Cool- here's my issue: I need low noise pFET pairs with matched Idss of 15-20mA for a project I've been trying to design. They have to be available to anyone building the project. Selected J74s in the right range would have been ideal but the available ones have a high percentage of fakes, and LS doesn't seem to be anywhere near production. Suggestions?
Go single ended JFET input. Very easy to get high loop gains ( or not if you don't want that), high slew rates and ultra low distortion. You can set the output stage up to run heavily into class A with all the attendant benefits. It seems to me that this fully balanced thing for small signal stages is totally overrated - its just not a prerequisite for top notch sound. LS LSK389C make a great input pair.
Unfortunately, that won't work for this application. I absolutely need p-channel, unless you know about some positronic tubes. I prototyped it with parts on hand, and got stellar performance, but can't publish the design unless others can build it as well.
LS finally in production of j74
Not cheap at $6.37 each
Volume production in less than 12 weeks.
6 different idss levels
Not cheap at $6.37 each
Volume production in less than 12 weeks.
6 different idss levels
Wow a new element discovered
LSj74 - gone from unobtanium to unaffordium
I have Toshiba j74 I better use or my kids will throw them in the dumpster when I'm gone
LSj74 - gone from unobtanium to unaffordium
I have Toshiba j74 I better use or my kids will throw them in the dumpster when I'm gone
Last edited:
How low is low? These J174-177 look like they could be good for something. Noise is too high for a phono stage, but should be OK for line level. (or am I being a cretin again)
Really.
If these LS parts turn out to be real, they're absolutely viable, given that the tube part of the circuit is a lot more expensive than the FETs (D3a goes for ~$15 each these days). But LS has pulled this one before, so I won't believe it until I actually can order parts, receive them, and find that they actually do what's claimed.
If these LS parts turn out to be real, they're absolutely viable, given that the tube part of the circuit is a lot more expensive than the FETs (D3a goes for ~$15 each these days). But LS has pulled this one before, so I won't believe it until I actually can order parts, receive them, and find that they actually do what's claimed.
Godfrey, you made a good choice, but I suspect the noise is a little too much for SY, suspecting that he wants to use them as a self biasing follower with a tube as a 'cascode', but then I am never completely sure of SY's intentions.
I wish to say that I am almost shocked that these 'exotic' complementary jfet pairs are so hard to find, and NOT just with amateurs.
With regard to amateurs, we might have to go back to the J175-j113 combination, probably from LSI, but maybe many other sources, like Fairchild.
I found, and started to use these devices 40 years ago, yes, in 1973, when I was working with the GD, and passed them on to Mark Levinson who made the JC-2 with them.
This part-pair is pretty good, all around. They are similar sized die, which means that the N always has more Gm than the P, but it doesn't matter that much.
You have to sort them for Idss, but the sorting can be made relatively easy with just a 9V battery, and a simple voltmeter and a 100 ohm resistor.
I lived with these parts, almost exclusively, for 5 years. You can too!
Now what about the alternative, of using all N channel parts?
That's OK too.
It would be interesting what complementary transistors are still easily available?
Personally, I would still choose an N channel pair (don't tell me that you can't find one) and then use complementary bipolar devices for the following stages. Did it for a decade, myself.
Many here tell me that it is virtually 'impossible' to get good parts, so I would appreciate what your price expectations are, and what you have found?
For example, my boss at Parasound located a number of Pch 2SK109 devices in Europe at $20 ea, and he laughed it off. Is that too much for an amateur to pay? 'Have you ever tried to buy a replacement part for an old Porsche? '-)
I wish to say that I am almost shocked that these 'exotic' complementary jfet pairs are so hard to find, and NOT just with amateurs.
With regard to amateurs, we might have to go back to the J175-j113 combination, probably from LSI, but maybe many other sources, like Fairchild.
I found, and started to use these devices 40 years ago, yes, in 1973, when I was working with the GD, and passed them on to Mark Levinson who made the JC-2 with them.
This part-pair is pretty good, all around. They are similar sized die, which means that the N always has more Gm than the P, but it doesn't matter that much.
You have to sort them for Idss, but the sorting can be made relatively easy with just a 9V battery, and a simple voltmeter and a 100 ohm resistor.
I lived with these parts, almost exclusively, for 5 years. You can too!
Now what about the alternative, of using all N channel parts?
That's OK too.
It would be interesting what complementary transistors are still easily available?
Personally, I would still choose an N channel pair (don't tell me that you can't find one) and then use complementary bipolar devices for the following stages. Did it for a decade, myself.
Many here tell me that it is virtually 'impossible' to get good parts, so I would appreciate what your price expectations are, and what you have found?
For example, my boss at Parasound located a number of Pch 2SK109 devices in Europe at $20 ea, and he laughed it off. Is that too much for an amateur to pay? 'Have you ever tried to buy a replacement part for an old Porsche? '-)
I have heard they are still having production problems. Seems Jfet maunfacturing is a bit of an art.
Whoops, yes, "complementary" is the proper spelling. My my mistake. Thanks for catching it.
se
Yes, buzzforb, LSI is still limping along with making the Toshiba 2SK74 for the masses. It has been a big headache for them, and 'promises' have been made in the past, but you can't just make a similar part, it has to be interchangable, I suppose.
Here is an example of an ALTERNATE topology that doesn't have pchannel jfets, that is akin to what Scott Wurcer used for his AD797.
I am very concerned about the availability of P ch jfets. Somebody, somewhere is grabbing ALL of them up. Please input here as to your problems and what you have done to find these pch parts.
I am very concerned about the availability of P ch jfets. Somebody, somewhere is grabbing ALL of them up. Please input here as to your problems and what you have done to find these pch parts.
Attachments
- Status
- Not open for further replies.