Addressing John Curl's concerns on low noise designs

Lumba Ogir · 2009-05-23 3:51 pm

syn08,

I don't see any reason why a low noise device should necessary be high current capable.

High current gain lowers the base shot noise. High Ic reduces rB, thus the thermal noise.

Edmond,
right?

Sigurd Ruschkow · 2009-05-23 3:53 pm

SYn08,
thanks for sharing your noise measurements. The noise from the 2240 was higher than I expected, and the noise from the 2SC2547 was lower than I expected.

Very interesting to see that one of my favorite BJTs, the 3601, has such low noise! What a BJT!

Would be interesting to see you measure some classic low noise BJTs like the 2SD786/SB737 and SC3329/SA1316 and also the BJTs from THAT Corp.

Sigurd

Terry Demol · 2009-05-23 4:59 pm

Sigurd Ruschkow said:
SYn08,
thanks for sharing your noise measurements. The noise from the 2240 was higher than I expected, and the noise from the 2SC2547 was lower than I expected.

Very interesting to see that one of my favorite BJTs, the 3601, has such low noise! What a BJT!

Would be interesting to see you measure some classic low noise BJTs like the 2SD786/SB737 and SC3329/SA1316 and also the BJTs from THAT Corp.

Sigurd

I would be interested in MAT01. The data sheet is all over the place
and really needs some attention. It looks like all the noise specs
were done at 300uA or less.

T

PH104 · 2009-05-23 5:18 pm

I have a couple MAT02s and a small collection of the 2SD786/SB737 plus some other medium power Toshibas and ROHMs I will try to measure using syn08's setup if I find some time next week. I have some of the A1407/C3601's and will use them to benchmark my measurements against syn08's.

Good info - thanks syn08. Some useful stuff popping up here. The 4PCB and bdent links plus transistor info.

anatech · 2009-05-23 9:49 pm

Hi John,

Thanks SYN08, I think that is the first real info you have ever posted in my presence.

Wouldn't a straight "thank you" do, no qualifiers? Worded as you have, that is not really a compliment at all.

Let's stay away from personal attacks and comments. Not very pretty.

-Chris

anatech · 2009-05-23 9:52 pm

Hi Syn08,
Thank you for all the excellent information you tend to share with us. Especially the experimenting you do that backs up your comments.

I've also found your results interesting.

-Chris

syn08 · 2009-05-23 11:58 pm

HPS 3.1 Jung regulators ready and working beautifully.

Now, on to the signal part...

syn08 · 2009-05-24 12:02 am

2SD786/SB737 and SC3329/SA1316 and also the BJTs from THAT Corp.

MAT01, MAT02
[/B]

Unfortunately I have none of these devices in my hands. In fact, I don't have anything else that would have chances to qualify as low noise.

Terry Demol · 2009-05-24 1:08 am

PH104 said:
I have a couple MAT02s and a small collection of the 2SD786/SB737 plus some other medium power Toshibas and ROHMs I will try to measure using syn08's setup if I find some time next week. I have some of the A1407/C3601's and will use them to benchmark my measurements against syn08's.

Good info - thanks syn08. Some useful stuff popping up here. The 4PCB and bdent links plus transistor info.

MAT02 is well documented and at a broad range of currents, no need
to bother.

T

syn08 · 2009-05-24 4:01 am

You guys are amazing

As soon as I posted the Hitachi devices information and a dealer in the NA (but not before I sent my own order) www.bdent.com (to which I am in no way associated) ran out of 2SA1084 and raised the price for future deliveries to absurd levels ($3.57 from $0.57). It happened before, but now was faster than ever.

Lumba Ogir · 2009-05-24 4:35 pm

John,
did you say something?

Sigurd Ruschkow · 2009-05-24 6:01 pm

Yeah, I noted that, too. I tried to order 500 but they only had 6 pcs left....

