 bit 44 KHz capable ADC/DAC in 1971 and the application had nothing at all to do with audio. It also cost about a half million dollars so I knew for sure it was never going to be used for audio. BTW it was acutally capable of a true 16 bits and 200 KHz sampling and had analog multiplexors in and out. It was attached to this:



and somone snuck in one night and did use it to digitize a LP... ;-)
Someone did? Hmmm. This technology is before my time smile.gif. Your comments that it was able to digitize audio with such high resolution and the ADC/DAC cost half a million dollars got my antenna up smile.gif. So I did some research. This is what I found.

The picture is that of Electronic Association Inc (EAI)'s 680. This was a hybrid computer that was analog but had a set of analog to digital and digital to analog converters that would let it interface with a separate digital computer. I assume this is the interface you are talking about. The interface was called the EAI 693 Hybrid Linkage Interface. I found a CMU 1971 document that was offering the 680 and 693 for sale. In there, they provided the original costs for each. The analog computer was $161,000 and the Linkage interface was $81,000. In the above you say that the cost of the ADC/DAC was $500,000. Even I throw the cost of the analog computer in there, we are at half the dollar amount you state. So that math doesn't add up, especially if one considers the cost of the ADC/DAC module along.

You next say that it was capable of 16 bits and 200 Khz sampling. That is remarkable resolution for its time. I researched this further and found a couple of university papers that had built FORTRAN libraries to access the 680/693 combo. One reference there set a limit of 30 Khz for the ADC the other said this:

"the A/D conversion system consists of a fifteen channel multiplexer and a single high-speed analog-to-digital converter. The settling time of the multiplexer is 10 usec. The ADC converts signals between +- 10 volts to 13 bits (plus sign bit) resolution; conversion time is 20 usec."

So clearly this was not a 16 bit converter. 13 is a far cry from 16. 20 usec conversion times translates to 50 Khz. Adding the MUX time of 10 usec, we get down to 33 Khz -- pretty close to the other reference (not sure if the MUX had to be programmed every time or not). Either way a far cry from 200 Khz that you state in your post.

Then there is the issue of voltages and feasibility of hooking up a turntable to this system. You would have to build a level translator to drive an ADC that wants +- 10 volts. Hard to imagine you building such a thing just for this overnight adventure smile.gif. Then there is a matter of getting access to a computer system of this cost and screwing around with it. Even if you had, this is not a stand-alone ADC. It simply generates data that must be captured by the digital system. You would have had to write a program to capture that data. But where would you have put it? I am sure no one was giving you permanent storage to put the bits there for the data rates you are talking about. The "Winchester" IBM hard disk of 30 megabytes was not to be invented until a few years later after your experiment.

Our EAI 680 didn't have the standard EAI DAC/ADC conversion unit, ours was much better. That is probably why our ADC/DAC cost so much more. I don't remember the make or model but it occupied a full-height 19" rack.
Quote:
And what would you have done if you did record the digital samples anyway? The DAC would not have had the proper anti-aliasing filter if your 200 Khz spec is right.

The DAC/ADC formed the interface between the EAI 680 and an IBM 1130 16 bit digital computer. The samples were visible from within FORTRAN programs running on the 1130. We could store and retrieve the samples from the 1130's internal single disk hard drive, or if you want a real laugh, we could punch them out and read them from decks of IBM cards. I used to joke that the 1130 was my first personal computer because I logged so many hours on it that it almost seemed like that was so, but truth was that it was shared among the entire engineering department.

The ADC/DAC was mostly used for very prosaic things such as setting potentiometers and switches on the 680 from the 1130. The right front panel (with the wires) of the 680 was a patch panel. There were enough of those that lab teams (including me and my buds) and grad students (like me the next year) could have their own for the semester or whatever. You could slap the patch panel on, run a little FORTRAN program to set the pots and switches, and off you went!

http://en.wikipedia.org/wiki/IBM_1130
Quote:
"1976 - The first 16-bit digital recording in the US was made at the Santa Fe Opera on a handmade Soundstream digital tape recorder developed by Dr. Thomas G. Stockham."

