LOL…did same google and was surprised and studying over results.
Now just wondering about pin length on transistors being long enough to make it through quarter inch aluminum “T” bar and still seat properly in socket.
I’m assuming I’d have to do a chassis type mounting of the transistors on the extrusions and remote mount OS boards and run wire jumpers to the pins and case of these TO-3 devices?
Now just wondering about pin length on transistors being long enough to make it through quarter inch aluminum “T” bar and still seat properly in socket.
I’m assuming I’d have to do a chassis type mounting of the transistors on the extrusions and remote mount OS boards and run wire jumpers to the pins and case of these TO-3 devices?
try to choose sockets looking as having shortest length from surface to pin clamps
then when you get it, you'll know is there enough pin length to leave pcb sandwiched
though, it's best to practice some desoldering, or even better - to buy desoldering station
life is too short to not have best tool for job
then when you get it, you'll know is there enough pin length to leave pcb sandwiched
though, it's best to practice some desoldering, or even better - to buy desoldering station
life is too short to not have best tool for job
I wonder how those impact heat dissipation? I’d make sure they do that well before using - it’d be a shame to socket only to loose them to overheating.
I'm sure what you're thinkin' now ........ damn optocoupler-flipping greenhorns......... you can't with them, you can't without them .....
ZM, what about the current? Can the sockets be trusted to connect solidly enough and keep the connection for 1.5A plus of continous current?
Forgot to mention… as I was thinking and googling at the same time as Mighty ZM was answering and scolding me for being a lazy de-soldering Padawan…
I believe the same type sockets were used to mount TO-3 transistors to the heatsinks in my old Counterpoint SA12
Developed plenty of heat and drew substantial current…nevver an issue with it…and I’m sure countless other Mil-spec applications it seems that the TO-3 sockets have been used in.
I found some nice sturdy looking bakelite versions with gold plate contacts on eBay
I believe the same type sockets were used to mount TO-3 transistors to the heatsinks in my old Counterpoint SA12
Developed plenty of heat and drew substantial current…nevver an issue with it…and I’m sure countless other Mil-spec applications it seems that the TO-3 sockets have been used in.
I found some nice sturdy looking bakelite versions with gold plate contacts on eBay
Papa stated the above in post 77.
I just want to make sure I interpret it correctly…
Matching VFETs help soothe audio nervosa.
It is possible to build these iterations of the VFET amps using unmatched 2sj28, 2sk82, 2sj18 and 2sk60…possibly others…as long as you make adjustments to a few (gate?) resistors?
I ask because I’m considering purchase of unmatched VFETs to keep the 2016 push/pull kit intact to build afterwards.
regarding ranking:
I have unmatched 2SK182. Worse, one has gate current. [Specs say max 1 mA; my case, up to 200micro).
- The biasing differs to get 2.1 A: one channel -6.5V, the other -7.3V. I have the SIT amplifiing, choke in drain.
The result: 0,5 dB difference between channels. My Alps Royal Blue pot also has a channel difference of 0,5 dB. The amp goes flat from 6 Hz to very high. No overshoot what so ever.
This is all totally inaudible.
With the SIT as a source follower there will be maybe < 0,1 dB difference. Moot.
Why matching might be important: if there is feedback and hence risk of current hogging between paralleled devices (e.g. the SONY amps).
I have unmatched 2SK182. Worse, one has gate current. [Specs say max 1 mA; my case, up to 200micro).
- The biasing differs to get 2.1 A: one channel -6.5V, the other -7.3V. I have the SIT amplifiing, choke in drain.
The result: 0,5 dB difference between channels. My Alps Royal Blue pot also has a channel difference of 0,5 dB. The amp goes flat from 6 Hz to very high. No overshoot what so ever.
This is all totally inaudible.
With the SIT as a source follower there will be maybe < 0,1 dB difference. Moot.
Why matching might be important: if there is feedback and hence risk of current hogging between paralleled devices (e.g. the SONY amps).
difference is in xconductance/internal resistance of SITs, having nothing with biasing voltage
anyway, combine channels in that way, with ALPS - stronger channel getting weaker signal from ALPS
anyway, combine channels in that way, with ALPS - stronger channel getting weaker signal from ALPS
Fwiw, the 2016 amp doesn’t require tight matching either. My parts (sourced, not kit) were all measurably different and the end result measured and sounded great.
IMHO, I would go for a Lazy Singing Bush for that devise as it needs more voltage to get to the sweet spot as a Follower.
What I did is make a 'balance pot' with a 5k resistor and a 10 k pot that can be turned either way, for that balance. The tube pre-amp itself again has some slight imbalance. [I used to have 300B SE and without NFB it was within 1 dB].difference is in xconductance/internal resistance of SITs, having nothing with biasing voltage
anyway, combine channels in that way, with ALPS - stronger channel getting weaker signal from ALPS
I used to tell meself I could hear the difference on my 1957 Electrostats. Indeed, panning gave a shift of image.
But that imaging was also possible because I had a CD (Meridian) and preamp (lots of 2nd harmonics??) that together gave just that tad of sparkle.
Now I do not hear this Nirvana it anymore (small full-range horn, high WAF). 😢
I am exploring a French-style "Némésis" stage with these: with a 100 to 150 ohms/8 ohms transformer. The Némésis with the original 2SK135 was the best amplifier I ever heard - in Paris as built by the designers.
Nelson Pass wrote some words about Jean Hiraga's Némésis too. "The Pass SIT1 improved the [2SK135 Mosfet and . .] SiC based Arch Nemesis design another notch. "
Alazira: Because my speakers are something between 8 and 6 ohms (higher = lower) I opted for now for R1 = =0,82 ohm and R2 = 0,12 ohm. I thought it gives this little bit of boost.
But of course in time I will parallel the 0,82 with some 10 ohms or so and short ot the 0,12 sooner or later to make a real CCS. And get your figures.
But I hope I will keep out of the amplifier. I should put up a sign: KEEP OUT.
🙄
But of course in time I will parallel the 0,82 with some 10 ohms or so and short ot the 0,12 sooner or later to make a real CCS. And get your figures.
But I hope I will keep out of the amplifier. I should put up a sign: KEEP OUT.
🙄