I still have at least 450pcs of each J271 and J309 (TO92 case), but no way that I'm going this Beast route .......
What are you think of for those devices?
I have plenty of those kinds of items. It is an illness.
For the J271/J309, the Idss looks similar to the J113. Too much Idss for many of the Pass-style simple circuits unless you are operating at 5V.
If you connect 350 pcs in parallel and use as a follower, noise = 50nV/RtHz / sqrt(350) = 2.7 nV/RtHz. Should be ok, yes?
Last edited:
> = 2.7 nV/RtHz
3.5x higher than a single BF862.
Patrick
The question is if the J270 will have audible noise as a massively parallel power follower. It can be obtained as a production part.
The BF862 appears to be unobtanium
Your point is well taken. LSK170/J74 are sweet parts but qty 2300 of them would be spendy.
No problem if you want to use Jxxx in parallel instead of a power JFET (say LU1014).
But statements like "perfect instead of Toshibas" deserves some fact-checking.
Patrick
"considering all"
makes world of difference,
which means - if you don't have real McCay, who's gonna say no to Moonshine?
we all know what's Real McCay, so what's fuss?
several threads , all resulted in conclusion that there is really no cheap and available substitution, while here context is established - instead of audio parts, playing with switchers etc.
so - either Toshiba or current Linear Systems parts, everything else is and will stay dumpster divers fun
No problem if you want to use Jxxx in parallel instead of a power JFET (say LU1014).
But statements like "perfect instead of Toshibas" deserves some fact-checking.
Patrick
For sure the LU1014 is on the road map. But it will never become a complementary buffer. The LU1014 is a different animal for a different rodeo. I also don't consider the LU1014 just a JFET. The extensive curve tracing I have done tells me it is a static induction transistor so it has a more special place on my list than just a JFET.
I think that there is an appetite in this community for pedestrian JFETs in a highly parallel complementary follower. Nelson's disclosure of 4 different topologies is evidence.
So, I am searching for some candidates.
Here is a new question. Does anyone have noise spec for or an actual measurement of a Sanyo 2SK304?
J270 P-channel JFET 360 pcs tested Vds 5V
Idss ranged from 4.4 mA to 6.1 mA
Average Idss 5.1 mA Total ganged Idss 1.8A
2SK304E N-channel JFET 272 pcs tested Vds 5V
Idss ranged from 2.8mA to 4.6 mA
Average Idss 3.7 mA Total ganged Ids 1.0A
Next step is to create a board that can gang enough parts to make a complementary buffer with 250mA bias.
Idss ranged from 4.4 mA to 6.1 mA
Average Idss 5.1 mA Total ganged Idss 1.8A
2SK304E N-channel JFET 272 pcs tested Vds 5V
Idss ranged from 2.8mA to 4.6 mA
Average Idss 3.7 mA Total ganged Ids 1.0A
Next step is to create a board that can gang enough parts to make a complementary buffer with 250mA bias.
At line levels and higher, this is not an issue at all. If we were talking
about moving coil cartridge levels it would be a big consideration.
Thank you, Nelson. This is great news.
I have decided to do an array of 8 x 10 buffers. The board will be ~ 8" x 10" and bias may get to 300 mA. It will be an asymmetric population N vs P devices. All devices are through hole for simplicity of assembly.
It will be easy to do single ended or balanced like the original Beast. Voltage rails can be anything that the devices can handle.
No degeneration resistors. 1 ohm drain resistors. 1k ohm gate resistors.
I am figuring out how to route the In/Out/V+/V- traces. After 3 drawings, I think I have it figured out.
Plenty of devices to build some boards.