[pinnahertz]

Of course, with the speed like that, slew rate is bound to be great. Recent(ish) op-amps also usable in this circuit can swing up to 3000 V per microsecond. This can not be utilized to the full, since at least I have never heard of coupling capacitors that could even approach this figure - and the speed thus depends on coupling caps. Although the original price was relatively low $ 500 back in the day ( when Audio Research and Mark Levinson charged double or more for their premium efforts ), this did not allow for film coupling caps - stock units all have electrolytics. But - an important BUT - there IS ample space on PCB - indicating that prototypes have used film caps, which had to be replaced in production run with electrolytics because of economics. Needless to say, in my unit electrolitycs are used in signal path only in one location - because I have not yet found a film cap of required value and small enough size to fit in the available space.

This nonsense doesn’t even approach real electronics. There are thousands of fine capacitors available for very little cost that can easily pass a signal slewing that fast. Caps used in RF and video designs do it regularly, or those devices couldn’t work. The reason you have not found one is you don’t know what you’re looking for or why. And, to make a fine point of it, the best solution if caps bother you is to design so coupling caps are not necessary. AGI didn’t do that.

High speed opamps that slew over 4000 V/uS are useless for audio because their noise figure is more than 20dB higher than much more appropriate opamps that slew fast enough for any audio application. Fortunately there are no audio signal that tax, even slightly, and opamp slewing at 30V/uS, which are easy to come by and entirely appropriate for audio.

You are off on the AGI 511 date by 4 years, it was 1977-78. I remember the review in Audio, in fact I probably still have it. Oddly, it was never adopted as the “gold standard”, and is largely now forgotten. So much fo MHz bandwidth.

The need for RF frequency response in audio gear is unproven, undocumented, and vastly overrated. Far more important that brute-force bandwidth is linearity in the area above 20kHz where distortion products lie and can, if an amp is significantly non-linear, intermodulate and create products in the audible spectrum. The rather severe error here is to equate high frequency linearity with high frequency response, which is not universally correct, but is easy to test with non-military-secret equipment.

In short - it is great because it is great in stock form - and can be significantly improved still further. It is unlikely it will be "heard" in any system that does not place totally out of ordinary requirements for preamp.

There are no audio systems that place out of ordinary requirements on preamps. Frankly, preamps just aren’t that hard to design, even with ridiculous HF bandwidth. And that means your “holy grail” 511 with its advantages cannot be heard...ever.

And, to date, nothing you’ve posted has ever been “in short”, much to everyone’s dismay.

 

[analogsurviver]

This nonsense doesn’t even approach real electronics. There are thousands of fine capacitors available for very little cost that can easily pass a signal slewing that fast. Caps used in RF and video designs do it regularly, or those devices couldn’t work. The reason you have not found one is you don’t know what you’re looking for or why. And, to make a fine point of it, the best solution if caps bother you is to design so coupling caps are not necessary. AGI didn’t do that.

High speed opamps that slew over 4000 V/uS are useless for audio because their noise figure is more than 20dB higher than much more appropriate opamps that slew fast enough for any audio application. Fortunately there are no audio signal that tax, even slightly, and opamp slewing at 30V/uS, which are easy to come by and entirely appropriate for audio.

You are off on the AGI 511 date by 4 years, it was 1977-78. I remember the review in Audio, in fact I probably still have it. Oddly, it was never adopted as the “gold standard”, and is largely now forgotten. So much fo MHz bandwidth.

The need for RF frequency response in audio gear is unproven, undocumented, and vastly overrated. Far more important that brute-force bandwidth is linearity in the area above 20kHz where distortion products lie and can, if an amp is significantly non-linear, intermodulate and create products in the audible spectrum. The rather severe error here is to equate high frequency linearity with high frequency response, which is not universally correct, but is easy to test with non-military-secret equipment.


There are no audio systems that place out of ordinary requirements on preamps. Frankly, preamps just aren’t that hard to design, even with ridiculous HF bandwidth. And that means your “holy grail” 511 with its advantages cannot be heard...ever.

And, to date, nothing you’ve posted has ever been “in short”, much to everyone’s dismay.

Well, those caps that do slew that fast have other disadvantages - directly with audio. In case you have not moticed by now, I am trying to balance everything into something that , in the end, sounds better. That means achieving the best sounding compromise. And for RF frequencies, values of capacitors are usually MUCH smaller than 47-68 microfarad ( I have seen both, batches from different years of manufacture ) used as coupling cap by AGI.

I agree the best way is to do away with coupling capacitors - if possible. That quickly leads to incompatibility with the vast majority of other equipment. And increases the danger of blowing something up, if the operator does not know EXACTLY what he/she is doing. AGI did not decide to market its power amp ( a 1000 V per microsecond design ) , because it required too tigtly matched semiconductors in order to work at all - and that was too costly. So, they had to throw in some form of "protection" against the direct DC while still passing signal below 1 Hz ( it can be any frequency below 20 Hz, depending on input impedance of the load, it can drive 20 Hz at 0 dB into 600 ohm ) . And, it works - and definitely can be ( or better said, can NOT be heard ) - as it behaves, for all practical purposes, like a DC coupled preamp - but without the possibility to generate excessive DC offset in also DC coupled power amp due to impedance mismatch. I like this solution better than DC servo circuits - YMMV.

AGI 511 is clean in its entire operating range even in stock form - and the advantages of 40 years younger op-amps only enhance the already great performance. The last thing it will do is to intermodulate the products in the audible spectrum.

There ARE audio systems that do place out of ordinary requirements on preamps. One I came across was the necessity to reduce the gain of the "power amp" significantly - in order to allow BW product to within what I have been comfortable with using componentry at hand and available at the time. If it were a commercially available product, an additional gain stage would have to be used - and that has detrimental effects on sound quality. A preamp with more output voltage swing than possible on + - 15 VDC rails of AGI had to be used - and that combination proved to be audibly superiour to whatever lower voltage preamp plus additional gain stage.

It is true that you can not "hear" the 511 - because, in a truly good audio system, it will sound transparent - MY standard, not, for example, that of @bigshot. The difference in soundstage made possible by such high speed but clean electronics as AGI has to be heard - no blah blah the size of entire head fi can not even approach the real listening experience. But it has to be used in "fast" systems that are also as phase coherent as possible. It was no coincidence I first heard it in Milan, Italy ( in 1979, it was THE audio show in the world, manufacturers often preffered premiering their latest/greatest in Milan than at CES in Las Vegas ) - driving Acoustat X electrostatics with matching electrostatic amps.

Since I do have schematics, 1974 is among the first entries in the documentation - even if mentioning as the beggining of the development of 511. It took something like 2 years ( for sticklers to the exact dates, I would have to check my archive ) to perfect it - and most reviews do date from 1979.

Short version - try as you might, it will be VERY difficult to find another preamp that will match it, let alone surpass it - in a real world situation. Remember, this is, first and foremost, a PHONO preamplifier ; there are two versions with different gain for MM cartridges and there is an additional MC stage ( rare as hen's teeth - it did/does exist, I do have schematics furnished by AGI, but have not been able/willing to build it yet ) that can be paired with any of the two MM gain versions. Needless to say, it requires the best possible cartridge/tonearm/turntable front end in order to truly appreciate the 511 . It really has to be heard to be believed.

 

[pinnahertz]

Well, those caps that do slew that fast have other disadvantages - directly with audio.

Again, complete nonsense. There are MANY solutions. You are simply out of your depth here. The problems with capacitors and audio are well known...and also HIGHLY over-rated. I've actually tested...with real caps and test gear...what caps do to audio performance. The series of articles "PIcking Capacitors" by Walt Jung and Richard Marsh is a good place to start, though slightly out of date on the values available in certain types.

I agree the best way is to do away with coupling capacitors - if possible. That quickly leads to incompatibility with the vast majority of other equipment. And increases the danger of blowing something up, if the operator does not know EXACTLY what he/she is doing.

Yet again, compete nonsense! A proper capacitor-less design requires absolutely nothing from an operator, not even the knowlege that the thing doesn't use coupling capacitors! Ever heard of a DC servo circuit?

<snipped irrelevant info>

There ARE audio systems that do place out of ordinary requirements on preamps. One I came across was the necessity to reduce the gain of the "power amp" significantly - in order to allow BW product to within what I have been comfortable with using componentry at hand and available at the time. If it were a commercially available product, an additional gain stage would have to be used - and that has detrimental effects on sound quality. A preamp with more output voltage swing than possible on + - 15 VDC rails of AGI had to be used - and that combination proved to be audibly superiour to whatever lower voltage preamp plus additional gain stage.

The example makes no technical sense at all. Reducing the gain of a power amp? Passive pad. Achieving it with a preamp with even more output swing? Silly, no engineering involved. As usual.

It is true that you can not "hear" the 511 - because, in a truly good audio system, it will sound transparent - MY standard, not, for example, that of @bigshot.

That's just silly. How could you possibly now what is "transparent" if your only reference is other device that you claim to be non-transparent? Did you compare with/without the 511?

The difference in soundstage made possible by such high speed but clean electronics as AGI has to be heard - no blah blah the size of entire head fi can not even approach the real listening experience. But it has to be used in "fast" systems that are also as phase coherent as possible. It was no coincidence I first heard it in Milan, Italy ( in 1979, it was THE audio show in the world, manufacturers often preffered premiering their latest/greatest in Milan than at CES in Las Vegas ) - driving Acoustat X electrostatics with matching electrostatic amps.

As soon as I hear "soundstage", I stop listening to the rest because it's applied as a 100% subjective term. I could make the same impression by connecting an unknown black box along with a powerful suggestion.

Short version - try as you might, it will be VERY difficult to find another preamp that will match it, let alone surpass it - in a real world situation. Remember, this is, first and foremost, a PHONO preamplifier ; there are two versions with different gain for MM cartridges and there is an additional MC stage ( rare as hen's teeth - it did/does exist, I do have schematics furnished by AGI, but have not been able/willing to build it yet ) that can be paired with any of the two MM gain versions. Needless to say, it requires the best possible cartridge/tonearm/turntable front end in order to truly appreciate the 511 . It really has to be heard to be believed.

Really? You're now citing gain differences for MM and MC? Why? What does that have to do with anything?

I'll tell you exactly what happened here. You read the reviews, saw the specs, and were highly impressed, as was I. You saw the construction quality, and was again impressed. So you believed with conviction that it should be the best sounding preamp in the world. Yet, there's no data to substantiate the impression! It didn't survive as a product. It's principles of design weren't universally adopted, even in the expensive stuff.

There's no need for RF response in audio equipment. The AGI is a great piece of gear, but so what? There are other great preamps that don't have (and don't need) RF response. You couldn't tell them apart in an ABX test, but don't believe the guy who's tried extensive ABX tests, try it yourself. I know, I'm asking for a miracle.

 

[bigshot]

Pinnahertz, I'm very glad you're able to address all the stuff that I can't bring myself to even read. I feel a little guilty when I dismiss people out of hand. Especially when they are at least at the level where they are using proper grammar and spelling. But there are some brains that are pretzel shaped and I just can't force my brain to contort itself that far to try to understand their "pretzel logic". I know when I reply to people, it takes me a little while to organize my comments and construct a clear post. Your posts are always well constructed and clear. There's effort behind what you do, as opposed to people who vomit out stream of consciousness nonsense. I guess what I'm trying to say is... "You're a better man than I am, Gunga Din!"

 

[analogsurviver]

Again, complete nonsense. There are MANY solutions. You are simply out of your depth here. The problems with capacitors and audio are well known...and also HIGHLY over-rated. I've actually tested...with real caps and test gear...what caps do to audio performance. The series of articles "PIcking Capacitors" by Walt Jung and Richard Marsh is a good place to start, though slightly out of date on the values available in certain types.
Yet again, compete nonsense! A proper capacitor-less design requires absolutely nothing from an operator, not even the knowlege that the thing doesn't use coupling capacitors! Ever heard of a DC servo circuit?
The example makes no technical sense at all. Reducing the gain of a power amp? Passive pad. Achieving it with a preamp with even more output swing? Silly, no engineering involved. As usual.
That's just silly. How could you possibly now what is "transparent" if your only reference is other device that you claim to be non-transparent? Did you compare with/without the 511?
As soon as I hear "soundstage", I stop listening to the rest because it's applied as a 100% subjective term. I could make the same impression by connecting an unknown black box along with a powerful suggestion.

Really? You're now citing gain differences for MM and MC? Why? What does that have to do with anything?

I'll tell you exactly what happened here. You read the reviews, saw the specs, and were highly impressed, as was I. You saw the construction quality, and was again impressed. So you believed with conviction that it should be the best sounding preamp in the world. Yet, there's no data to substantiate the impression! It didn't survive as a product. It's principles of design weren't universally adopted, even in the expensive stuff.

There's no need for RF response in audio equipment. The AGI is a great piece of gear, but so what? There are other great preamps that don't have (and don't need) RF response. You couldn't tell them apart in an ABX test, but don't believe the guy who's tried extensive ABX tests, try it yourself. I know, I'm asking for a miracle.

Well, "Picking Capacitors" is the very source of "capacitor blues" - and, if I am not mistaken, there is NO mention of slew rate ( dV/dt ) at all ! Everything else BUT the slew rate ... - and I know the article from the first issue of Audio in which this series run back then. And, unfortunately, it is dated - because of digitalization and miniaturization, the best capacitors are now either NOS, produced at really premium prices - or worse of all, not available at all. Today, a manufacturer can not build with the best that USED TO BE AVAILABLE - as there simply is no quantity for continuing running production. If you heard an amp Serial # 2345 at a friend and based on impression bought one from the recent batches with say Serial # 123456, you would expect it to sound at least similar to the unit that convinced you into buying in the first place - wouldn't you ? That's why only readily available parts that are reasonably expected to be available for at least another few years are used in any currently produced equipment. That does NOT mean that these readily available parts are the best in absolute terms.

Did you read that I wrote I do not like the DC servo circuits ? Would you put one on the INPUT of a moving coil preamplifier ? Would you risk a customer for your preamp suing you for blowing up an expensive phono cartridge ( 5 figures, the first one not necesarilly being one ) - because your preamp failed and took the cartridge with itself to ever haunting grounds ?

No, you misundersood again. The gain of the "power amp" had to be lowered in order for it to work as fast as required - and the front end did not have enough gain bandwidth product to do so at a normal input sensitivity. That was what has been possible 30 or so years ago, today most probably there ARE electronic parts capable of doing what was required while allowing for the gain to be raised enough to allow for the normal 1 V or so sensitivity for the full output.

Well, get an 511. Although not more than about 5000 units have been built ( and LOTS of those are in Japan, Italy ( VERY dedicated distributor made an unheard of effort at the time promoting it ) and elsewhere in Europe ( Germany, France , etc...), there must be some of them still within USA. Compare, listen, AB(X) ( remember: the inventor of 511 is the originator of double blind ABX ) - and you should be able to hear the difference. It is not night and day - more sunny and light overcast. I hope you can differentiate such difference(s).

And I can relate WHY you frown upon the word "soundstage". I know that in most studio gear , the soundstage is the firs thing that falls victim - due to the "20-20k is enough" syndrome being so firmly imbedded in minds of studio people. Place in series as many "20-20k" minded components as required in a typical studio workflow - and it is a miracle in itself whatt little of soundstage that does come trough - at all.
Imagine the sound from a studio built to AGI 511 standard ...

The gain differences for MM ( MC is another gain stage inserted between the MC cartridge and regular MM input, something that back then was called a head amp - FAR before the headphone world adopted the same name for an entirely different device with another task ) in AGI 511 is significant - because it is not acomplished by a single resistor change, but entire feedback RIAA filter. As these components - both resistors AND capacitors are EXACT values, NO tolerance parts ( resistors having exact value PRINTED on them , down to a single ohm ) are critical for the EXACT adherence to RIAA curve. There never has been - or is likely to be - a better/lower deviation from the ideal RIIA response in a real world analogue circuit. Only Peter Moncrieff has been able to measure exactly how small deviation from perfect AGI511 has - an it was 4 mB, with the test gear having limit to measure at 2 mB. The mB is NOT a typo - in words, miliBel . Every other test published you are likely to see will be perfectly flat line - below the limit of measuring equipment. That kind of dedication to perfection IMO does have to do with - everything ... - including, of course, the highest possible quality capacitors with zero tolerance as used in AGI 511 phono section.
Just for the record : there were 511, 511 A ( normal MM gain ), 511 H ( 6 dB higher gain for low output MM carts, like Technics EPC 100 Series ) and there was 511M - either low or high gain MM paired with MC head amp. 511M is, by far, the rarest of them all - never saw even a picture.

There is another, VERY important feature of the 511 phono preamp. How it deals with picking up the radio. First, radio transmission pickup occurs if the slew rate of the amp is lower than that of the RF input. OK, 90 MHz up to which 511 is specified ( it does exceed this figure, up to two times, depending on sample ) is still too low for FM. Normal phono preamps ( and FTC requirement in late 80s) place simply high enough value capacitor parallel to the phono input - so that no audible RF induced radio pickup can occur. That is usually around 500 pF and higher - which wreaks absolute havoc in frequency response with most, but most notably best MM phono cartridges ( which happen to see the lower capacitance, the better ). It is also one of the reasons why MC cartridges ( which are largely immune to capacitive loading, used only for fine tuning, not basic performance ) then really took off - under new legislation MM carts have been all but reduced to an afterthought ...

AGI does it differently. Its phono input is made like a two way loudspeaker crossover - up to certain ( for CD lovers still FAR too high frequency ...100 kHz and above ) , the LF pass portion is the phono audio output - and HF pass connects all RF garbage above that simply to - ground . You can connect ANY phono cartridge ( with their vastly different inductance and resistance parameters ) through ANY tonearm/turntable ( with their vastly different cabling/shielding/grounding schemes ) to the AGI 511 - and even if you tune your RF transmitter in the very same room EXACTLY to the cartridge/tonearm/turntable ELECTRICAL resonance - there will be absolutely zero radio transmission in your audio output. And it does not have to use a single additional pF on its input to accomplish this feat - meaning that cartridge capacitive loading still can work as intended for performance and not FTC requirements . One thing you will never, under any circumstances, hear trough AGI 511 is - radio - except that intentionaly fed from a tuner output to its tuner input, of course

Try that with any other phono preamp ...

I have seen and heard MANY preamps that dwarf the AGI 511 in novelty factor, in price, in X ( insert your number here ) thick front plates, in GodKnowsWhatElse, etc - and regardless of owning a few, when it comes to the actual listening, the old and trustworthy 511 gets >> 80 % listening time. For a conventional MM phono input, it is still - beyond any shadoe of a doubt -

“One of The Most Significant Preamps of All Time”

http://whathifi.whatgroupmag.com/old-school-vintage-sound-agi-model-511/

 

[bigshot]

Who cares about LPs? The format is demonstrably inferior to digital in every aspect except for the size of album covers and selection of titles at swap meets. I have tens of thousands of records myself, but I swear all this torrent of words on the subject bores me stiff. Please create a thread just for discussion of LPs and you can talk to yourself like this all day long without derailing this thread.

Castle, is it within your power to move off topic posts? If so, we could create a special thread for this topic and you could just move them into that thread when they inevitably pop up like weeds.

 

[71 dB]

Who cares about LPs? The format is demonstrably inferior to digital in every aspect except for the size of album covers and selection of titles at swap meets.

Some people like the distortions caused by vinyl's inferiority to digital audio. Since they don't understand digital and analog audio well, they assume vinyl must be superior because they like it more. The vinyl distortions should on the master if people love them, just as music production uses all kind of effects that people like. Why have the media produce the final steps in your product? Makes now sense, because that way the final step is not controlled. People have different TTs with different distortions etc.

 

[pinnahertz]

Well, "Picking Capacitors" is the very source of "capacitor blues" - and, if I am not mistaken, there is NO mention of slew rate ( dV/dt ) at all ! Everything else BUT the slew rate ... - and I know the article from the first issue of Audio in which this series run back then. And, unfortunately, it is dated - because of digitalization and miniaturization, the best capacitors are now either NOS, produced at really premium prices - or worse of all, not available at all.

Incorrect, but see below.

Today, a manufacturer can not build with the best that USED TO BE AVAILABLE - as there simply is no quantity for continuing running production. If you heard an amp Serial # 2345 at a friend and based on impression bought one from the recent batches with say Serial # 123456, you would expect it to sound at least similar to the unit that convinced you into buying in the first place - wouldn't you ?

No, not without knowing what changes were made in production.

That's why only readily available parts that are reasonably expected to be available for at least another few years are used in any currently produced equipment. That does NOT mean that these readily available parts are the best in absolute terms.

Lets get specific rather than using massive and irrelevant generalities. What parameters in a capacitor are no longer available? And "the sound quality" doesn't count! Be specific!

Did you read that I wrote I do not like the DC servo circuits ? Would you put one on the INPUT of a moving coil preamplifier ? Would you risk a customer for your preamp suing you for blowing up an expensive phono cartridge ( 5 figures, the first one not necesarilly being one ) - because your preamp failed and took the cartridge with itself to ever haunting grounds ?

That statement indicates a full lack of understanding of how a DC servo works. I'm not going to explain it, as that's a huge tangent, but for a start: you don't apply it to an input! No need, and no point. So, no chance at all of "blowing up" a cart. Works just fine, and is a very nifty way to eliminate huge caps in the audio path.

No, you misundersood again. The gain of the "power amp" had to be lowered in order for it to work as fast as required - and the front end did not have enough gain bandwidth product to do so at a normal input sensitivity. That was what has been possible 30 or so years ago, today most probably there ARE electronic parts capable of doing what was required while allowing for the gain to be raised enough to allow for the normal 1 V or so sensitivity for the full output.

It doesn't matter. The entire scenario of needing to reduce gain because of gain/bandwidth product, then having to hit it so hard that you needed a higher output preamp (like what, +/- 50V???) is irrelevant and stupid.

Well, get an 511. Although not more than about 5000 units have been built ( and LOTS of those are in Japan, Italy ( VERY dedicated distributor made an unheard of effort at the time promoting it ) and elsewhere in Europe ( Germany, France , etc...), there must be some of them still within USA. Compare, listen, AB(X) ( remember: the inventor of 511 is the originator of double blind ABX ) - and you should be able to hear the difference. It is not night and day - more sunny and light overcast. I hope you can differentiate such difference(s).

The response above was fully expected, but disappointing. It's typical audiophile: "You haven't hear it, so you don't know. You probably couldn't hear it anyway because you are 1)deaf 2)an idiot 3)your system is far to unsophisticated. Thanks for following suit with expectations, but none of that supports even a single one of your arguments.

And I can relate WHY you frown upon the word "soundstage". I know that in most studio gear , the soundstage is the firs thing that falls victim - due to the "20-20k is enough" syndrome being so firmly imbedded in minds of studio people. Place in series as many "20-20k" minded components as required in a typical studio workflow - and it is a miracle in itself whatt little of soundstage that does come trough - at all.

There it is again. This time it's the "20-20k minded".

Imagine the sound from a studio built to AGI 511 standard ...

I do. It's exactly what we get now. You've fallen into the "circle of confusion" once again. So you think everything needs to be capable of some RF frequency like 100mHz, but you want everything in the chain to do that. Most mics don't, can't and won't, and are chosen for other far more audible and important parameters. The mics that do are useful for very limited applications. We could digitize up to 100mHz, but we're not going to because there's no evidence it makes any difference above 20kHz or slightly more. No analog recorder will record above 20kHz reliably and consistently, and none at all above 25kHz that don't have some other fatal issues. No speakers can deliver anything to your ears above 25kHz at all, even if stated response is 40kHz, the ultrasonic beam width and air absorption blows it away. The whole argument collapses if you look at any individual segment of the chain, but you want the WHOLE CHAIN to play up to 100mHz!

The gain differences for MM ( MC is another gain stage inserted between the MC cartridge and regular MM input, something that back then was called a head amp - FAR before the headphone world adopted the same name for an entirely different device with another task ) in AGI 511 is significant - because it is not acomplished by a single resistor change, but entire feedback RIAA filter.

RIAA EQ doesn't dictate a feedback network, but ok...

As these components - both resistors AND capacitors are EXACT values, NO tolerance parts ( resistors having exact value PRINTED on them , down to a single ohm ) are critical for the EXACT adherence to RIAA curve.

There's no such thing as zero-tolerance parts. Even the ones with values printed on them have a tolerance, usually 1%, but possibly lower. But you clearly don't understand either the RIAA curve or what tolerances are necessary.

There never has been - or is likely to be - a better/lower deviation from the ideal RIIA response in a real world analogue circuit. Only Peter Moncrieff has been able to measure exactly how small deviation from perfect AGI511 has - an it was 4 mB, with the test gear having limit to measure at 2 mB. The mB is NOT a typo - in words, miliBel . Every other test published you are likely to see will be perfectly flat line - below the limit of measuring equipment.

Well, that may have been true in 1977, but not true today. Easily measured, with non-exotic equipment. But again, you're in the circle of confusion! So, we have a preamp with precise RIAA match to the mDB. Great. But how does it perform with a cartridge on it? Oops. You might measure it if you had a test records recorded with no EQ and a perfectly flat lathe. But then we get into the real problem, the record RIAA eq may not be that good. And since the entire produced response is based on what was heard in a studio with monitors that are not flat to 2mdB, likely not to 1dB, we have another random error built into the system. Then we hit your speakers in your room, and response errors become HUGE because I'm 100% certain you make no attempt to equalize them in the room (an audiophile atrocity). So why on earth would we need or benefit from 2mdB response accuracy anywhere? We cannot possibly achieve it at all, not even close.

There is another, VERY important feature of the 511 phono preamp. How it deals with picking up the radio. First, radio transmission pickup occurs if the slew rate of the amp is lower than that of the RF input. OK, 90 MHz up to which 511 is specified ( it does exceed this figure, up to two times, depending on sample ) is still too low for FM.

Wow. You couldn't BE more wrong! US FM (and most other countries) runs from 88-108mHz! AM radio is between 535-1605kHz! Analog TV channels 2-6 are between 55-83mHz! CB radio is centered around 27mHz (AM and SSB), HAM radio frequencies are all over, from 135 meters (135kHz), up to 50mHz in 13 separate frequency bands!

But...I also think you're misquoting the specs. You've confused mHz with kHz. From memory I recall that the AGI, while pretty wide band, didn't really perform much above 100kHz. That's Kiloherts, not Megahertz. So you're right, it rolls off below most high power RF. But it's far from the only preamp that would do that. For example, the Apt-Holman preamp was -3dB at 150kHz with the ultrasonic filter switched out. I recall the Marantz 7T being good to 100kHz too. So, no big deal there. However, RF proofing is quite another matter.

Normal phono preamps ( and FTC requirement in late 80s) place simply high enough value capacitor parallel to the phono input - so that no audible RF induced radio pickup can occur. That is usually around 500 pF and higher - which wreaks absolute havoc in frequency response with most, but most notably best MM phono cartridges ( which happen to see the lower capacitance, the better ). It is also one of the reasons why MC cartridges ( which are largely immune to capacitive loading, used only for fine tuning, not basic performance ) then really took off - under new legislation MM carts have been all but reduced to an afterthought ...

Wrong again, on several points. In fact, the only thing right in the above paragraph is that a MM cart won't work well into 500pf! But they don't have to. Simply placing a 500pf cap on the front end won't RF proof it at all. Several preamps, including one by Kenwood, and the notable Apt-Holman, solve the RF problem by not using a bipolar transistor at the front end. A bipolar junction acts as a diode detector/demodulator at RF frequencies, becoming the root cause. Eliminate that junction, you don't have a detector.

