Distribution of Idss in a bunch of random 2SJ74?

[Corvus corax]

I have received a lot of 50 Toshiba 2SJ74 from a vendor in HK, unmatched, and this evening took the bunch of them to the bench to match them up. They were grade "BL" devices, and for the most part, with the exception of maybe 2 of them, they all fell within what I believe to be the range specified for the part. I was able to get 18 relatively well matched (within .1- .05mA) pair out of the bag of parts, ranging from about 7mA to 17mA in Idss.

I was somewhat surprised to see that much variation, but I'm betting I have received the remains of some cherry picked stock, because they didn't come on a tape strip or the like. I'm going to install them in my BA3 gain stage and see what happens. So far, they look good, and appear to be genuine.

Anything else I should be looking for in my anti counterfeit tests? If it turns out that these are real, I may want to order some more…

[EUVL]

See :

The next best thing to 2SK389 / 2SJ109 ?


Patrick

[MRupp]

ranging from about 7mA to 17mA in Idss.

hmmm ... I am surprised you get any BL grades outside the 6 - 12 mA BL range. Did you measure them properly and how many did you get above 12 mA?

[Nelson Pass]

Originally Posted by Corvus corax

I was somewhat surprised to see that much variation, but I'm betting I have received the remains of some cherry picked stock, because they didn't come on a tape strip or the like. I'm going to install them in my BA3 gain stage and see what happens. So far, they look good, and appear to be genuine.

They were probably cherry picked. The usual yield is the bulk of the parts
within 2 ma of average. Most of my BL's are 8 to 10 mA.

That said, put them to work.

[qusp]

yep, never seen a BL over 12 and usually not over 10, maybe there were some legit and some dodgy parts mixed, do they all have the same markings?

[Corvus corax]

They all have the same markings, and it looks as if I ended up with somebodies leftovers after they were done matching a bunch. Most of them are between 8 and 10 mA. I had them held down to a temperature controlled aluminum block, that was set for 78 degrees, and they had the power applied for exactly 3 seconds before the readings were taken. I rounded up my matches to the nearest .1 mA- I figured that was good enough for my amplifier experiments.

The outliers of the bunch measured a minimum of 5.5mA, and a max of 17. There was only 1 that high, but there were a couple more that measured 13- 14mA.

As soon as I get a chance to match up my 170s, I'm going to fire the gain stage up and see what happens…

[AndrewT]

Some of the "error" will be down to the testing/measuring method.
The two that were 13 to 14mA could well be within specification.
The one odd ball @~17mA sounds like a 1% batch mistake.

I would expect unsorted random samples to cover the whole range of Idss for that device.

[Corvus corax]

I looked through the lot today, and found the highest was 17.5mA, the lowest was 6.5mA, and the majority of them settled in two groups: 7-8mA, and 10-15mA.

[Lazybutt]

Funny that everyone thinks they are from a cherry picked batch... when 2SJ74 are (among) the most faked devices in Audio history....

Having so many off-specs (above 12mA).... especially with same date code !!! is enough for me to think they are not genuine

Where did you buy them from ? and at what price ? I'm sure that will tell us more about those parts

[Floric]

Hi,

even when the sj74 was in production, I alway observed a distribution of idss values, that was not uniform. There were always "clusters" at one or two values. Those values varied in every bunch. Perhaps the properties depend on the production run, the wafer the sj74 came from or...

In greater bunches there were always one or two out of the spec.

I don't think that this must be an indication for fakes.

Regards

Flo