Well that may have been the first live 16 bit recording, but the work I described was done in the very early 1970s. I was out of school and working as an engineer for Chrysler by 1973 and we digitized that LP several years earlier.
arnyk is offline  
post #553 of 637 Old 02-25-2014, 08:19 AM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 17,739
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 459 Post(s)
Liked: 318
Quote:
Originally Posted by DonH50 View Post

...And that is why I don't post pictures of memories here...

Arny didn't say he built it, he said he saw it.
I am not following your comment Don. Arny didn't just say he saw something. He said he digitized LPs on it. That requires full knowledge of the system including how to program it. Later he posts this in response to the well written article by Randy Hoffner http://www.tvtechnology.com/opinions/0087/digital-audio-sample-rates-the--khz-question-/184354:
Quote:
Originally Posted by arnyk View Post

However the article makes a few missteps such as:

[...]

"At that time, state-of-the -- art 16 -- bit digital-to -- analog and analog -- to-digital converters could operate at sample rates up to 60 kHz, and the sample rates under consideration therefore ranged from about 45 to 60 kHz"

I was personally hands-on with a 16 bit 200 KHz DAC/ADC over a decade earlier and there were no pretenses about it being SOTA
.

Given the strong statement Arny is making to counter the history there, I think it merits some discussion. Or at least it does when someone revives the thread. biggrin.gif
Quote:
The mux settling time should not be included when calculating the sampling rate. It probably switched, then sat whilst conversion took place. It represents a delay before conversion can start, then conversion proceeds at whatever rate the ADC can do.
That would be the case if the analog computer was not used. I was not sure if Arny did or did not use it and hence my comment that I was not sure. There is also delay in capturing the samples on the digital minicomputer which if it uses programmed PIO (programmed IO) and there is no pipelining in the ADC.
Quote:
I hooked my turntable up to a similar system in college, using a (tube) amplifier to get up to +/-10 V. Not much different in gain than the chain from cartridge to speaker today.
I didn't say it was hard Don. Just that you can't just take a turntable to that computer lab and hook it up on a whim.
Quote:
Good for you finding the info on that old beast, that is really cool! I wouldn't beat anyone over the head with a 40'ish year old memory, however. I have many times said "I may be wrong, but I'm not uncertain." And I have been wrong!
You certainly set the standard when it comes to modesty Don smile.gif. That is not what I am responding to. I am responding to Arny using this as clear evidence that some other person's article is wrong. And no softening words were used to downplay the specifics, and the statement repeated in another thread. I also mentioned that the inaccuracies may have been due to it being so long ago.

Amir
Founder, Madrona Digital
"Insist on Quality Engineering"

amirm is offline  
post #554 of 637 Old 02-25-2014, 09:14 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by amirm View Post

Quote:
Originally Posted by DonH50 View Post

...And that is why I don't post pictures of memories here...

Arny didn't say he built it, he said he saw it.
I am not following your comment Don. Arny didn't just say he saw something. He said he digitized LPs on it. That requires full knowledge of the system including how to program it. Later he posts this in response to the well written article by Randy Hoffner http://www.tvtechnology.com/opinions/0087/digital-audio-sample-rates-the--khz-question-/184354:
Quote:
Originally Posted by arnyk View Post

However the article makes a few missteps such as:

[...]

"At that time, state-of-the -- art 16 -- bit digital-to -- analog and analog -- to-digital converters could operate at sample rates up to 60 kHz, and the sample rates under consideration therefore ranged from about 45 to 60 kHz"

I was personally hands-on with a 16 bit 200 KHz DAC/ADC over a decade earlier and there were no pretenses about it being SOTA
.

Given the strong statement Arny is making to counter the history there, I think it merits some discussion. Or at least it does when someone revives the thread. biggrin.gif

I don't see the conflict.

