LU1014D Power JFET DISCONTINUED

[Melon Head]

Sorry if this is already common knowledge.
I just received an email from qspeed the other day saying LU1014D is now discontinued.
So get in fast, cause once they're all gone, they're all gone.
Sorry for the bad news.

 

[Nrik]

Are you sure it is not only Qspeed who has discontinued them?
Some-one should ask Lovoltech or what the producer is called.

Anyway: I am glad I allready got some 😀

 

[h_a]

Qspeed = Lovoltech

The Lu's are already for a rather long time discontinued now, unfortunately.

Have fun, Hannes

 

[Melon Head]

Thats right QSPEED and LOVOLTECH are the same company.

 

[Nrik]

Okay okay - got the message.
Sadly that must be the ultimate end of power JFETs , or does anybody else produce some...?

I bought some recently from www.audioprojekte.de, and they still lists them for 3.60€/piece. So I guess one should be quick if they wanna join the Power JFET club.

I am just really looking forward to getteing time, to make some PassDIY F3/Z9 for my tweeters, and some 100W class AB á lá Yamaha B1 for my woofers.

 

[GRollins]

And no, don't ask...I've been out of the Lovoltech/Qspeed JFET supply business since last year. Yes, I still have some. No, you can't have any. They're the ones I saved so I could build a thing or two.
You can't build a Yamaha B1 or B2 with the Lovoltech/Qspeeds--they're N-ch only. The Yamahas used a complementary output stage. No, Lovoltech/Qspeed never made a P-ch.

Grey

 

[Nrik]

hey Grey

From the Pass Z8 article:
http://www.6moons.com/industryfeatures/zen8/1.gif
...it seems that the Yamaha B1 only is with N-channels.

Eitherway - I know that there is no P-channel, and one would have to make some changes to use the 'old' Sony and Yamaha Power JFET circuits with Lovoltech.
Among other change bias voltage, cascoding for protecting the LU1014s for excessive drain-source voltage and power dissipation=heat, and maybe more changes.

Here is some good inspiration with the old 2SK82:
http://www.est.hi-ho.ne.jp/jack-morita/sub02/sub02.html
http://www.est.hi-ho.ne.jp/jack-morita/sub06/sub06.html

 

[GRollins]

I know I've got the B-2 schematic, and I think I've got the B-1. I'm pretty sure the B-2 is complementary, but I may have slipped a cog mentally. If the B-1 is quasi, then that's going to be a more appropriate topology.
Cascoding isn't an option, it's a must. If you're going to run any amount of bias at all, you're going to be very limited in terms of voltage swing and/or heat dissipation. The only exceptions would be something like headphones or some of the more efficient full-range drivers out there.

Grey

 

[Nrik]

Cascoding isn't an option, it's a must. If you're going to run any amount of bias at all, you're going to be very limited in terms of voltage swing and/or heat dissipation.
Grey

I imagine something like a Class AB quasi with +/- 35Volts supply and the speaker would be a 6 ohm (8 inch bass only driven from 80-700Hz, with active crossover)

Say one cascode both the negative and postive LU like NP did in the F3/Zen, leaving some 3-4 volts over the LU1014 - that would take care of the voltage swing- right?

At maximum voltage swing, the average current thorugh the transistor will theoretically 35/6/2= 2,9ampere and power dissipation 4x2.9=11,6 watts.
(edit: i havent even subtracted the voltage lost in the cascode MOSFET)

I am only in the stage of basic thumbruling-achitecture, but does it sound completely nuts?
I could ofcourse also paralell two LU1014s in each leg.

 

[EchoWars]

Yes, the Yamaha B-1 is quasi-comp, using huge 2SK77 JFET's.

The B-2 is full complementary (but I'm too lazy to dig out the SM to see what the devices were).