AGI does it differently. Its phono input is made like a two way loudspeaker crossover - up to certain ( for CD lovers still FAR too high frequency ...100 kHz and above ) , the LF pass portion is the phono audio output - and HF pass connects all RF garbage above that simply to - ground . You can connect ANY phono cartridge ( with their vastly different inductance and resistance parameters ) through ANY tonearm/turntable ( with their vastly different cabling/shielding/grounding schemes ) to the AGI 511 - and even if you tune your RF transmitter in the very same room EXACTLY to the cartridge/tonearm/turntable ELECTRICAL resonance - there will be absolutely zero radio transmission in your audio output. And it does not have to use a single additional pF on its input to accomplish this feat - meaning that cartridge capacitive loading still can work as intended for performance and not FTC requirements . One thing you will never, under any circumstances, hear trough AGI 511 is - radio - except that intentionaly fed from a tuner output to its tuner input, of course

Again...the actual electronics involved is very different here. What you describe (actually not in the schematics available on-line which show a bare-foot uA749 opamp as the front end!) is just a low pass RF filter. The function of which would be necessary for any preamp operating in a high RFI field.

Try that with any other phono preamp ...

Oh, yes. Been done. The Apt-Holman was designed within 1 mile of a high power broadcast transmitter! And I have actual hands-on experience with phono preamps and all sorts of other audio gear at both a 20kW FM transmitter site and a 50kW AM transmitter site. There are plenty of other preamps that will work out there, but none (AGI included!) will work without very special attention to installation and grounding.

I have seen and heard MANY preamps that dwarf the AGI 511 in novelty factor, in price, in X ( insert your number here ) thick front plates, in GodKnowsWhatElse, etc - and regardless of owning a few, when it comes to the actual listening, the old and trustworthy 511 gets >> 80 % listening time. For a conventional MM phono input, it is still - beyond any shadoe of a doubt -

“One of The Most Significant Preamps of All Time”

I thought that thing was special too...in 1977. But time moved on, it's not special today at all, nor has it been for 30 years.

 

[castleofargh]

Castle, is it within your power to move off topic posts? If so, we could create a special thread for this topic and you could just move them into that thread when they inevitably pop up like weeds.

no, I have the dominatrix pack. I only punish or say out loud, "now I have become death, the destroyer of words", anytime I delete a post.
for constructive actions, you'd have to ask a real admin.

personally I don't see an issue with people sharing their experiences. that IMO is on topic, either as a myth or as a mean to try and dispel one. I only wish people would try to provide actual data about conducted experiences instead of just some "trust me I tried and it is so because I say so".

 

[bigshot]

My main objection is that we can all be talking about ice cream or aardvarks and he'll reply about phono preamps and cartridges that go up higher than bats can hear. It's kind of like a bot at this point.

 

[TheSonicTruth]

My main objection is that we can all be talking about ice cream or aardvarks and he'll reply about phono preamps and cartridges that go up higher than bats can hear. It's kind of like a bot at this point.

Well, let me ask this: Audio 101 - can harmonics of a specific tone or combination appear on both sides of that fundamental, on the audio spectrum?

IE: A 300Hz tone might have multiple harmonics, say, at 600, 900, etc.

So, can a 40kHz tone generate harmonics BELOW its fundamental frequency, and possibly down into the range of human audibility?

 

[Zapp_Fan]

Well, let me ask this: Audio 101 - can harmonics of a specific tone or combination appear on both sides of that fundamental, on the audio spectrum?

IE: A 300Hz tone might have multiple harmonics, say, at 600, 900, etc.

So, can a 40kHz tone generate harmonics BELOW its fundamental frequency, and possibly down into the range of human audibility?

In nature this is very rare. Things have resonant frequencies, and harmonics always appear *above* this frequency as far as I know.

In recorded and reproduced audio, it can happen. There is intermodulation distortion which will create sidebands (not really harmonics, but the same general idea) lower than the fundamentals. There is also a type of distortion (foldover aliasing) that does this too. It is caused by trying to play a tone that's higher than the nyquist frequency in digital audio. Generally it doesn't happen unless something has gone extremely wrong in your system. Like if you sample something that's 30Khz at 44.1khz, you get this problem. This is one of the big reasons why you need filters on ADCs.

Let's be clear these are types of distortion, and if you reproduce ultrasonics and then get these effects, it's a bad thing.

 

[analogsurviver]

Incorrect, but see below.
No, not without knowing what changes were made in production.
Lets get specific rather than using massive and irrelevant generalities. What parameters in a capacitor are no longer available? And "the sound quality" doesn't count! Be specific!
That statement indicates a full lack of understanding of how a DC servo works. I'm not going to explain it, as that's a huge tangent, but for a start: you don't apply it to an input! No need, and no point. So, no chance at all of "blowing up" a cart. Works just fine, and is a very nifty way to eliminate huge caps in the audio path.
It doesn't matter. The entire scenario of needing to reduce gain because of gain/bandwidth product, then having to hit it so hard that you needed a higher output preamp (like what, +/- 50V???) is irrelevant and stupid.
The response above was fully expected, but disappointing. It's typical audiophile: "You haven't hear it, so you don't know. You probably couldn't hear it anyway because you are 1)deaf 2)an idiot 3)your system is far to unsophisticated. Thanks for following suit with expectations, but none of that supports even a single one of your arguments.
There it is again. This time it's the "20-20k minded".

I do. It's exactly what we get now. You've fallen into the "circle of confusion" once again. So you think everything needs to be capable of some RF frequency like 100mHz, but you want everything in the chain to do that. Most mics don't, can't and won't, and are chosen for other far more audible and important parameters. The mics that do are useful for very limited applications. We could digitize up to 100mHz, but we're not going to because there's no evidence it makes any difference above 20kHz or slightly more. No analog recorder will record above 20kHz reliably and consistently, and none at all above 25kHz that don't have some other fatal issues. No speakers can deliver anything to your ears above 25kHz at all, even if stated response is 40kHz, the ultrasonic beam width and air absorption blows it away. The whole argument collapses if you look at any individual segment of the chain, but you want the WHOLE CHAIN to play up to 100mHz!
RIAA EQ doesn't dictate a feedback network, but ok...
There's no such thing as zero-tolerance parts. Even the ones with values printed on them have a tolerance, usually 1%, but possibly lower. But you clearly don't understand either the RIAA curve or what tolerances are necessary.
Well, that may have been true in 1977, but not true today. Easily measured, with non-exotic equipment. But again, you're in the circle of confusion! So, we have a preamp with precise RIAA match to the mDB. Great. But how does it perform with a cartridge on it? Oops. You might measure it if you had a test records recorded with no EQ and a perfectly flat lathe. But then we get into the real problem, the record RIAA eq may not be that good. And since the entire produced response is based on what was heard in a studio with monitors that are not flat to 2mdB, likely not to 1dB, we have another random error built into the system. Then we hit your speakers in your room, and response errors become HUGE because I'm 100% certain you make no attempt to equalize them in the room (an audiophile atrocity). So why on earth would we need or benefit from 2mdB response accuracy anywhere? We cannot possibly achieve it at all, not even close.
Wow. You couldn't BE more wrong! US FM (and most other countries) runs from 88-108mHz! AM radio is between 535-1605kHz! Analog TV channels 2-6 are between 55-83mHz! CB radio is centered around 27mHz (AM and SSB), HAM radio frequencies are all over, from 135 meters (135kHz), up to 50mHz in 13 separate frequency bands!

But...I also think you're misquoting the specs. You've confused mHz with kHz. From memory I recall that the AGI, while pretty wide band, didn't really perform much above 100kHz. That's Kiloherts, not Megahertz. So you're right, it rolls off below most high power RF. But it's far from the only preamp that would do that. For example, the Apt-Holman preamp was -3dB at 150kHz with the ultrasonic filter switched out. I recall the Marantz 7T being good to 100kHz too. So, no big deal there. However, RF proofing is quite another matter.
Wrong again, on several points. In fact, the only thing right in the above paragraph is that a MM cart won't work well into 500pf! But they don't have to. Simply placing a 500pf cap on the front end won't RF proof it at all. Several preamps, including one by Kenwood, and the notable Apt-Holman, solve the RF problem by not using a bipolar transistor at the front end. A bipolar junction acts as a diode detector/demodulator at RF frequencies, becoming the root cause. Eliminate that junction, you don't have a detector.
Again...the actual electronics involved is very different here. What you describe (actually not in the schematics available on-line which show a bare-foot uA749 opamp as the front end!) is just a low pass RF filter. The function of which would be necessary for any preamp operating in a high RFI field.
Oh, yes. Been done. The Apt-Holman was designed within 1 mile of a high power broadcast transmitter! And I have actual hands-on experience with phono preamps and all sorts of other audio gear at both a 20kW FM transmitter site and a 50kW AM transmitter site. There are plenty of other preamps that will work out there, but none (AGI included!) will work without very special attention to installation and grounding.

I thought that thing was special too...in 1977. But time moved on, it's not special today at all, nor has it been for 30 years.

Well, in "Picking capacitors" slew reate may have been mentioned ( don't care to check it, 3 metres away in the library ) - but VERY briefly. Since it was also a thinly disguised advertisement for what later became known as Wonder Cap, there was no warning saying film caps can have VERY large differences in slew rate. From just a few (single digit ) volts per microsecond to aprox 600 ( at least back at the article publishing, but is not much different today ).

Regarding differences made in production; IF that first amp heard at friend's contained "vintage, definitely in limited supply, not to be available again- ever" caps of superiour quality, and the one bought later with whatever best available from the current production, but inferiour to vintage caps - you WOULD mind. That's why manufacturers avoid such situations at all costs - and build with whatever is currentlyproduced and likely to remain available in at least few years ahead. That means an old(er) design may well be, ultimately, superiour sounding for this very reason.

With the miniaturization and advent of SMT /SMD, unfortunately MOST of the best capacitors once used in trough hole variety have been dropped. Simply because they can not withstand the temperatures of the SMD/SMT soldering process - further made worse with the introduction of the lead-free solder ( which requires yet even higher temperature ). Only the most THERMICALLY robust varieties of film capacitors remain available in SMD form - and those are NOT the equal in quality of say polystyrene as dialectric material. And, yes, capacitors DO have different sound quality - even IF the electrical measurements, down to the last DA and whatever electrical parameter ( triboelectrics, anyone ? NO mention in "Picking Capacitors" , THE fatal flaw of teflon caps ... ) are identical. Now try to figure this one out ...

I do know how a DC servo works. And did post its use at the input as the ultimate no-no. Sooo - if your preamp input is directly coupled to a phono cartridge AND there is a catastrophic failure of preamp ( one of the rails blown, latching to whatever remains, blown input semiconductor, etc, etc ) - would you still insist on direct coupling ( best for sonics, no doubt about that ) - or would you insure yourself from blowing up an expensive cart by using ( best quality you can get ) cap ?

No, that scenario is not stupid. Because it does SOUND better than having to insert another gain stage between the preamp input and - call it that way - driver of the "power amp". And yes, although not +-50 V rails preamp, it was close : +- 45V rails. But, I agree it is not the normal way of doing things and incompatible with most other equipment.

You are , of course, free to think this way. But I ( or anybody else ) can not convince to the contrary anybody - online. Only on the real demo floor, with real equipment of high enough quality.