I certainly had no intent or interest in countering history, then or now. What happened, happened, and that is that. ;-)

The technical efforts aren't comparable for the reasons I have already explained.
Quote:
Just that you can't just take a turntable to that computer lab and hook it up on a whim.

Why not? This was the early 1970s - long before 9/11! In those days the only security between the parking lot and the airplane was the person who checked boarding passes at the top of the jetway. Security was far more relaxed than that on the campuses of of any number of small universities. Security for this microcomputer lab was zilch. No locked doors during normal classroom hours, no gatekeepers, not even a grad student or prof keeping an eye on things with any degree of reliability. Were there likely co-conspirators with a full set of keys for after hours access? LOL!

I suggest finding another teapot to examine for nascent tempests! ;-)
arnyk is offline  
post #555 of 637 Old 02-27-2014, 07:30 AM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 17,739
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 459 Post(s)
Liked: 318
Quote:
Originally Posted by arnyk View Post

Our EAI 680 didn't have the standard EAI DAC/ADC conversion unit, ours was much better. That is probably why our ADC/DAC cost so much more. I don't remember the make or model but it occupied a full-height 19" rack.
The DAC/ADC formed the interface between the EAI 680 and an IBM 1130 16 bit digital computer.
Thanks for the new details. Based on that, I asked my friend, "Dr. Google" what he thinks is going on there. He reported back at first that he had better things to do, like searching for Miley Cyrus videos than look up things in dusty manuals from 1960s smile.gif. After some persuasion, explaining that the posts here lead to more ads, he capitulated and explained that there was a version of IBM 1130 called the 1800. Here is a picture of it:

IBM1800-9sm.jpg

Indeed it is in a 19 inch rack as you say. It had a pair of analog to digital converters with a resolution of 8, 11 or 14 bits (doesn't say if this included a sign bit or not). End to end conversion rate was 9,000 to 25,000 samples per second with the former being the 14 bit bits and latter, 8 bits.

So again, we are long, long ways away from 16 bits and 200 Khz. Which makes sense as this predates monolithic ADC/DACs by a few years and that technology was needed to get speed and accuracy. Building discrete would have been very hard if not impossible.
Quote:
Well that may have been the first live 16 bit recording, but the work I described was done in the very early 1970s. I was out of school and working as an engineer for Chrysler by 1973 and we digitized that LP several years earlier.
"We?" I thought it was just you sneaking into the lab Arny with your turntable. So this was a group effort then?
Audionut11 likes this.

Amir
Founder, Madrona Digital
"Insist on Quality Engineering"

amirm is offline  
post #556 of 637 Old 02-27-2014, 07:42 AM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 17,739
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 459 Post(s)
Liked: 318
Quote:
Originally Posted by arnyk View Post

Why not? This was the early 1970s - long before 9/11! In those days the only security between the parking lot and the airplane was the person who checked boarding passes at the top of the jetway. Security was far more relaxed than that on the campuses of of any number of small universities. Security for this microcomputer lab was zilch. No locked doors during normal classroom hours, no gatekeepers, not even a grad student or prof keeping an eye on things with any degree of reliability. Were there likely co-conspirators with a full set of keys for after hours access? LOL!
Well, your situation must have been different than ours. I was a system admin and programmer for our PDP-11 computer lab at college and my colleague and I were there anytime the lab was open. We would both lock it if we had to go out for any reason. The IBM 360 at the other campus was behind full lock and key and students could only work outside on punch card machines. You handed the cards to the geeks on the other side of the glass doors through a box like you have at bank drive-ins. And sometime later they gave you your print out. Likewise, the Univac system at University was totally inaccessible and even the computer terminals were under full security. The only thing that was not secured was punch card machine which I am sure no one was interested in stealing smile.gif. This is all late 1970s/early 1980s so not that far past your time with those systems.
Quote:
I suggest finding another teapot to examine for nascent tempests! ;-)
I think it is important for us to know the proper history of digital audio. Playing fast and loose with facts doesn't help. Folks need to know that it was not easy to achieve 16 bits of accuracy. The notion that it was all done in 1960s and without much effort rewrites the history and I don't think that is proper. Saying stuff like the ADC/DAC costing $500,000 which in today's dollars is $2.7 million dollars didn't help either. The computer company I worked for in early 1980s sold systems that would fill an entire room, were as loud as a rocket taking off, and was built using discrete ECL transistors and it still only costs $400,000. So the notion that an ADC/DAC module would cost that much just doesn't pass muster.

