More FET noise measurements (for EUVL)

Just out of curiosity, were there ever depletion MOSFETs made for audio purposes (small signal parts not high power)? I know some odd ducks were made for RF purposes such as 2SK241 and 2SK359 and those have pretty low noise in their intended frequencies but I probably shouldn't expect them to be any good in audio. I'm just curious if audio equivalents existed.

(I have some of the above mentioned ones left over from a past RF project; the datasheets are interesting, reads like a JFET datasheet but they're MOSFET)
 
Fairchild no longer exists.
Does that AN not look like a re-issue of a very old one ?
Low noise without numbers is a very vague definition.

We measured J111s before and they are not bad, but not 2SK170 level.
I do not understand why people only keep looking for 2SK170 when the likes of 2SK369 or 2SK372 can still be found.


Patrick
Hi Patrick,

is there a difference between 2sk372 and 2sk363? Their datasheet seems to be completely identical:confused:
 
I'm just asking some of the OCD types here if this could be of interest.

I got curious today to see what kind of thermal effects might occur when heating up resistors (immersed in a heated solution) and looking at the change in resistance.

One of the results was very surprising (bad).

I measured Panasonic 3W 0.47 Ohm, Vishay CPF 3W 0.47 Ohm, Futaba 0.47 Ohm

I have short 20 second videos I could post in a new thread.

If people are interested I will post, if not, no worries.
 
The datasheet for the device should have tempco stated in ppm / deg C. Measuring the tempco should yield no surprises.

I have visited a primary standards lab. The most important resistors are in a temp controlled oil bath. They are sent to NIST every year to be measured.

They are used with a bridge that is in a rack that costs hundreds of thousands of dollar.

It is a rat hole. Beware.
 
is there a difference between 2sk372 and 2sk363? Their datasheet seems to be completely identical:confused:
I put this together some time ago based on the various data sheets / old Japanese FET Manuals (LN means low noise, GEN is general purpose); to answer your specific question - K363 & K372 are the same die / different package, but for low noise you want K369 & K371.
 

Attachments

  • 20200726.jpg
    20200726.jpg
    311.4 KB · Views: 495
Well, the data sheets for 363, 369, 371 & 372 are all but identical - there are only 2 differences:

1) 363 & 369 are bigger package / higher dissipation than 371 & 372
2) 369 & 371 are designated as "super low noise", while 363 & 372 aren't - I wonder why the engineers at Toshiba wouldn't notice 372 is even more "super low noise" :)

My guess is they are the same chip, but 363 & 372 are not screened for noise, which wouldn't necessarily mean you can't find a good batch there...?

In Scott's measurement, the difference between the 372 and the 369 is like 2 dB at the lower end - that would be well within the typical / max spread as given in the data sheet.
 
Screening for noise in such a mass-produced jfet at the time seems pretty labor consuming though..

Anyway, after looking for 369/363/372/371 in different sourcing companies, the price differ somewhere from 0.9$ all the way to 0.2$ -> Indication for monkey business here. I will try building a simple cookie-can LNA first, then buy them in quantities.
 
post 4852 under
https://www.diyaudio.com/forums/pass-labs/124889-b1-buffer-preamp-486.html#post6746451
shows various threads concerning jFET's.

I am looking for currently available jFET's (and MOSFET's) both for replacement obsolete ones in amplifiers and new diy- so as modification projects.

It should have this features:
1) low input (G-S) capacity
2) good availability (i. e. from currently production)
3) not expensive (like INTERFET parts)
4) low noise (in most cases actually only really important for microphone pre-amps, but less important for line stages and phono RIAA input stages due the still present record surface noise - so I think).

summarized good replacement for hard to find and obsolete jFET's (e. g. from Toshiba, ROHM, NEC, Hitachi and others) - in any case better than the usual known replacements e. g. from LSK series.

I start already this thread, but mainly in order for a 2SK/2SJ replacements in true complementary input differential amps.
Better Audio by UHF/RF jFET Replacement (<5pF) for 2SK389/2SJ109+2SK170/2SJ74

There are no small signal FET's to find, where the input (gate-source) capacitance is low like e. g. AD817 (which have according the datasheet only 1,5pF at both inputs) and appropriate for low frequency applications.

