I have a reel of the 2SK209s I've yet to open - that's good news about the consistency. I've been balking at putting together a clamshell sorting rig for Id vs. Vds for SOT-23 parts. I'll have to do it sooner or later to preserve my sanity, as I have a bunch of SOT-23 parts that won't be anywhere near as consistent.
Another way to make SOT-23 parts more amenable to sorting is to get some plated-through pad-per hole breadboard and cut it into little squares with 4 holes per square. You can then mount a SOT-23 part on 3 of the holes with bare wire leads so that you can test it as a normal leaded part. I usually use a low value 1/8W resistor as the gate lead (47-100 ohms) so that the part does not oscillate in the sorting jig. The oscillation can give you squirrelly results while you're trying to sort parts. The other alternative is to add the stopper resistor on the sorting jig. Since you are testing under DC conditions it's OK to use a 1k stopper resistor in the jig so that the DUT is quiet under all circumstances
Another way to make SOT-23 parts more amenable to sorting is to get some plated-through pad-per hole breadboard and cut it into little squares with 4 holes per square. You can then mount a SOT-23 part on 3 of the holes with bare wire leads so that you can test it as a normal leaded part. I usually use a low value 1/8W resistor as the gate lead (47-100 ohms) so that the part does not oscillate in the sorting jig. The oscillation can give you squirrelly results while you're trying to sort parts. The other alternative is to add the stopper resistor on the sorting jig. Since you are testing under DC conditions it's OK to use a 1k stopper resistor in the jig so that the DUT is quiet under all circumstances
OK, so that's not a good choice.
How about the 2SK369 - still available as thru-hole, I am told?
Andy
Just as a matter of interest, wrenchone, the original 'Le Pacific' circuit has the Drains of the jfets connected to the +ve DC rail (via a resistor).
AIUI, this setup removes any Miller capacitance - which is important, in the case of the 1st gain stage of a MM phono stage (otherwise this Miller capacitance would be loading the cart).
Does the 1st gain stage setup in your revised circuit (which is cascoded) also remove Miller capacitance?
Nice one your jig. I have this useful thingy for SOT-23.
Attachments
Aah, OK, Salas ... I didn't know this.
I thought that when Drain was connected to the +ve DC rail (via a resistor which sets the gain) - see attached circuit - this setup removed Miller capacitance?
Thanks,
Andy
Attachments
No, its a classic common source arrangement.
Watch this video. You will understand.
Basics of a Cascode Amplifier and the Miller Effect
Watch this video. You will understand.
Basics of a Cascode Amplifier and the Miller Effect
Hi Salas,
I notice that the circuit which the guy shows in his video has the cascode as an NPN transistor sitting on top of a second NPN (so the Emitter of the top transistor is connected to the Collector of the bottom transistor).
But I have seen in other circuits, a PNP transistor sitting on top of a 2SK170 (so the Collector of the top transistor is connected to the Drain of the bottom 2SK170).
So can either a PNP or an NPN transistor be used as the cascoding transistor?
Thanks,
Andy
Yes, but a PNP must be used when the DC voltage at its collector is lower than the one's it cascodes. Liike in "folded" cascode that has the load resistor connecting to ground. My UFSP phono's input stage for instance. Which is NJFET+PNP hybrid fold. The straight to rail cascode as in the video, which is most commonly met, is also sometimes named "telescopic".
P.S.
Always emitter input on top to be a common base for cascoding. Collector to drain you saw maybe it was a constant current source load.
P.S.
Always emitter input on top to be a common base for cascoding. Collector to drain you saw maybe it was a constant current source load.
Can you point me to the thread about your UFSP phono stage, Salas?
I looked in "Analogue Source" but couldn't find it.
Andy
iteration next is 2x2SK209 in parallel per position. developed board, fets matched, separate g.stoppers and s. resistors per fet. pcb is done fast , in a few h, and components mounted in 3D. voltage gain is fine, 35dB, DC supply with Rs=30 ohms must be over 24V, from 28-32V.. or Rs to be increased to 40ohms and use lower DC supply (with lower total gain ). found some nos polystyrene Rifas which i will test through RIAA and coupling positions.
on first there is no big difference between this and previous version with 2SK2394.. need to make A/B switcher to find out differences fast.
on first there is no big difference between this and previous version with 2SK2394.. need to make A/B switcher to find out differences fast.
Attachments
Can you point me to the thread about your UFSP phono stage, Salas?
I looked in "Analogue Source" but couldn't find it.
Andy
In analogue source, find the simplisitc njfet riaa thread, in its post#1 there are links for its evolution phases, about UFSP is the bottom link.
Nice test board work. Its for MM cartridge, yes? Because gate stoppers and highish source pin resistors contribute directly to the noise budget. Signal source impedance must be much higher than those to sufficiently swamp them.