P.S. an annoying story regarding the Univac system. In the terminal room you had 30 minutes to work before someone else was supposed to use it. The pain in the neck admins would come and yank the power plug from your terminal if you were just one minute late without warning! frown.gif Here you were right in the middle of getting your program running and bam, the screen would go blank! You looked up and the guy was standing there with the cord saying your time is up!

Amir
Founder, Madrona Digital
"Insist on Quality Engineering"

amirm is offline  
post #557 of 637 Old 02-27-2014, 09:22 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by amirm View Post

Quote:
Originally Posted by arnyk View Post

Why not? This was the early 1970s - long before 9/11! In those days the only security between the parking lot and the airplane was the person who checked boarding passes at the top of the jetway. Security was far more relaxed than that on the campuses of of any number of small universities. Security for this microcomputer lab was zilch. No locked doors during normal classroom hours, no gatekeepers, not even a grad student or prof keeping an eye on things with any degree of reliability. Were there likely co-conspirators with a full set of keys for after hours access? LOL!


Well, your situation must have been different than ours.

Smaller school (at the time) and maybe what 10 years earlier?
Quote:
I was a system admin and programmer for our PDP-11 computer lab at college and my colleague and I were there anytime the lab was open. We would both lock it if we had to go out for any reason.

There were no such things for a lab computer in many schools in the earlier years. If you haven't noticed, our culture changes. ;-) Things have come full circle. Now, compuer security for this level of computational power has changed back and everything that we did on the six-figures system can probably now done on a smartphone with the right ap.
Quote:
The IBM 360 at the other campus was behind full lock and key and students could only work outside on punch card machines. You handed the cards to the geeks on the other side of the glass doors through a box like you have at bank drive-ins. And sometime later they gave you your print out. Likewise, the Univac system at University was totally inaccessible and even the computer terminals were under full security. The only thing that was not secured was punch card machine which I am sure no one was interested in stealing smile.gif. This is all late 1970s/early 1980s so not that far past your time with those systems.

This was a period of rapid change. Our University's first computer was a 1620 that was the be-all and end-all about 5 years earlier. The 1130 could easily blow it away on a any relevant grounds but size, cost, and power usage.The 1602 got the full mainframe treatment with gatekeepers, locked doors and the whole 9 yards.
Quote:
Quote:
I suggest finding another teapot to examine for nascent tempests! ;-)

I think it is important for us to know the proper history of digital audio. Playing fast and loose with facts doesn't help.

The irony of being lectured on that topic in this particular context is funny, even ludicrous. There has been nothing but the most accurate recitation of the facts possible. The phase " Playing fast and loose with facts" suggests dishonesty which in this context is highly inappropriate and insulting.
Quote:
Folks need to know that it was not easy to achieve 16 bits of accuracy. The notion that it was all done in 1960s and without much effort rewrites the history and I don't think that is proper.

Some people's questionable ideas about propriety don't dictate what the true facts were. The true facts were as I said.
Quote:
Saying stuff like the ADC/DAC costing $500,000 which in today's dollars is $2.7 million dollars didn't help either.

Please explain? Is this some dark truth that needs to be kept away from the public?
Quote:
The computer company I worked for in early 1980s sold systems that would fill an entire room, were as loud as a rocket taking off, and was built using discrete ECL transistors and it still only costs $400,000. So the notion that an ADC/DAC module would cost that much just doesn't pass muster.