Yes - because a SINGLE "20-20K" minded component in the entire audio chain, from the microphone to the whatever used as the end transducer back to sound, is enough to throw the spanner into soundstage works. Now - go to any studio - and count 'em ...

No, I do not want the whole chain to play up to 100MHz - because I am realist. It is unlikely to happen - ever, but certainly not in my lifetime. However, ANY stage that can be made large bandwidth will contribute to more accurate recording and reproduction. And I have NEVER said that < 20 Hz and > 20 kHz is more important than 20 - 20K, or even more important 100 to 10K band. Only if this core band is OK, would I try to stretch to either of the extremes. As the easiest way to start doing it is in amplifiers, I started with them.

There are microphones made specially for music that go to 100 kHz. There is any number of headphones that go - at least - to 40 kHz; and an inch or so of air can not filter the driver output before it reaches the ear in any significant way. Furthermore, that the attenuation of high frequencies in air is so severe I find hard to believe; my recordings ( never use anything that might be categorized as close miking ) do show sound clerly related to music - all the way to 55 or so kHz ( where the quantization noise of the ADC starts to keep rising ), depending on the instrument(s).

As I type this, I am lidstening to and looking at Voxengo Span display of 192/34 transfer of yesteday's DSD128 recording ( clarinet as lead/solo, violoncello, viola, 3 violins, acoustic bass, accordeon, piano and percussion ) - from the rehearsal, where there was less noise than during the concert ( damn lights, audience was generally VERY quiet ...) I remember a light ( reflector, something like that ) has been switched on during the 12 minute take of rehearsal - and, presto, a constant peak of 19.0 kHz at - 66dB crept in ... and I dread converting DSD to PCM and looking at what has been switched on during the second part of the concert - EVERY damn light in sight dimmed - yuck...!

However, I did not know clarinet has so much output above 20 K - at very least to 40 kHz, during the solo clarinet passage. One learns every day...
And the beat box used by the percussionist goes to 50 kHz on this recording - some at least 5 metres from the mike; clarinet between 2-3 metres, depending on player's position on the stage.

RIAA can be implemented in many circuit configurations - and I have only described the one as used in AGI 511.

I agree there is no such thing as zero tolerance parts - but the best approximation, as most probably used for the AGI, is some Hewlet Packard RLC meter. OK, " measured to be exactly 123456 ohms, with the tolerance from the absolute correct value as specified by XY measurning instrument" - would that satisfy you ? And, yes, the values of resistors in RIAA circuit are printed down to the last ohm - like 23732 ohms and NOT 24 K 1% ( or any other value/tolerance) . Caps are not - because, for those less familiar with film caps, even picking them up with anything more than feather touch can change their value - let alone re-printing them. For this reason, no desoldering and measuring the super precise ( aka close to zero tolerance ) parts from AGI 511 ( or anything else built with even approaching this quality of parts ) is NOT recommended.

I am well aware of the cartridge/preamp interferace problems. And so was/is Mr. Spiegel - 511 works exactly as well as any other correctly designed preamp when it comes to interfacing with real world MM cartridge and its RL characteristics .

No, I did NOT confuse kHz with MHz - correct as written. However, you are right, AGI 511 overall HF limit is set by its line stage, which is approx 100 kHz.

You are correct in stating that whatever has been used in a studio has been way off the accuracy of the RIAA precision of AGI. Then again, it is not AGI's fault to try to strive for the precision - and studio folks being satisfied with sloppiness.

I do not know which schematics of 511 you did find online. It is NOT a single op-amp - in addition, there are ( bipolar?- would have to check it ) transistors used as feed-forward - that's where 100 MHz+ capability of the circuit is coming from, not from the slow IC op-amp.

I agree installation and grounding are paramount in any phono ptreamp. But none I am aware of can reject this much RF garbage as 511. Enough is to see the open box - plenty images online.

I disagree - there still is nothing comparable up to today. The only gripe I have with 511 is the use of that ribbon cable for all the connections - because that adds approx 80 pF of capacitance per each input, unfortunataly sometimes this is too much for some of the best MM phono cartridges - after the cabling from TT to preamp has already contributed at least 100 pF - if not more.

 

[analogsurviver]

Well, let me ask this: Audio 101 - can harmonics of a specific tone or combination appear on both sides of that fundamental, on the audio spectrum?

IE: A 300Hz tone might have multiple harmonics, say, at 600, 900, etc.

So, can a 40kHz tone generate harmonics BELOW its fundamental frequency, and possibly down into the range of human audibility?

Yes - it can. To be precise - THEY can. Say overtones at 41 and at 43 kHz - the difference in this case being 2 kHz - definitely audible.

And will be missing in recording>equipment that goes only to 20 kHz.

 

[analogsurviver]

In nature this is very rare. Things have resonant frequencies, and harmonics always appear *above* this frequency as far as I know.

In recorded and reproduced audio, it can happen. There is intermodulation distortion which will create sidebands (not really harmonics, but the same general idea) lower than the fundamentals. There is also a type of distortion (foldover aliasing) that does this too. It is caused by trying to play a tone that's higher than the nyquist frequency in digital audio. Generally it doesn't happen unless something has gone extremely wrong in your system. Like if you sample something that's 30Khz at 44.1khz, you get this problem. This is one of the big reasons why you need filters on ADCs.

Let's be clear these are types of distortion, and if you reproduce ultrasonics and then get these effects, it's a bad thing.

Here it has to be diferentiated between the natural sound > 20 kHz - AND possible IMD created by electronics itself.

The first type of NATURALLY occuring IMD is welcome - it is, after all, what we are exposed whenever in air and sound foeld, that is to say always.

The second is an enemy - and needs to be avoided. That is why competently performing electronics above 20 kHz are required in the first place.