I think these is the company I wrote, asked more pictures (from the actual items) unfortunately I never get any answer (yet)
Based on the picture they posted look authentic but can be the usually Chinese method (trick) post original pictures and ship counter faith items.
How did you ( try) to communicate with them? In English? Maybe I'll give it a try also...
Why the hell can't an American company produce those fets? If Linear Systems can do 2sk170s and 2sj74s, why not 2sk82s and 2sj28s? Oh, BTW, is Linear Systems still an American company?
It is indeed a sad state of affairs. The Holy Grail of amplification shows up, but the crucial V-FETs (SITs) are sold out.
I guess such transistors could be made in 2015. The manufacturing process is not entirely lost, I would guess.
Another question is if could be made within reasonable cost, whatever that is. And if some company has any interest in this.
I guess it would take a fierce Kickstarter campaign to re-launch these devices!!!
I will pitch in here. I did buy a number of pairs of 2SK KE and 2SJ KF of these in anticipation of this project. What I think should happen is that we should pool the ones we do have together and match them the best we can, for those that have enough to build an amp, buy maybe not enough to meet .2vgs.
Further if the front end is standalone useful as a pre, then a pcb could be started for that. While we wait for possible variations on a back end circuit.
Further if the front end is standalone useful as a pre, then a pcb could be started for that. While we wait for possible variations on a back end circuit.
Hello Teabag, we meet again. Now I have new guardians for my treasure: palace pets!!!, and of course little ponies.....It is a joke.
Pool was an idea that my friend dady told me some time ago and I proposed for doing with tokins. I know only that two friends exchanged their vfets. Logistics is complex. What happen if some packet is lost?.
Look curves from 2sj28 mines. A lot of variability. We need to do a big pool for matching a big number of devices. A lot of work. Who wants to do?.
For one person is too much. We need some persons within different regions of the world. A big effort.
Is it enough matching using one point only? I do not know. This is a question that we need solve. And other question, if we have two 2sj28 matched, for example, now we need matching to other pair of 2sk82, or not?
When I did lamp with thf51 I used two randomly selected. It was a shock for me. Then I exchanged with other devices with curves almost equals, but they not sounded well. I do not know why. When I used the first pair all sounded well again.
I offer to collaborate for pool devices... but we need do it well (logistic, selection criteria...).
Ideas?
Pool was an idea that my friend dady told me some time ago and I proposed for doing with tokins. I know only that two friends exchanged their vfets. Logistics is complex. What happen if some packet is lost?.
Look curves from 2sj28 mines. A lot of variability. We need to do a big pool for matching a big number of devices. A lot of work. Who wants to do?.
For one person is too much. We need some persons within different regions of the world. A big effort.
Is it enough matching using one point only? I do not know. This is a question that we need solve. And other question, if we have two 2sj28 matched, for example, now we need matching to other pair of 2sk82, or not?
When I did lamp with thf51 I used two randomly selected. It was a shock for me. Then I exchanged with other devices with curves almost equals, but they not sounded well. I do not know why. When I used the first pair all sounded well again.
I offer to collaborate for pool devices... but we need do it well (logistic, selection criteria...).
Ideas?
Attachments
It's a good idea Teabag and jama.
But I think matching can be done by most of us individual. If you are skilled enough to build this amp, then you must also be able to do the matching with the VFET's on a big chunk of heatsink.
We have to limit and agree to the parameters though.
So match all at Vds of24V or 28V as in the application. And measure Vgs versus Id only. At a certain temperature...
Then everybody can publish their results here.
I have measured 16 VFETs and found 4 reasonable pairs. 2 P-channels and 2 N-channel pairs.
Because I bought some more then needed, I'm willing to sell the possible useful VFET's to others here. Or even give them to you
I built the CSX1 VFET Amp with my worse VFETs, and it sounds great.
P and N don't have to be matched, the differences are to big anyway.
Walter
But I think matching can be done by most of us individual. If you are skilled enough to build this amp, then you must also be able to do the matching with the VFET's on a big chunk of heatsink.
We have to limit and agree to the parameters though.
So match all at Vds of24V or 28V as in the application. And measure Vgs versus Id only. At a certain temperature...
Then everybody can publish their results here.