The logical connection does not seem clear to me. The fact of the matter is that performance that is close to or at the state of the art is usually costly. True in 1965, true in 1975, true today.
Quote:
P.S. an annoying story regarding the Univac system. In the terminal room you had 30 minutes to work before someone else was supposed to use it. The pain in the neck admins would come and yank the power plug from your terminal if you were just one minute late without warning! frown.gif Here you were right in the middle of getting your program running and bam, the screen would go blank! You looked up and the guy was standing there with the cord saying your time is up!


Umm sounds pretty Gestapo to me! AFAIK the school I attended generally had more than enough computational power at the disposal of both the administration and the students.
arnyk is offline  
post #558 of 637 Old 02-27-2014, 03:03 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,874
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 81 Post(s)
Liked: 83
High end from the early days:

Very expensive at the time but interestingly the 16 bit dac was discrete and the linearity of the lower bits could be adjusted.

http://www.accuphase.com/cat/dp-80en.pdf
Frank Derks is online now  
post #559 of 637 Old 02-27-2014, 07:13 PM
AVS Special Member
 
DonH50's Avatar
 
Join Date: Feb 2010
Location: Monument CO
Posts: 5,900
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 81 Post(s)
Liked: 229
Heard it many times, but the components Arny and Amir are discussing predate it by quite a bit...

"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley
DonH50 is offline  
post #560 of 637 Old 02-28-2014, 05:11 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by Frank Derks View Post

High end from the early days:

Very expensive at the time but interestingly the 16 bit dac was discrete and the linearity of the lower bits could be adjusted.

http://www.accuphase.com/cat/dp-80en.pdf

Discrete converters of the day were based on precision resistor ladders and were inherently thermally unstable.

There was a widely used SOTA digital recorder made by 3M that needed to be adjusted frequently. An audiophile legend hangs on an assertion that it was improperly adjusted in this area for the recording Ry Cooder's "Bop Till You Drop".

Chip converters helped with that a great deal by making the resistors so small and tightly packed together that they all drifted together. It's the ratios of the parts values that matter.

The first Sony CD player was based on a chip converter and needed no adjustment. I've tested a few that have survived to this day and over 30 years later they are still adequately linear:



Now a moot issue because Sigma-Delta converters don't depend on highly precise resistor ladders nearly as much.

Dynamic range test of a CDP 101 ca. Y2K. The test signal was FS -60 dB so if there was appreciable low level nonlinearity it would stick out like a sore thumb.
arnyk is offline  
post #561 of 637 Old 02-28-2014, 07:33 AM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 17,739
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 459 Post(s)
Liked: 318
Quote:
Originally Posted by arnyk View Post

This was a period of rapid change. Our University's first computer was a 1620 that was the be-all and end-all about 5 years earlier. The 1130 could easily blow it away on a any relevant grounds but size, cost, and power usage.The 1602 got the full mainframe treatment with gatekeepers, locked doors and the whole 9 yards.

The irony of being lectured on that topic in this particular context is funny, even ludicrous. There has been nothing but the most accurate recitation of the facts possible. The phase " Playing fast and loose with facts" suggests dishonesty which in this context is highly inappropriate and insulting.
I apologize for making you feel that way Arny. In this forum, we critique each other's assessment of audio technology and that is what I am doing. This is what started our discussion:
Quote:
Originally Posted by arnyk View Post

I'll play the "Derivative Technology" card against that one. ;-) I saw my first 16 bit 44 KHz capable ADC/DAC in 1971 and the application had nothing at all to do with audio.

Your comment was that digital audio existed way earlier in the form of above ADC/DAC and you digitizing LPs with it. Just like hearing someone say their three ton truck gets 100 miles/gallon, when I see stats like that I cringe and want to know how they were possible. You said your adventure was in 1971 so the system would have been designed and built at the latest in late 1960s. It is quite remarkable for analog to digital conversion to have advanced so much and hence my questioning of the historical perspective that you are providing for digital audio recording capability.

