Benchmark Mic Pre Schematic from 1984

[bcarso]

[quote author="Samuel Groner"]IIRC the LM194 was just the better spec version of the LM394 and later renamed to a LM394-whathever-postfix.

Samuel[/quote]

National's convention was that the better parts with guaranteed specs over a wider temp range started with a "1" after the "LM", then 2 and 3 as things got looser.

 

[Samuel Groner]

Interesting, the temperature control--didn't noticed that at first. Datasheet for the part: www.datasheetarchive.com/pdf/3509834.pdf

Samuel

 

[bcarso]

[quote author="Samuel Groner"]Interesting, the temperature control--didn't noticed that at first. Datasheet for the part: www.datasheetarchive.com/pdf/3509834.pdf

Samuel[/quote]

A later part from National (of course the uA726 was Fairchild) temp-controlled a zener, and came in a TO46 with a plastic enclosure around it: LM199/299/399. I used it for one of the systems within the UCLA Reticon spectrometer, although it wasn't quite low enough 1/f for the main system reference.

I think maybe National also did a temp-controlled diff pair of their own (?)

 

[Guest]

[quote author="mediatechnology"]

I'm starting to get the feeling that neither Cohen nor Wurcer cited their source for this because no one really knew where it came from. SO they used words like "new" and "alternative." Thanks to our crack research department - no not you CJ - we have Samuel to thank for the oldest example. Way cool.[/quote]

Or they did not know that somebody already did the same thing. It happens very often. As I said, I did the same in 1979 for a microwave alarm system coming to such a solution logically thinking. I needed to put everything inside of a hybrid IC so tried to eliminate all needs for electrolytics and trimpots.

 

[magicchord]

[quote author="jdbakker"]...Pretty much Rod Elliot's Project 66:

discussed here by PRR et al. It's also used in at least one Behringer product (and likely several more; I suspect this is their 'IMP')...[/quote]

This looks similar to a design I've been kicking around:

In this one, the opamps form a superbal circuit and the input stage is cap-coupled to the zero-ohm nodes.

 

[JohnRoberts]

I am not surprised that the topology in instrumentation applications way predates the wide use in mic preamps.

Not to be repetitive but IMO the low noise discrete devices were the threshold event to make this older topology finally make sense vs. transformers in mic preamp application. The ROHM parts (737,786) were cost effective and quieter than LM394s. I had used 394s in a handful of console DOAs but never considered them practical for mic preamps.

As I have also mentioned previously there are some subtle and perhaps unexpected interactions between inputs using that instrumentation topology wrt current noise, but this too may have been well known to the ancients.

I see no justification to remotely link my name to this topology. IMO the more creative work in that published phono preamp was how I managed the RIAA EQ, but that's a different subject.

I move that we stop calling this "Cohen". I privately call it something like "hybrid instrumentation amp". Hybrid for the mixture of discrete and IC parts, but will gladly defer to any better suggestions.

JR

 

[Guest]

[quote author="magicchord"]
In this one, the opamps form a superbal circuit and the input stage is cap-coupled to the zero-ohm nodes.[/quote]
Now, this one is more symmetrical! Both nodes are zero-ohm! :thumb:

 

[Samuel Groner]

A design I've been kicking around.

I might suggest to use a complementary implementation (i.e. PNPs as input); 2N4403 has considerably lower voltage noise than 2N4401.

Samuel

 

[bcarso]

[quote author="Wavebourn"][quote author="magicchord"]
In this one, the opamps form a superbal circuit and the input stage is cap-coupled to the zero-ohm nodes.[/quote]
Now, this one is more symmetrical! Both nodes are zero-ohm! :thumb:[/quote]

They are not zero-ohm nodes, exactly, at least in the way one might at first assume. I had hoped, when I looked at this circuit a while back, that they might be, as it would eliminate common-mode distortion.

Maybe another thread to analyze and establish priority on SuperBal? (here he suppresses an evil cackle :twisted: )

 

[Guest]

[quote author="bcarso"][quote author="Wavebourn"][quote author="magicchord"]
In this one, the opamps form a superbal circuit and the input stage is cap-coupled to the zero-ohm nodes.[/quote]
Now, this one is more symmetrical! Both nodes are zero-ohm! :thumb:[/quote]

They are not zero-ohm nodes, exactly, at least in the way one might at first assume. I had hoped, when I looked at this circuit a while back, that they might be, as it would eliminate common-mode distortion.

Maybe another thread to analyze and establish priority on SuperBal? (here he suppresses an evil cackle :twisted: )[/quote]

Yes, I was wrong believing in what was written, until looked at the schemo...
One more opamp is needed.

 

[magicchord]

That's why I like to post here every once in a while. I learn quite a lot from your picking apart my circuits and calling me on my ignorance.

Using PNPs as input devices is something I'll definitely try.

 

[bcarso]

[quote author="magicchord"]
Using PNPs as input devices is something I'll definitely try.[/quote]

Other good parts (E-C-B pinout) are the 2SA1015 (KSA1015) and 2SA970. Get the highest beta brackets available. They have comparable rbb' to the 4403 and as well higher and flatter beta, typically.