Sigurd

syn08 said:
You guys are amazing

As soon as I posted the Hitachi devices information and a dealer in the NA (but not before I sent my own order) www.bdent.com (to which I am in no way associated) ran out of 2SA1084 and raised the price for future deliveries to absurd levels ($3.57 from $0.57). It happened before, but now was faster than ever.

john curl · 2009-05-24 7:17 pm

Perhaps I should have said nothing about buying up these exceptional parts, as Syno8's excellent measurements and my experience with these Hitachi devices found them to be a rare example among good parts. However, I did think that they were not easily available, but sometimes a distributor might some existing stock on hand, perhaps years old, and priced accordingly.
What usually happens is that the distributer realizes that these parts have become popular, and that it is time to reorder, and then is surprised by the manufacturer's cost/availability response.
Then, prices change, because they have to. This happens a lot in this business.

lineup · 2009-05-24 8:09 pm

john curl said:
Perhaps I should have said nothing about buying up these exceptional parts,
as Syno8's excellent measurements and my experience with these Hitachi devices found them to be a rare example among good parts.

However, I did think that they were not easily available, but sometimes a distributor might some existing stock on hand, perhaps years old, and priced accordingly.

What usually happens is that the distributer realizes that these parts have become popular,
and that it is time to reorder, and then is surprised by the manufacturer's cost/availability response.
Then, prices change, because they have to.
This happens a lot in this business.

If we are lucky
another dealer, making 2 of them at least,
takes on the service to supply.

Comparer with Microsoft - Linux, Firefox - Internet Explorer, INTEL - AMD.
Good helthy business is not promoted by monopoly.
In Europe we have very strict laws against this ugly thing.

Now is the chance for these to do some good service to audio community:
http://ampslab.com/components.htm
http://www.tech-diy.com/
http://www.aussieamplifiers.com/
http://www.audiosector.com/
.. and while doing thier good duty, make a few bugs for themselves.
... cause:

when John Curl reveals he prefers some good devices
they can not be bad
especially when, for once, syn08 is on the same track!
These 2 guys most often know what they are talking.
At least for low noise .. if this is so important for audio listening??[/b]

Please all dealers out there.
Do not let us resort to one of those GROUP BUYS, again.
Some of them end of in endless waiting and even some crooks have made false money ..

/lineup - this time a post with a clear & useful message!!
hehe - for once

Sigurd Ruschkow said:
SYn08,
thanks for sharing your noise measurements. The noise from the 2240 was higher than I expected, and the noise from the 2SC2547 was lower than I expected.

Very interesting to see that one of my favorite BJTs, the 3601, has such low noise! What a BJT!

Would be interesting to see you measure some classic low noise BJTs like the 2SD786/SB737 and SC3329/SA1316 and also the BJTs from THAT Corp.

PH104 said:
I have a couple MAT02s and a small collection of the 2SD786/SB737 plus some other medium power Toshibas and ROHMs I will try to measure using syn08's setup if I find some time next week. I have some of the A1407/C3601's and will use them to benchmark my measurements against syn08's.

Good info - thanks syn08. Some useful stuff popping up here. The 4PCB and bdent links plus transistor info.

Terry Demol [/i][B] MAT02 is well documented and at a broad range of currents said:
You guys are amazing

As soon as I posted the Hitachi devices information and a dealer in the NA (but not before I sent my own order) www.bdent.com (to which I am in no way associated) ran out of 2SA1084 and raised the price for future deliveries to absurd levels ($3.57 from $0.57). It happened before, but now was faster than ever.

syn08 · 2009-05-24 8:22 pm

Done

And works like a charm, 0.3nV/rtHz.

Now swapping opamps, measuring, etc...

syn08 · 2009-05-24 9:05 pm

First measurement

Gain is 10,000 (no RIAA caps installed yet), probe is 1:10 so equivalent scale factor is 1,000. Opamps are LT1128.

Under 0.3nV/rtHz. That's pretty good

syn08 · 2009-05-24 9:23 pm

Low frequency noise.

Probe is now 1:1, scale factor is 10,000 so vert scale is in uV/rtHz at the output, of course.

john curl · 2009-05-25 12:35 am

Looks as good as anything I have seen, Syn08. Now, I will have to catch up.

john curl · 2009-05-25 5:36 am

Ok, I caught up.

syn08 · 2009-05-25 11:46 am

john curl said:
Ok, I caught up.

Not enough time to even handmatch (TM) your JFETs.

Addressing John Curl's concerns on low noise designs

Member

Member

Member

Member

Administrator

Administrator

R.I.P.

Attachments

R.I.P.

Member

R.I.P.

Member

Member

Member

Member

R.I.P.

Attachments

R.I.P.

Attachments

R.I.P.

Attachments

Member

Member

R.I.P.