Anyway, I solved the puzzle! The key was remembering that you went to school at Oakland University in Michigan. Combining that with the fact that you said they used an IBM 1130 and EAI 680 popped up this published technical paper from two researchers at Oakland University: "A Hybrid Computer Program for the Visual Display of Compensatory System Model Parameters" by Glenn A. Jackson and Gerald Brabant published in 1972. In there they show the computer they used for this program:

i-8pLhX5x-XL.png

So the two computing devices, the analog EAI 680 and IBM 1130, are there confirming that we are looking at the same system you used. But it doesn't say what the analog to digital and digital to analog converters. Fortunately it does so in the summary at the end:

i-2QMVVpz-XL.png

So as you see, my first guess was right that the ADC/DAC was courtesy of EAI 693 interface option. This makes sense if you think about it: they would have wanted to properly capture the output of the analog computer and provide the reverse so why would they buy an ADC/DAC from any other manufacturer? They wouldn't. This is the spec that I provided in my original post on the capabilities of this interface module:

"The A/D conversion system consists of a fifteen channel multiplexer and a single high-speed analog-to-digital converter. The settling time of the multiplexer is 10 usec. The ADC converts signals between +- 10 volts to 13 bits (plus sign bit) resolution; conversion time is 20 usec."

Inverting 20 microseconds we get max of 50 Khz. Allowing for settling time of the ADC (not the MUX) will roughly get us to the 30,000 samples/sec that the other report stated. The ADC/DAC interface cost was $81,000.

Putting it all together, the best your system could do is one channel at 14 bits at 30 Khz sampling. This is a far cry from 16-bits, 200 Khz that you stated the system was capable of performing. And I already talked about challenges of capturing data and storing them on that system in order to play them back to have some semblances of digital audio as we know it today.

So on any basis we look at, your recollection of the facts were not correct. Digital audio as we know it did not come about until we could play MP3s in real-time at CD resolution and sample rate. Combine that with CD ripping and software to manage it (e.g. Winamp), and we off to the races. Add to it the Diamond RIO audio player making it portable and good bit of piracy (Napster) and we had the full revolution on our hands. None of this dates back to any analog hybrid computer and instead rests solidly on VLSI developments way past your timeline. Hope you agree now.

Amir
Founder, Madrona Digital
"Insist on Quality Engineering"

amirm is offline  
post #562 of 637 Old 03-22-2014, 01:54 AM
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 29
I have a question on class a amps. Please correct me if I'm wrong, but as far as the ability for the amp to control the speaker, is it not the ability of a beefier class ab amp to stay in class a for longer than a slimmer avr? My understanding might be wrong, but in class a the amp has more instantaneous power and control over the speaker?
Heinrich S is offline  
post #563 of 637 Old 03-22-2014, 03:25 AM
 
Jonathan Parker's Avatar
 
Join Date: Mar 2014
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10

The Class rating has nothing to do with it. Your Class AB amplifier does not behave like a VTEC engine which I think you are trying to imply in your previous post? Behind every amplifier lies a well designed (read very clean) power supply that has the capacity to provide an adequate reservoir of power that might be needed once you factor in the power stage requirements and losses/efficiency (such as heat) into the equation.

 

Consider the power supply as the baseline from where everything starts. As AVR manufacturers have to walk a finely balanced tightrope when it comes down to AV features, licensing fees, parts count and complexity as both directly affect cost and reliability, the power supply usually ends up as a bit of compromise solution. It might work fine with 90% of the speaker loads out there and 90% of the audio situations you'll find. If you find yourself in the latter category.

 

This is in particular why I like Anthem gear as you essentially get the same processor throughout the range, which means you can buy an entry level Anthem AVR and hook it up to a quality 3 channel amplifier for the price of most top of the range AVRs and it will handle anything you can throw at it. You get great power, great processing and room calibration. My 2c.

Jonathan Parker is offline  
post #564 of 637 Old 03-22-2014, 06:08 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by Heinrich S View Post

I have a question on class a amps. Please correct me if I'm wrong, but as far as the ability for the amp to control the speaker, is it not the ability of a beefier class ab amp to stay in class a for longer than a slimmer avr?