The super-low rbb' 2SA1316 is hard to get these days although not impossibly so.

 

[JohnRoberts]

[quote author="bcarso"]

Maybe another thread to analyze and establish priority on SuperBal? (here he suppresses an evil cackle :twisted: )[/quote]


:grin:

good one..

JR

 

[Jim Zuehsow]

Not to be repetitive but IMO the low noise discrete devices were the threshold event to make this older topology finally make sense vs. transformers in mic preamp application. The ROHM parts (737,786) were cost effective and quieter than LM394s.

I don't know John, If the 394 is good enough for Hardy, it's good enough for me! Oh, and my old Harrison 4040 has 12 channels of the standard Jensen mic pres, but all the others are the TransAmp LZ retrofit, and I still love their sound.

 

[Samuel Groner]

If the LM394 is good enough for Hardy, it's good enough for me!

That attitude is lethal for progress. I always expect to know it better... :green:

Samuel

 

[JohnRoberts]

The LM394 at 1nV/rt Hz is respectable and IIRC about the noise level of the IC mic pre from THAT corp, that plenty of people find acceptable.

When I prototyped up a mic preamp using the LM394 back in the day, I didn't find it equivalent to the Transamp I was already using in place of transformer based preamps in consoles at the time. I don't recall the part number of the devices used inside the Transamp but it was a GP medium power transistor (2Nxxxx) and hand selected for noise. This was decades ago so I don't recall specifics of my decision, just the general sense.

The John Hardy preamp I'm aware of uses a transformer input so the ein of the active circuitry is not a critical performance parameter. IIRC he uses the very good for it's day, and still respectable discrete opamp designed by Deane Jensen (RIP). Using a LM394 input stage is not shabby for noise level even compared to modern super opamps, but perhaps overkill when an input transformer has already done the heavy lifting.

In reality a 1nV/rt Hz front end is probably OK in the context of 200 ohm source impedance and typical room ambient noise, but I really liked using the MC head amp parts that were perhaps better than needed, and much cheaper than the LM394. While I am not very rigorous about tracking what everybody else uses, I think I've seen LM394 front ends, where the parts were paralleled to drop the ein lower. This would surely make the part lower noise voltage except you have just doubled the cost of an already pricey part (Note: the Transamp paralleled input devices too but they were relatively inexpensive transistors before you add all the human handling).

I am very much about value engineering and not putting cost into a product that the customer doesn't appreciate, but I used the very low noise (2SD786) in a lot of entry level Peavey mixers for years before they were obsoleted. I may sound overly focussed on cost but mic preamps inside consoles are often a 24x or more cost factor, so individual parts costs add up. I put the 786 based pre in small mixers retailing as cheap as $299 (Unity 300). I could never match that sonic performance with a cheap transformer and/or LM394 for anywhere close to that price.

JR

 

[JohnRoberts]

[quote author="mediatechnology"]Anyone have this?

The so-called current-feedback operational amplifier technological breakthrough or engineering curiosity?

Authors: Bowers, D.F.

Description: Circuits and Systems, 1993.,
ISCAS '93, 1993 IEEE International Symposium
on Start Page: 1054 End Page: 1057 ISSN:
ISBN: 0-7803-1281-3 Volume: Issue:[/quote]

Close,, I stopped taking the journals years ago when I found I was never even looking at them. The newest Circuits and systems I could find on my bookshelf was '91, so we just missed by 2 years.

JR

 

[Gus]

Great, a good thread with parts deleted why? I read this thread and am interested in the history of things. Good stuff in this thread.

What happened?

 

[JohnSiau]

This is John Siau, Vice President of Benchmark Media Systems, Inc.

Here is the schematic of the Benchmark Mic Pre, hand-drawn by Allen H. Burdick March 1, 1984.  This schematic had been posted at the begining of this thread, but was later deleted.  Benchmark did not request the removal of this schematic.  Fortunately, I made a copy of it berfore it was deleted.  I am reposting it due to the multiple requests on this thread.  Benchmark would like the schematic to be available to members of this forum.

All Benchmark Mic Preamplifiers still follow a similar topolgy.  We have made a number of improvements over the years.  These include better transistors, better opamps, lower impedances, DC servos, improved protection circuits, and improved CMRR trim circuits.  Nevertheless, the schematic as it apeared in 1984 is sufficient to build an outstanding Mic Preamp.

Our newer designs have lower EIN, and maintain a low EIN over a wide range of gains.  Bandwidth has also been extended to 500 kHz.

[Thanks John!  I've added to our "Technical Documents" section - Ethan]

Here is the link:
http://www.groupdiy.com/index.php?action=dlattach;topic=46424.0;attach=10299

 

[dissonantstring]

hi john,
first of all, welcome to the forum.  very pleased to have you here and hopefully you'll find enough interest to continue to add to this forum. 
secondly, thank you for the generous re-post of the schematic and for the support of benchmark media systems, inc. for donating the schematic to the diy community.  i always feel so glad to see support like this from established companies who make well-known and well-respected products.  i have so much respect for you and your company for the supportive culture you are perpetuating. 
cheers!
-grant