Benchmark Mic Pre Schematic from 1984

Thanks for sharing, Wayne!!

I met him at 2001 AES... Smart guy.

save those 737s, they stopped making them a few years ago.. AFAIK nothing that low noise V anymore, but they were almost lower than needed for 150-200 ohms. I have about 100 pcs of 737-S in my stash for my limited edition "halucination series" pre.

JR

[quote author="JohnRoberts"]save those 737s, they stopped making them a few years ago.. AFAIK nothing that low noise V anymore, but they were almost lower than needed for 150-200 ohms. I have about 100 pcs of 737-S in my stash for my limited edition "halucination series" pre.

JR[/quote]

In principle the Toshiba 2SA1316's are even a bit lower rbb', and last I looked still in production. I have breadboarded with them but not done definitive measurements.

[quote author="bcarso"]In principle the Toshiba 2SA1316's are even a bit lower rbb'[/quote]
Out of interest I dl'd both datasheets and (at least as stated there) I saw they're both 2 Ohm typ. Is there more to it ?

Regards,

Peter

What actually is the purpose of a Common mode choke here. I can not really figure out what sort of purpose these things serve on the input of a mic pre.

The 2SA1316 is a newer part, and looks quite good. When dealing with extreme low noise these parts will be more similar than different. The 737 may spec .55nV rt Hz, vs. the 1316's .6 nV rt Hz, but I doubt that's a very audible (if real) difference. OTOH the graded beta of the 1316 at 350-700 'BL' is lightly better than 737's 270-560 'S' grade.

I suspect there's bigger differences from how you apply these parts (current density et al) than due to differences between the parts themselves.

I wouldn't throw away my 737s to use these, but they look like a good replacement especially now that 737s are obsolete. My old employer who was using 2SD786 changed over to this series.

JR

[quote author="adamasd"]What actually is the purpose of a Common mode choke here. I can not really figure out what sort of purpose these things serve on the input of a mic pre.[/quote]

RFI

Ahhhh!

It was to obvious, over thinking it.

thanks,
adam

OTOH the graded beta of the 1316 at 350-700 'BL' is lightly better than 737's 270-560 'S' grade.

Click to expand...

Unfortunately you'll have a hard time shopping the BL grade--so far I've found them to be impossible to source at least in DIY quantities.

According to www.beis.de/Elektronik/LNPreAmp/LNPreAmp.html (and other sources) some power parts might be even better with respect to voltage noise--probably at the cost of lower hFE, increased 1/f noise and wider parameter spread.

Samuel

[quote author="Samuel Groner"]

OTOH the graded beta of the 1316 at 350-700 'BL' is lightly better than 737's 270-560 'S' grade.

Click to expand...

Unfortunately you'll have a hard time shopping the BL grade--so far I've found them to be impossible to source at least in DIY quantities.

According to www.beis.de/Elektronik/LNPreAmp/LNPreAmp.html (and other sources) some power parts might be even better with respect to voltage noise--probably at the cost of lower hFE, increased 1/f noise and wider parameter spread.

Samuel[/quote]

Well I've got 'S's in my stash but I used to buy them in decent quantity. In the beginning I was one of the few people in the US buying them (I read about them in an IEEE journal). ROHM thanked me for the publicity when I used them in a MC phono pre in early '80s published in Popular Electronics.

Yes, power devices will also have low Rbb (to handle high current), but typically not very high beta etc. Modern processes may be better for 1/F noise. In the bad old days you had to select though lots of devices for decent noise and could have plenty of useless parts.

For truly low source impedance (like MC 10-20 ohms) maybe power devices could eek out slightly lower noise but for 150-200 ohm microphones, these parts are probably the right choice.


JR

[quote author="clintrubber"][quote author="bcarso"]In principle the Toshiba 2SA1316's are even a bit lower rbb'[/quote]
Out of interest I dl'd both datasheets and (at least as stated there) I saw they're both 2 Ohm typ. Is there more to it ?

Regards,

Peter[/quote]

Maybe I had seen a max spec on the 737 and remembered that. I think actually it was recollection of Horowitz and Hill saying 4 ohms but I don't have that at my fingertips right now.

And then to get crazy one can buy a bunch and select close enough matches and parallel. I went in with a friend who got quite a few of these as well as the NPN complement the 2SC3329. I looked at a few out of the batch and Vbe's were tightly distributed. I think we had no difficulty getting the highest beta sort parts.

Amusingly he didn't really understand current noise, and said that it was high for the parts he tried. It was about as expected when I got the numbers.

2SD786 has 4 ohm typical--perhaps you remember that figure.

Samuel

[quote author="Samuel Groner"]2SD786 has 4 ohm typical--perhaps you remember that figure.

Samuel[/quote]

That's it, thanks---that must be the one discussed in H&H.

Yup... When I first started using the ROHM part from a blank sheet of paper, I went with the PNP due to lower noise... When I went to work at PV the NPN complement was already in their system so I used it (difference is too small to justify bringing in a different part number).

FWIW this morning looking a 737 data sheet (in Japanese) it notes the different Rbb but suggests a similar noise voltage for 786 and 737 at 10mA, which is different than my recollection but, at this point it's just history so academic and I don't read Japanese.

JR

[quote author="bcarso"][quote author="clintrubber"][quote author="bcarso"]In principle the Toshiba 2SA1316's are even a bit lower rbb'[/quote]
Out of interest I dl'd both datasheets and (at least as stated there) I saw they're both 2 Ohm typ. Is there more to it ?

Regards,

Peter[/quote]

Maybe I had seen a max spec on the 737 and remembered that. I think actually it was recollection of Horowitz and Hill saying 4 ohms but I don't have that at my fingertips right now.[/quote]

Cool, and as you know it wan't meant as nitpicking but just to keep the discussion going (not that that was necessary though :wink: )

Regards,

Peter

I think I recall the mention of a Teledyne Crystallonix ultra low noise JFET. Something like .7nV rt/Hz in the same IEEE paper where I read about the 737/786 bipolar parts... They BTW were originally made by some small Japanese company (Toyo something) that ROHM bought.

I never got my hands on the JFET, pretty expensive and not consumer oriented stuff. I later found some 1 nV toshiba JFETs (2SK117) that were more than adequate to keep me happy for higher Z applications.

JR

Yup, I don't see any anti-zener diodes,, low noise transistors can degrade if phantom or turn on/off transients cause the b-e junctions to zener. Easy to protect against with simple clamp e-b.

JR

Easy to protect against with simple clamp e-b.

Click to expand...

Which is built-in in the LM394, at least for the TO-78 package.

Samuel

Maybe not. These people claim to have 27,000.

http://www.digitroncorp.com/part/2SB737

Click to expand...

They also have 7,500 2SD786 in stock.....

Peter