Since NXP (and Nexperia) have discontinued manufacture of JFETS (EOL a couple of years ago), and in particular the BF862, what is the suggested low noise n-channel jfet? Well OK the LSK170 - but is there another tested alternative lurking?
Having asked the question I found BF862 is now End Of Life. Well that sucks. | Details | Hackaday.io so the demise has caused difficulties in the physics community. There is a suggestion in that link that the 2SK2394 is a viable alternative. It certainly has the same application area in am tuner and low noise applications. Has anyone looked at this device?
Craig
Having asked the question I found BF862 is now End Of Life. Well that sucks. | Details | Hackaday.io so the demise has caused difficulties in the physics community. There is a suggestion in that link that the 2SK2394 is a viable alternative. It certainly has the same application area in am tuner and low noise applications. Has anyone looked at this device?
Craig
Good question. Who can say Hans. But it cuts deep.
Renesas, which was what Hitachi semiconductors became, and which later acquired NEC, has obsoleted the vast majority of its extensive discrete semiconductor catalog. All JFETs and bipolars have gone, although they seem to be keeping MOSFET power devices going, presumably because of automotive applications.
On Semi, which acquired Fairchild and Sanyo have been working through a "rationalization" process that has seen the demise of many devices, among them the long time standard BC560 (although it lives on in SOT23 as the BC860). And daft things like continuing with a PNP but obsoleting the NPN, or complementary pairs the are now only available in different gain bands.
I note that Litteldiode has acquired IXYS fairly recently. Just a matter of time...
At least LIS is keeping the low noise JFET flame alive, including monolithic duals.
INTERFET too, but with punishing prices. The one detailed in HH, the IF3602 dual with stunning noise performance of measured 0.35nV/rootHz rising to a mere 0.6nV/rootHz at 10Hz, currently costs $100 each from Mouser. Idss is not well controlled, datasheet says 30-300mA. And of course the price to pay is very high capacitance.
Craig
Renesas, which was what Hitachi semiconductors became, and which later acquired NEC, has obsoleted the vast majority of its extensive discrete semiconductor catalog. All JFETs and bipolars have gone, although they seem to be keeping MOSFET power devices going, presumably because of automotive applications.
On Semi, which acquired Fairchild and Sanyo have been working through a "rationalization" process that has seen the demise of many devices, among them the long time standard BC560 (although it lives on in SOT23 as the BC860). And daft things like continuing with a PNP but obsoleting the NPN, or complementary pairs the are now only available in different gain bands.
I note that Litteldiode has acquired IXYS fairly recently. Just a matter of time...
At least LIS is keeping the low noise JFET flame alive, including monolithic duals.
INTERFET too, but with punishing prices. The one detailed in HH, the IF3602 dual with stunning noise performance of measured 0.35nV/rootHz rising to a mere 0.6nV/rootHz at 10Hz, currently costs $100 each from Mouser. Idss is not well controlled, datasheet says 30-300mA. And of course the price to pay is very high capacitance.
Craig
Last edited:
3R is 0,22nV/rtHz or -193dB.
Together with the 0.3nV/rtHz amp this gives -188dB, or 2dB more than the -190dB noise from the ZTX amp alone.
In that case the 1R cabling in series with the Cart can be almost completely ignored.
But the noise from the Cart alone is -194dB, so in combination with the 1R lead and the ZTX amp, you loose 6dB.
Depending on the output of the Cart, the question is whether a 0.3nV/rtHz ZTX amp is acceptable for this cart instead of worrying anout the 1R cabling.
Hans
Together with the 0.3nV/rtHz amp this gives -188dB, or 2dB more than the -190dB noise from the ZTX amp alone.
In that case the 1R cabling in series with the Cart can be almost completely ignored.
But the noise from the Cart alone is -194dB, so in combination with the 1R lead and the ZTX amp, you loose 6dB.
Depending on the output of the Cart, the question is whether a 0.3nV/rtHz ZTX amp is acceptable for this cart instead of worrying anout the 1R cabling.
Hans
Rule of thumb: a series resistor 3 times smaller than the source resistance can be, noise wise, safely ignored.
brianco's Kondo Io has a coil resistance of 1 ohm - in which case the 1 ohm tone arm wire resistance degrades the source noise by 6dB.
However, with an output voltage of 0.12mV there is certainly a discussion to be had about the best method of amplifying such a small output from a 1 ohm source R.
There is no argument that as the cartridge R increases, the 1 ohm wiring resistance becomes rapidly less of an issue.
However, with an output voltage of 0.12mV there is certainly a discussion to be had about the best method of amplifying such a small output from a 1 ohm source R.
There is no argument that as the cartridge R increases, the 1 ohm wiring resistance becomes rapidly less of an issue.
If a connection only adds 10 mOhm, why not use a thicker wire inside the armtube and graft on a proper thin wire just before the exit?
Using Cardas 33 guage only adds about half an ohm loop resistance, so even with a 1 ohm cartridge only adds 1.75dB to the source noise.
There is a secondary effect. If such a 1 ohm cartridge is feeding a typical 10 ohm load, the half ohm lead resistance leads to a potential divider effect, reducing the signal voltage at the input by -0.4dB. Under those specific conditions the effective source noise is degraded by 1.75 + 0.4, or ~2.1dB
Whether that is worth a hill of beans in a practical context is debateable. However fakeall's concept of thicker wires in the arm tub transitioning to much thinner ones in the region of the arm pivot is not at all a bad idea, in the admittedly extreme example of a ~1 ohm cartridge.
There is a secondary effect. If such a 1 ohm cartridge is feeding a typical 10 ohm load, the half ohm lead resistance leads to a potential divider effect, reducing the signal voltage at the input by -0.4dB. Under those specific conditions the effective source noise is degraded by 1.75 + 0.4, or ~2.1dB
Whether that is worth a hill of beans in a practical context is debateable. However fakeall's concept of thicker wires in the arm tub transitioning to much thinner ones in the region of the arm pivot is not at all a bad idea, in the admittedly extreme example of a ~1 ohm cartridge.
Wow,But, as the standing joke goes 'I was attracted to vinyl because of the expense and inconvenience of it'
and all this time I thought it was because I didn't have any other choice.
But then as my wife and kid remind me, I'm the "OLDIE".
The good news is the people who know me don't tell me,
you have the cutest granddaughter. Then my wife tells them
that he's my father, don't let me kid you. -- Oh the joi.
Cheers,