Apparently the following rule exists: the higher the Gm value, the higher the input capacity - go for examples to
IF3601 (750mS 300pF - very expensive)
https://www.interfet.com/jfet-datasheets/jfet-if3601-interfet.r00.pdf
2SK364 (28mS 30pF)
2SK364 pdf, 2SK364 description, 2SK364 datasheets, 2SK364 view ::: ALLDATASHEET :::
2SK389 20mS 25pF
2SK389 pdf, 2SK389 description, 2SK389 datasheets, 2SK389 view ::: ALLDATASHEET :::

Until now I have found only this small range of MOS-/j-FET's (actually not for audio application, but for RF resp. UHF/FM- applications)

CE3512 50-70mS 1pF
https://www.mouser.de/datasheet/2/286/CE3512K2-965091.pdf
BFP405/BPF520F 0,31pF
https://www.infineon.com/dgdl/Infin...N.pdf?fileId=5546d462689a790c01690f03150a390c
https://www.infineon.com/dgdl/Infin...N.pdf?fileId=5546d462689a790c01690f03696a391c
TC2623 35-45mS 0,5pF
https://datasheetspdf.com/pdf-file/787187/UnitedMonolithicSemiconductors/TC2623/1
BF961 (MOSFET dual gate) 12-15mS, 1,6pF
https://pdf1.alldatasheetde.com/datasheet-pdf/view/95965/VISHAY/BF961.html
BFR1105 (MOSFET dual gate) 30mS, 2,2pF
https://www.nxp.com/docs/en/data-sheet/BF1105_R_WR.pdf
BF904 (MOSFET dual gate) 25mS 2,6pF
https://docs.rs-online.com/e4ba/0900766b80b4bc7d.pdf
BF998 (MOSFET dual gate) 24mS 2,1pF
https://www.nxp.com/docs/en/data-sheet/BF998.pdf
BF999 (MOSFET) 20mS 2,5pF
https://www.infineon.com/dgdl/Infin...n.pdf?fileId=db3a30431441fb5d0114936ed6171043
2SK606 5mS, 5-8pF
http://pdf.datasheetcatalog.net/datasheets/panasonic/2SK606.pdf
2N3819 2-6,5mS 8pF
https://pdf1.alldatasheet.com/datasheet-pdf/view/171937/FAIRCHILD/2N3819.html
BF510-513: 5pF, 2,5-4mS
https://www.nxp.com/docs/en/data-sheet/BF510_511_512_513_CNV.pdf
2N5484-5486, BF244, BF245 3-8mS 3-5pF
https://pdf1.alldatasheet.com/datasheet-pdf/view/173154/FAIRCHILD/BF244B.html
https://www.mouser.de/datasheet/2/308/1/MMBF5486_D-2316174.pdf
2SK161 9mS 6pF
https://cdn-reichelt.de/documents/datenblatt/A100/2SK161_TOS.pdf
J308-J309-J310 12-13mS 4-5pF
https://pdf1.alldatasheet.com/datasheet-pdf/view/11649/ONSEMI/J308.html
J304/J305 3-5mS 2-4pF
https://datasheetspdf.com/pdf-file/1115622/Vishay/J305/1
https://datasheetspdf.com/pdf/1125543/InterFET/J304/1
J210-212: 7-12mS 4pF
https://www.mouser.de/datasheet/2/308/j210-1192555.pdf
2N6908-6910 0,1-4mS (very large spread in Gm) 3-5pF
https://datasheetspdf.com/pdf/1315961/Siliconix/2N6908/1
BF256 3mS, input capacity not mentioned (maybe <3pF)
https://pdf1.alldatasheetde.com/datasheet-pdf/view/172151/ONSEMI/BF256A.html
https://pdf1.alldatasheet.com/datasheet-pdf/view/77367/MOTOROLA/BF256C.html

Due the low Gm value I would like to expand BF256, BF961 or BFR1105 by a normal or Sziklai darlington variant with a BjT like BC550C or BC560C (to increase Gm) - go to fig. 3 of the first attached PDF file and fig. 4 of second PDF file.

With regard to the circuit design, the question of the distribution of the currents remains.

So far I've always had the experience that 10-12mA for the Bjt transistor and about 1mA for the FET gave very good results in listening tests both for buffer stages and differential input (LTP) amplifiers (Idss selecting was not necessary in case of additional fitted offset servos).
Concerning noise unfortunately I don't know, whether 1mA for the FET is good (maybe 0,5mA or 2mA provide better results for SNR).