The idea that operating amps in Class A improves sound quality is an audiophile myth. According to Cordell and Self who are leading authorities about power amp design, the lowest distortion is obtained by an optimally-biased Class AB amplifier. I've found this to be true in my own power amp design projects. Class A is a near-total waste of both electrical power and the capabilities of the most expensive hardware in the power amp.
Quote:
My understanding might be wrong, but in class a the amp has more instantaneous power and control over the speaker?

Your understanding may be somewhat flawed. Running a power amp in Class A wastes about half the useful capabilities of the power supply and output stage. IOW take the same parts, build a Class AB amp and set it aside and then build a Class A amp. The Class AB amp will have about twice the current output capability just because it isn't wasting so much power fighting against itself. That's what a class A amp does - it wastes huge amounts of power with the positive half of the output stage battling with the negative half of the output stage and none of that power ever getting near the load. With class AB amps almost all of the power either goes into the load or it stays in the AC power lines in your house.

The beef to control the load, and drive reactive loads is concentrated in the amp's output stage. The intelligence that shapes both load control and reactive load handling is in the amp's negative or inverse feedback system which may be in a loop around most of the amp or local to the output stage or both.

So far I haven't mentioned the power transformer which is a major source of bulk and weight that so many audiophiles seem to lust over. That's because sizing the power supply so it can drive a resistive load is where most of that bulk and weight goes.

There is an interesting factoid, which is that resistive loads dissipate the energy they receive from the amplifier, but reactive loads quickly return the energy they receive to the source.

This is just one more way that claims about the desirability of big heavy power supplies fails. The other way they fail relates to the crest factor of music which I have been belaboring for months.
arnyk is offline  
post #565 of 637 Old 03-22-2014, 06:28 AM
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 29
Quote:
Originally Posted by arnyk 
Your understanding may be somewhat flawed. Running a power amp in Class A wastes about half the useful capabilities of the power supply and output stage. IOW take the same parts, build a Class AB amp and set it aside and then build a Class A amp. The Class AB amp will have about twice the current output capability just because it isn't wasting so much power fighting against itself. That's what a class A amp does - it wastes huge amounts of power with the positive half of the output stage battling with the negative half of the output stage and none of that power ever getting near the load. With class AB amps almost all of the power either goes into the load or it stays in the AC power lines in your house.

But then why do Class A amplifiers exist if they are such a compromise? Why do manufacturers charge huge premiums for designs that are compromised as you say? That doesn't make sense to me.
Quote:
So far I haven't mentioned the power transformer which is a major source of bulk and weight that so many audiophiles seem to lust over. That's because sizing the power supply so it can drive a resistive load is where most of that bulk and weight goes.

But a good power supply is essential in the design of the amplifier. Correct? An AVR tries to be a jack of all trades and a master of none. So it's power supply design will be compromised compared to a dedicated power amp that cuts fewer corners?
Heinrich S is offline  
post #566 of 637 Old 03-22-2014, 06:35 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by Heinrich S View Post

Quote:
Originally Posted by arnyk 
Your understanding may be somewhat flawed. Running a power amp in Class A wastes about half the useful capabilities of the power supply and output stage. IOW take the same parts, build a Class AB amp and set it aside and then build a Class A amp. The Class AB amp will have about twice the current output capability just because it isn't wasting so much power fighting against itself. That's what a class A amp does - it wastes huge amounts of power with the positive half of the output stage battling with the negative half of the output stage and none of that power ever getting near the load. With class AB amps almost all of the power either goes into the load or it stays in the AC power lines in your house.

But then why do Class A amplifiers exist if they are such a compromise? Why do manufacturers charge huge premiums for designs that are compromised as you say? That doesn't make sense to me.

Class A amplifiers exist for the same reason that people "improve" their AVRs with expensive DACs.

