I ended up buying a large number of Toshiba 2SJ201 2SK1530 to get some close matches for a project I am working on. Kudos to Patrick (EUVL) on turning me onto these. These are definitely the most complimentary MOSFETs I've tested and would highly recommend them for the Nelson Pass projects. The result is jaw dropping in my opinion.
Clearly, this required a large cash outlay on my part, as these aren't cheap devices. My matching pool was 72 n-ch and 72 p-ch. What I'd like to do is sell these and acquire more and try and keep these in steady supply to the DIY community.
I tried to carefully document my matching process. It has produced very close matches that prove themselves outside of the test rig. I found that good test equipment and carefully controlling the temperature to be the key.
I matched these at 2.000A at 16V rails at 60C. I used a block of aluminum and a temperature probe to keep the matching rig at the same temperature throughout the tests. I used a Fluke 287 50000 count meter to measure Vgs to 100nV (0.0001V) via the data acquisition mode the meter offers. I also used another meter to monitor the current within 1mA. I used a battery to bias the device. No passive components were in the matching circuit.
As a result, I have a number of devices that are matched complimentary pairs, as well as a number of matched devices all the way up to quintets.
Here are a few pics:
Here's a snapshot of the Vgs while I was warming up the test jig:
I was varying the 9V supply in order to keep a constant 2.000A current, hence the changes
Here's a snapshot of a 2J201 during test:
After initial adjustment, I probably undercompensated a bit and after the test time of 3 minutes the device was right at 2.000A, so I took the ending Vgs number and moved to the next.
Here's a snapshot of a 2SK1530 during test:
Looks like I had to compensate a bit towards the 2 minute mark and it was flat after that (i cannot compress the x-axis on the meter).
Anyways, I'd like to start the pricing at $17.50 USD a pair (one 2SJ201 and one 2SK1530) and that includes PayPal fees and worldwide first-class shipping. If you want anything exotic (please specify preferences, i.e. comp pair, matched 2,3,4,5 etc.) it might be more, depending on how rare the condition you are interested in happens to be. I'll provide the test data for the devices that match what you are looking at, and we can go from there.
Thanks everyone, I'm hoping to get the word out on these!
-Marc
Clearly, this required a large cash outlay on my part, as these aren't cheap devices. My matching pool was 72 n-ch and 72 p-ch. What I'd like to do is sell these and acquire more and try and keep these in steady supply to the DIY community.
I tried to carefully document my matching process. It has produced very close matches that prove themselves outside of the test rig. I found that good test equipment and carefully controlling the temperature to be the key.
I matched these at 2.000A at 16V rails at 60C. I used a block of aluminum and a temperature probe to keep the matching rig at the same temperature throughout the tests. I used a Fluke 287 50000 count meter to measure Vgs to 100nV (0.0001V) via the data acquisition mode the meter offers. I also used another meter to monitor the current within 1mA. I used a battery to bias the device. No passive components were in the matching circuit.
As a result, I have a number of devices that are matched complimentary pairs, as well as a number of matched devices all the way up to quintets.
Here are a few pics:
Here's a snapshot of the Vgs while I was warming up the test jig:
I was varying the 9V supply in order to keep a constant 2.000A current, hence the changes
Here's a snapshot of a 2J201 during test:
After initial adjustment, I probably undercompensated a bit and after the test time of 3 minutes the device was right at 2.000A, so I took the ending Vgs number and moved to the next.
Here's a snapshot of a 2SK1530 during test:
Looks like I had to compensate a bit towards the 2 minute mark and it was flat after that (i cannot compress the x-axis on the meter).
Anyways, I'd like to start the pricing at $17.50 USD a pair (one 2SJ201 and one 2SK1530) and that includes PayPal fees and worldwide first-class shipping. If you want anything exotic (please specify preferences, i.e. comp pair, matched 2,3,4,5 etc.) it might be more, depending on how rare the condition you are interested in happens to be. I'll provide the test data for the devices that match what you are looking at, and we can go from there.
Thanks everyone, I'm hoping to get the word out on these!
-Marc
Just an fyi, this price is cheaper than if you went to digikey and put them in your cart.... not even considering shipping.
I'd like two pairs for my F5.
doug
Can I replace with these directly to F5 ?
Good offer, but perhaps you should mention how good the matches are (10mV?)
yes, good question. I was assuming people would ask me directly about that. Here's a matched quartet of J201:
Code:
17 57 2.0010 59 2.2874
7 57 2.0010 60 2.2873
3 57 2.0060 59 2.2812
22 57 2.0010 59 2.2812
The columns are (device_num, starting_temp, ending current, ending temp, ending Vgs)
Here's a quartet of matched K1530 as well.
Code:
32 58 2.002 61 2.3909
35 59 2.000 61 2.3923
26 57 2.001 60 2.3959
3 58 2.000 62 2.3984
Still getting things in a spreadsheet and what not, but I can do at the very least 10mV difference, and in most cases better.
Last edited:
do these drop into an F5 or are there mods?
Not exactly drop in. You have to adjust drain resistors of the jfets to get the gain back.
Have fun, Hannes
okay, I have the following left (prices include PayPal fees and worldwide first-class shipping)
two untested K1530 $5 takes the pair
one matched J201 pair (2mV spread, very tight matching) $10 takes the pair
(1) F5 set $40 (4 devices)
(2) F5X (within 7mV across the quads) sets $80 a set (8 devices)
prices raised to help compensate for the waste in the matching process, and supply/demand. I'll split the quads if/when the pairs sell.
two untested K1530 $5 takes the pair
one matched J201 pair (2mV spread, very tight matching) $10 takes the pair
(1) F5 set $40 (4 devices)
(2) F5X (within 7mV across the quads) sets $80 a set (8 devices)
prices raised to help compensate for the waste in the matching process, and supply/demand. I'll split the quads if/when the pairs sell.
F5 set
Hi Luvdunhill,
Could I purchase the (1) F5 set of 4 ($40) - these will set up the new F5 project perfectly and hopefully the new version pcbs will be available in a few weeks too.
Could you direct me to your Paypal address and I'll get the money over to you straight away.
... James Hill (jameshillj@optsnet.com.au)
Hi Luvdunhill,
Could I purchase the (1) F5 set of 4 ($40) - these will set up the new F5 project perfectly and hopefully the new version pcbs will be available in a few weeks too.
Could you direct me to your Paypal address and I'll get the money over to you straight away.
... James Hill (jameshillj@optsnet.com.au)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.