In summary, the following questions remain:

1) which currently available FETs besides the above mentioned are interesting in terms of low input capacitance and low noise (low Gm isn't a disadvantage when using the approaches shown in the attached PDF files - so I think)

2) which idle (operating) current is the best choise for the FET's both in order for lowest noise and in order for best sound performance resp. lowest THD values ?

For small signal BjT's like BC550C or BC560C - go to
https://www.onsemi.com/pdf/datasheet/bc550c-d.pdf
the right idle current for lowest noise is arround 1-3mA and for best sound performance is those value for idle current from my experience ideal, which provide highest value of ft, i. e. around 8-12mA.

Thank you very much for some information.

P.S.:
This new jFET's from On-Semi have relatively low capacity and high Gm at the same time (maybe a special process in manufacturing):
CPH3910 40mS, 6pF
https://www.onsemi.com/pdf/datasheet/ena1965-d.pdf
and a dual type CPH6904 30-40mS, 6pF (maybe a replacement for 2SK389)
https://www.onsemi.com/pdf/datasheet/ena1966-d.pdf
BF862 (45mS 10pF) is obsolete in the meantime
https://www.nxp.com/docs/en/data-sheet/BF862.pdf
 

Attachments

  • sziklai National Semiconductor FET Databook 1977 (1).pdf
    43.9 KB · Views: 147
  • Ultra High Zin Buffer - National Semiconductor FET Databook 1977 (1).pdf
    43.4 KB · Views: 171
Last edited:
  • Like
Reactions: 1 user
I have the "currently ava
post 4852 under
https://www.diyaudio.com/forums/pass-labs/124889-b1-buffer-preamp-486.html#post6746451
shows various threads concerning jFET's.

I am looking for currently available jFET's (and MOSFET's) both for replacement obsolete ones in amplifiers and new diy- so as modification projects.

Due the low Gm value I would like to expand BF256, BF961 or BFR1105 by a normal or Sziklai darlington variant with a BjT like BC550C or BC560C (to increase Gm) - go to fig. 3 of the first attached PDF file and fig. 4 of second PDF file.

With regard to the circuit design, the question of the distribution of the currents remains.

So far I've always had the experience that 10-12mA for the Bjt transistor and about 1mA for the FET gave very good results in listening tests both for buffer stages and differential input (LTP) amplifiers (Idss selecting was not necessary in case of additional fitted offset servos).
Concerning noise unfortunately I don't know, whether 1mA for the FET is good (maybe 0,5mA or 2mA provide better results for SNR).

In summary, the following questions remain:

1) which currently available FETs besides the above mentioned are interesting in terms of low input capacitance and low noise (low Gm isn't a disadvantage when using the approaches shown in the attached PDF files - so I think)

2) which idle (operating) current is the best choise for the FET's both in order for lowest noise and in order for best sound performance resp. lowest THD values ?

For small signal BjT's like BC550C or BC560C - go to
https://www.onsemi.com/pdf/datasheet/bc550c-d.pdf
the right idle current for lowest noise is arround 1-3mA and for best sound performance is those value for idle current from my experience ideal, which provide highest value of ft, i. e. around 8-12mA.

Thank you very much for some information.
I have made one such list!

BC550 and BC560 are high noise small signal cheapistors. It's a continuation of Phillips cheap BC series. They have some use, but that is best when you want a million of those, otherwise, a more hightech transistor with lower parasitic resistances will fill your needs better.

ONsemi has bought Sanyo. So, apart of ONsemi offering the brilliant Motorola transistors, they have the 12A02MH PNP, 15C02MH npn, 50A02CH pnp.

For low noise JFET, you need high gm, low Vgs(off). And large area! That is why the table compares the transistorswith Interfet process, so that you can see what size are we talking about.

Art of Electronics has an input buffer with MMBFJ202 from ONsemi, and it was really good, study the system... they used 4 identical transistor, as it is simpler., I recommend you also study the cascode for exqample as described in the "Designing with Field Effect transistors" book. from 1981. There is a LOT to read! And I also attach one rare research paper that explains a lot.
 

Attachments

  • ONSEMI JFET class sorting simple.pdf
    116.3 KB · Views: 254
  • jfet noises joined picture.png
    jfet noises joined picture.png
    351.4 KB · Views: 228
  • A Low Noise Cascode Amplifier - JRNBS 1987.pdf
    258.3 KB · Views: 219