Quote:
So far I haven't mentioned the power transformer which is a major source of bulk and weight that so many audiophiles seem to lust over. That's because sizing the power supply so it can drive a resistive load is where most of that bulk and weight goes.
Quote:
But a good power supply is essential in the design of the amplifier. Correct?

Good does not necessarily mean big and heavy, right?

I didn't pick my wife because she was the biggest and heaviest girl in the room, if you catch my drift. Ditto for the house and car. ;-)
Quote:
An AVR tries to be a jack of all trades and a master of none.

So the high end tell us. They are always right, eh?
Quote:
So it's power supply design will be compromised compared to a dedicated power amp that cuts fewer corners?

That's another high end audio myth.

If AVR or amplifier power supplies were designed for playing music and drama they would be different.

The FTC rules and commercial pressures force all amplifiers whether in an AVR or stand alone, to be inefficiently designed and built.
arnyk is offline  
post #567 of 637 Old 03-22-2014, 06:42 AM
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 29
Quote:
Originally Posted by arnyk View Post


If AVR or amplifier power supplies were designed for playing music and drama they would be different.

The FTC rules and commercial pressures force all amplifiers whether in an AVR or stand alone, to be inefficiently designed and built.

Sorry, please explain what you mean by the above. Especially the bold part.
Heinrich S is offline  
post #568 of 637 Old 03-22-2014, 06:57 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by Heinrich S View Post

Quote:
Originally Posted by arnyk View Post


If AVR or amplifier power supplies were designed for playing music and drama they would be different.

The FTC rules and commercial pressures force all amplifiers whether in an AVR or stand alone, to be inefficiently designed and built.

Sorry, please explain what you mean by the above. Especially the bold part.

What I mean is that according to the USA Federal Trade Commission (FTC) audio amplifiers for consumer use must be advertised and have power ratings based on measurements that model music as a pure, steady sine wave being delivered to a resistive load. This has a strong influence on the world market. Music has a high crest factor while steady pure sine waves have a very low crest factor. Speakers typically have far higher average impedance than their rated impedance.

These asymmetries between law and reality make audio gear bigger, heavier, and more expensive than it needs to be.

While naive audiophiles lust over power amps that weigh a lot, amps that sound best and produce the most clean power possible at any price point would be smaller and lighter.
arnyk is offline  
post #569 of 637 Old 03-22-2014, 07:02 AM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,874
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 81 Post(s)
Liked: 83
Despite the inefficiency of class A it is still the most linear output topology for an output stage in an amplifier.
Optimally biased class B is a close second. Within a healthy feedback loop the distortion is negligible.

Class A power amplifiers are easy marketing. For example Pass amps depending on a linear output stage in Class A due to the fact the output stage is outside the feedback loop to claim the 'no feedback loop' certain audiophiles adore..
A problem with class A is the much higher ripple current from charging the additional buffer capacitors. It takes proper engineering in the input stages to prevent power supply ripple injecting into the signal.
For example in Pass amplifiers that have poor psrr due to the poor input stage design the resulting distortion easily exceeds that of an optimal biased class B output stage.

For preamplifiers or dedicated headphone amplifiers class A is still the best output topology. The energy lost is negligible in these cases.
Frank Derks is online now  
post #570 of 637 Old 03-22-2014, 07:20 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 13,807
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 376 Post(s)
Liked: 1027
Quote:
Originally Posted by Frank Derks View Post

Despite the inefficiency of class A it is still the most linear output topology for an output stage in an amplifier.

Have you ever designed power amps and/or had an existing design on the test bench and hooked up to a distortion measuring test set and varied the bias?

One obvious fact is that Class A operation halves the amount of current you can get out of a given output stage and power supply aside from heatsink considerations which only make things even worse. This has the effect of quartering the amount of power that can be delivered to the load. That means that the Class A amp made from the same parts starts clipping at a far lower output power, drastically increasing distortion.

So if you look at the big picture, class A is the least linear output topology for an output stage in a power amplifier. ;-)
arnyk is offline  
Reply Audio theory, Setup and Chat

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off