I am curious if the recent quads I got from watanabetomoaki are anything like other 2SK180s I have, which are really close to the 2SK77s. Might have to dust of the old Tex curve tracer....
watanabetomoaki references the 2sk77s in his listings. I did ask if the 2sk180s he is selling as precisely matched quads are suitable replacements for the 2sk77s and I did not get a response, so I am prepared for the worse....
watanabetomoaki references the 2sk77s in his listings. I did ask if the 2sk180s he is selling as precisely matched quads are suitable replacements for the 2sk77s and I did not get a response, so I am prepared for the worse....
Just got from Tomo as "rough matched quad (without data)". When I asked he said that it means "similar Vgs at Vds = 50V, Id = 0.5A".
Looking forward to test them by myself when I am back at the desk in few weeks.
Attachments
Now that I have looked at some SIT schematics, the circuit fundamentals are near identical to single-ended class A tube design, I am right at home.
I may use one pair of THF-51S for my own design and save the other for Nelson's. I think I would want the most basic, a common source SIT amplifier with capacitor coupled output. I have a tube curve tracer, but current is only rated up to 300mA, perhaps I need a second tracer...
I may use one pair of THF-51S for my own design and save the other for Nelson's. I think I would want the most basic, a common source SIT amplifier with capacitor coupled output. I have a tube curve tracer, but current is only rated up to 300mA, perhaps I need a second tracer...
I was told he can not guarantee how precisely matched because "preciseness" is not really quantitative. Then I got eight pcs of "matched" 182ES from him last year. I have not fully traced them, but all of them are very well matched (for me) at several different voltage / current points measured.
Last edited:
So how important is matching for any of the designs. I was looking at the price differential between matched and unmatched. If matching it’s a big deal then I will just get some of the unmatched ones. I was looking at the Tokin ones from Japan.
I picked up two pairs of the THF-51S to have for later, higher power, versions of the DIY VFET amplifier. There are several circuit designs that have been presented over the last few years that can make good use of the big SITs in single-ended class A designs. These include:
50w Single-Ended BAF2015 Schade Enabled
The Singing Bush
2SK82 -2SJ28 SE Amp
25W Single Ended Hammond 193V Choke Loaded 2SK180 L'Amp
The SIT-3X Amplifier
... and of course Most Greedy Boy, of them all... or (there is no) DEFiSIT of Papa's Koans
Just some light reading if you need help falling asleep at night. 😉
I'm sure that the next version from Nelson will be a little bit different.
Got my roughly matched pair in yesterday. Looking forward to see what you’ve got cooking here Nelson!
Thanks for the reading materials! Will definitely read into these designs. I feel a rabbit hole opening below me, wonder if I should purchase any more SITs while I'm at it 😀
So I spent most of today testing a population of 2SK182es and THF51s,
in a setup which was Common Source mode (Grounded Source, input on
Gate, Output from Drain) loaded into 8 ohms which is then attached to
a +50V supply. It is run without feedback, the Gate is driven directly by
the AP analyzer with a variable DC bias.
This would be an arrangement designed to magnify the differences
in gain and distortion.
Two lot codes of each were tested with sequential serial numbers.
The bias was adjusted for the lowest THD at 1 watt, which is the point
where the 2nd harmonic is nulled, and data and a distortion curve from
1 watt to 10 watts was taken at that bias point.
For THF51s the Vgs values ranged from -4.7 to -2.8V and Vds point was
36 to 40V. Distortion figures for 1w were from .15% to .25% and 10 watt
from 1.7% to .25%
For 2SK182es the Vgs values ranged from -5.4 to -3.3V and the Vds was
from 32 to 43V. Distortion figures for 1w were from .3% to .5% and 10
watt from 2% to 4.3%
The result is that there is a reasonable correlation between Vgs and
performance within a lot code, and not so much outside, but this is the
worst case scenario - Common Source without degeneration.
All the parts were good, and the Gate leakage was about 5 uA.