I have measured 16 VFETs and found 4 reasonable pairs. 2 P-channels and 2 N-channel pairs.
Because I bought some more then needed, I'm willing to sell the possible useful VFET's to others here. Or even give them to you
I built the CSX1 VFET Amp with my worse VFETs, and it sounds great.
P and N don't have to be matched, the differences are to big anyway.
Walter
Last edited:
I think this is our singular target.
from Vfet 2 article
" matched the VFETs for a Vgs within 0.1 volts at 0.5 amp current at 20 volts,
and 0.2V Vgs is good enough."
From 40th aniv article
"Fortunately these parts have a near-zero temperature coefficient. They require
no thermal compensation and can be relied to give the same performance over
a wide temperature range without altered performance."
So I am not sure we really need to be concerned with heatsinking at all, and maybe just a basic measure will do.
I think if we all wanted curve matching in a skeleton market, we are being too greedy.
And as mentioned, it may not even sounds best to human ears. (I would welcome Mike's Frankentracer based comments). I am willing to throw away perfection in order to get the most usable parts extracted and used. I am more pragmatic than perfectionist.
Please correct me if my assumptions are wrong or skewed, they often are.
I can answer some logistic questions from stuff I have done in the past, the problems are solved by assuming risk, and throwing money at it. My shipping history is about a 1.5-2% loss of packets from USPS. But that often includes the saavy consumers in some areas using Global Express mail for tracking which comes at a real premium cost. The rest means insuring packages (I did this with last of SemiSouth units I had). For international destinations this means full retail cost on customs forms.
WalterW standardizaton technique could lead to a real traders market and lower shipping costs.
Luckily, the packaging of the vfets is pretty good, so I have not heard of damaged unit in shipping.
from Vfet 2 article
" matched the VFETs for a Vgs within 0.1 volts at 0.5 amp current at 20 volts,
and 0.2V Vgs is good enough."
From 40th aniv article
"Fortunately these parts have a near-zero temperature coefficient. They require
no thermal compensation and can be relied to give the same performance over
a wide temperature range without altered performance."
So I am not sure we really need to be concerned with heatsinking at all, and maybe just a basic measure will do.
I think if we all wanted curve matching in a skeleton market, we are being too greedy.
And as mentioned, it may not even sounds best to human ears. (I would welcome Mike's Frankentracer based comments). I am willing to throw away perfection in order to get the most usable parts extracted and used. I am more pragmatic than perfectionist.
Please correct me if my assumptions are wrong or skewed, they often are.
I can answer some logistic questions from stuff I have done in the past, the problems are solved by assuming risk, and throwing money at it. My shipping history is about a 1.5-2% loss of packets from USPS. But that often includes the saavy consumers in some areas using Global Express mail for tracking which comes at a real premium cost. The rest means insuring packages (I did this with last of SemiSouth units I had). For international destinations this means full retail cost on customs forms.
WalterW standardizaton technique could lead to a real traders market and lower shipping costs.
Luckily, the packaging of the vfets is pretty good, so I have not heard of damaged unit in shipping.
Last edited:
Yes, curve matching is total overkill here. Look at the procedure I did on my Lab Hangout. It's the one NP suggested and I demonstrated it for soundhappy, or Frankentacer makes Vgs matching super simple as well (I can actually do it faster on Frankentracer so that's what I use.), but curves aren't needed for this endeavour.
Tokin SITs are much, much more widely varied than these Sony's, and there are some bad parts out there. Hence the curve tracing.
Tokin SITs are much, much more widely varied than these Sony's, and there are some bad parts out there. Hence the curve tracing.
Last edited:
Since a limited quantity of the parts does exist, making this happen is mostly
an exercise in logistics.
I contacted the ever-competent 6L6 and we are discussing it.
Three obvious options:
kit with matched dual pairs
kit with single pairs
kit using P channel only.
Acronman informs me that the P channel parts are moving.
an exercise in logistics.
I contacted the ever-competent 6L6 and we are discussing it.
Three obvious options:
kit with matched dual pairs
kit with single pairs
kit using P channel only.
Acronman informs me that the P channel parts are moving.
Oops, you are right the VFETs are running at 19V in this schematic, about 0.5- 0.75 Ampere each.
