Thanks for your reply,
I intend to use the modified diagram 2 with toshiba 2Sk170/2SJ74 fets, which can be found on ebay by the way.
I take it to increase the current i would increase the values of the 330R resistors R1 and R2? slightly?
I intend to use the modified diagram 2 with toshiba 2Sk170/2SJ74 fets, which can be found on ebay by the way.
I take it to increase the current i would increase the values of the 330R resistors R1 and R2? slightly?
Thanks for your reply,
I intend to use the modified diagram 2 with toshiba 2Sk170/2SJ74 fets, which can be found on ebay by the way.
expect some of them to be fakes
do not say this was my idea 😀 but if there would be enough to take the chance with buying a whole box, and someone would be willing to do the hard work and test them, well.....
ups, wrong subject

Hi
Yep, Unfortunatley i here some could be fakes, How bad is that
selling possibly fake fets to make a few pounds!
Yep, Unfortunatley i here some could be fakes, How bad is that

So, back to Can anyone suggest alternatives to the k170/J74? These are hard to source....
K246/J103 pair is a surprisingly good replacement. Unfortunately getting a matching pair for N and P devices is much harder than K170/J74 pair. Distributions of Idss for K246 and J103 are oddly very different.
Gentlemen, you could build this buffer with common and cheap 2n5459 and 2n5462 and still get pretty good performance. Its hard to hear a difference compaired to the toshibas although THD will be higher as well as noise but at 100 db down levels you cant hear it. Jlh built these buffers with this pair and I have used them a couple of times too. The other mentioned toshiba pair 103/246 is also a good alternative but agin THD increase because of their lower gm.
You can parallel some of the lower Gm devices.
I use this buffer since many years in various disguises. BC516, BC527 darligtons improve the performance but you have to watch the higher Ube drop of ca. 1.5V.
I use this buffer since many years in various disguises. BC516, BC527 darligtons improve the performance but you have to watch the higher Ube drop of ca. 1.5V.
Last edited:
You can keep the values of R1, R2 ( 390 Ohm ). Idle will simply double and that gives a bit more Class A.
Better still when you turn the bjt into a cfp, this is output I use for headphone amps to drive 30 ohm phones in open loop.
sorry to bring up a dead thread but what would be the current draw of the modded version?
i built one with same resistor values using K246/J103 and um... 2sa1268/c3200 (970/2240 sub) powered by salas shunt and not only does it oscillate but it the shunt doesn't even go active, only red leds coming on with under voltage.
i've noticed the same thing happening the salas shunt in the past when i tried to draw too much current.
and could those generic bjts work?
i built one with same resistor values using K246/J103 and um... 2sa1268/c3200 (970/2240 sub) powered by salas shunt and not only does it oscillate but it the shunt doesn't even go active, only red leds coming on with under voltage.
i've noticed the same thing happening the salas shunt in the past when i tried to draw too much current.
and could those generic bjts work?
Last edited:
- Status
- Not open for further replies.
- Home
- Amplifiers
- Solid State
- Another jfet buffer