For those of you who have matched parts or sequential serial #'s on
the parts, you are probably in good shape. If not, don't get excited,
because everyone will have to make some adjustments for optimal
performance in any case (once you decide what that means...)
I know you like some porn with your posts, so here's a photo of the
bench today and the test rig.
in a setup which was Common Source mode (Grounded Source, input on
Gate, Output from Drain) loaded into 8 ohms which is then attached to
a +50V supply. It is run without feedback, the Gate is driven directly by
the AP analyzer with a variable DC bias.
This would be an arrangement designed to magnify the differences
in gain and distortion.
Two lot codes of each were tested with sequential serial numbers.
The bias was adjusted for the lowest THD at 1 watt, which is the point
where the 2nd harmonic is nulled, and data and a distortion curve from
1 watt to 10 watts was taken at that bias point.
For THF51s the Vgs values ranged from -4.7 to -2.8V and Vds point was
36 to 40V. Distortion figures for 1w were from .15% to .25% and 10 watt
from 1.7% to .25%
For 2SK182es the Vgs values ranged from -5.4 to -3.3V and the Vds was
from 32 to 43V. Distortion figures for 1w were from .3% to .5% and 10
watt from 2% to 4.3%
The result is that there is a reasonable correlation between Vgs and
performance within a lot code, and not so much outside, but this is the
worst case scenario - Common Source without degeneration.
All the parts were good, and the Gate leakage was about 5 uA.
For those of you who have matched parts or sequential serial #'s on
the parts, you are probably in good shape. If not, don't get excited,
because everyone will have to make some adjustments for optimal
performance in any case (once you decide what that means...)
I know you like some porn with your posts, so here's a photo of the
bench today and the test rig.
Attachments
I am so appreciative of this work you are doing for us.lucky.few. Thanks for your generous gift of time and knowledge. And I'm so envious of that fantastic work bench. But at least I now have a goal to work toward.
- David
- David
That’s v useful info. The difference in THD performance between the THF and the 2SK182 seems substantial. Does anyone know how the 2SK180 fare in this regard?
Thanks for the data NP, happy to have jumped aboard this train early on, very curious what you are cooking up...
A question from one who is new to SIT amplifiers - is it generally thought that different devices give "flavors" that go beyond the distortion characteristics? Similar to SET amplifier voodoo where one might prefer a 2A3 to a 300B?
A question from one who is new to SIT amplifiers - is it generally thought that different devices give "flavors" that go beyond the distortion characteristics? Similar to SET amplifier voodoo where one might prefer a 2A3 to a 300B?
Last edited:
There is so much going on in this forum it is hard to keep up. What makes it doubly hard is that there are so many great projects already available and it seems many more just over the horizon it is hard to know what to start next.
Thank you Mr. Pass for making problems like deciding what is the next great sounding Pass design to build.
I get them while the last. A couple of pairs.
Thank you Mr. Pass for making problems like deciding what is the next great sounding Pass design to build.
I get them while the last. A couple of pairs.
I am curious if one of the Pass creations will be a 2SK182ES or THF51 DEF? The IXYS/Littlefuse P-channel hockey pucks seem a great partner for the Tokin SIT DEF. Higher output power than the Sony VFET creations. Simple to mount also. No heatsink bracket needed.
Vfet Lot numbers
I got 4 unmatched THF-51's from Tomo and they have these codes on them:
1X6924
1X6913
1X6906
1X6922
Are these all from different Lots?
Thanks,
Robert
I got 4 unmatched THF-51's from Tomo and they have these codes on them:
1X6924
1X6913
1X6906
1X6922
Are these all from different Lots?
Thanks,
Robert
Thanks Nelson for taking the time to test a few of the lots and for sharing pics of your test rig. I noticed you have a couple of capacitors and resistors set up in your test rig. I assume you’re using the same rig described in your matching FETs article, but, I don’t recall capacitors being described. Is this to smooth out the use of SMPS as a power source? Thanks.
Last edited: