Pease at AES SF 2006, very short summary

[bobkatz]

The man is very charismatic. He did not do this as an advertisement for new products, he talked about capacitor testing and wire testing and opamp testing and any device testing. Looks like he likes polypropylene or NFO ceramics best for coupling caps, no surprise. He showed how oil/paper caps suck, no surprise. His test jigs are all "analog" with scope horiz and vertical doing a form of transfer function analysis and the opamp to raise the gain up. Audio precision's order of magnitude for THD+n is so high that you have to amplify before putting it into it, no surprise.

To test the opamp he takes advantage of its own noise gain and uses a simple oscilloscope and a triangle wave is best, it produces a rectangular block on the screen and the smaller the block, the lower the distortion+noise of the D.U.T.. When I get a chance to scan his poop sheet on the schematic of how to analyse opamps and caps, but I'd not be surprised if on his website you can also find it.

Only if you need it, his email is rap[at sign]galaxy.nsc.com but first go to www.national.com/rap and search for

LM 4562 Data sheet (this is the new super low noise and distortion audio opamp). His challenge is for you to put 10 in series and see if you can hear a differene. With gain/atten/gain/atten etcetera, lather, rinse, repeat.

LM 4702 (new linear audio power amp/driver, to be followed by some Darlingtons)

AN1490 is the app note.

That's it for now!

 

[kvintus]

Interesting... That LM4562 looks cool. TO-5 case and everything! Too bad it's only +/- 17V.

More things to read, less time to read it...

/Anders

 

[clintrubber]

[quote author="kvintus"]Interesting... That LM4562 looks cool.[/quote]
Here's more about it: http://www.groupdiy.com/index.php?topic=18341&highlight=intriguing

 

[kvintus]

Ok. I missed that one. Should spend more time here and less time with my wife... :green:

/Anders

 

[Kev]

Thanks Bob
and I guess thanks also to the charismatic one

Bob,
have you had time to put your ear to the LM 4562
whether it is just a single one or the 10 in series test ??

 

[bobkatz]

[quote author="Kev"]Thanks Bob
and I guess thanks also to the charismatic one

Bob,
have you had time to put your ear to the LM 4562
whether it is just a single one or the 10 in series test ??[/quote]

THIS IS GOING TO BECOME A LONG POST TALKING ABOUT THD AND SUCH...

No... haven't done it. I'm out of the analog "development loop" at this stage in my career because I think that guys like Fred Forssell and Dave Hill can build far more robust-sounding stuff with their discrete opamps than I can imagine. Once in a while I put together a line amp with a Forssell Opamp. So I'm sitting on the sidelines waiting for someone with more time on their hands to listen to and report on this new and interesting opamp.

My personal take on the amplifier field is that there are two schools of thought, BOTH EQUALLY VALID at this point in audio circuit design. There is the school of thought that vanishingly-low THD is not necessarily good, because whatever residual is left can bite you (e.g. if 5th and 7th harmonic are significantly high relative to the 2nd), so an OPTIMUM amount of THD is ideal. Forssell may not voice that philosophy out loud (I'm not sure) but I believe that his components reflect that philosophy, with medium amounts of open loop gain and "reasonably" low distortion open loop.

I also believe that the more we learn about masking and how it works, the more we will learn how to better manipulate our signals. In my career I separate components into three classes: 1) Transparent (no character but no down side), 2) good character and 3) bad character.

I also know that by judiciously using components that have some "good character" I can mask or partially mask previous stages in a processing chain that had "bad character".

Without having listened, I don't know, but I surmise that this new LM4562 may turn out to be the best current opamp to use when we have to construct a multi-stage system, because even when distortion is vanishingly low, it can add up. This will allow us to choose how much "character" we need to add (if any) without being forced to use too much!

I once built a mixer with only one stage, using an OPA (forgot the number) and the mixer sounded harsh to me. It's sitting in the garage, unused. This opamp obviously did not make a good line driver, or I didn't implement or bias it properly. As an amateur designer who had a lot more time in my youth, I regret that I don't have the time anymore to go back to the "drawing board" and find out if I could make it better myself. So now I sit on the sidelines and depend on my friends Fred and Dave and Greg (Pendulum) and Hutch a few others to build real nice sounding pieces of gear that are either totally neutral and transparent or very pleasant sounding, or add a character that I need.

I like doing mods, though.... when I have time. But at this stage in my career(s) I have no more time to build stuff from scratch. Sometimes a real-talented intern comes along and I direct and help them to build a nice piece from scratch. We're working on an all-digital sample rate converter, but it's based on the 8420 eval board from Cirrus.

Fortunately, I still have a very strong grounding in circuit design, so I do a lot of helping out, beta testing, acting as a critical listener for some of these designers and we have a wonderful, symbiotic relationship.

BK

 

[Kev]

[quote author="bobkatz"]No... haven't done it. I'm out of the analog "development loop" at this stage in my career because I think that guys like Fred Forssell and Dave Hill can ...
So I'm sitting on the sidelines waiting for someone with more time on their hands ...
... I sit on the sidelines and depend on my friends Fred and Dave and Greg (Pendulum) and Hutch a few others to build real nice sounding pieces of gear that
... I like doing mods, though.... when I have time. But at this stage in my career(s) I have no more time to build stuff from scratch. Sometimes a real-talented intern comes along and I direct and help them to build a nice piece from scratch. We're working on an all-digital sample rate converter, but it's based on the 8420 eval board from Cirrus.

Fortunately, I still have a very strong grounding in circuit design, so I do a lot of helping out, beta testing, acting as a critical listener for some of these designers and we have a wonderful, symbiotic relationship.[/quote]
Thank Bob

I'm right there with you and also feel I just haven't the time to painstakingly go through all this stuff in detail.
I tend to lean on Joe M. as he is deep in this stuff at his time of his career, although a baby may change his time management in the future.

However
if someone does bring a prototype past you for a critical listen
do come back here to tell us.
AS I feel your ear is better than mine AND I have great trouble in deciding what is better ... once a piece of equipment is working to a good or excellent level I tend to glaze over when someone asks me what is better ....
mostly I just hear different once I stop hearing problems or mistakes
...
:roll:
if you get my drift

 

[SSLtech]

[quote author="mediatechnology"]We replaced a number of 4/6K consoles with Signetics 5534s after a bad run of Raytheons.[/quote]
Wayne,

I just recently came across a dated FedEx tube with a load of signetics 5534's in it round at our (and Bob's!) mutual friend's old stomping ground studio.

I told the new owner of the 5534's (the studio recently changed hands) not to give them away, and relayed the story of the chip-swapout tests.

Good, logical and repeatable test.

I like that kind of thinking.

Keith