i had a friend buy some RF transistors from ebay. these transistors were in a TO-5 can. the originals had the case connected to the emitter, the ebay fakes had the case connected to the collector (which is a bad thing to do for RF transistors, as all of that case surface area is capacitive).
US sellers aren't immune to fakes,
I bought two matched pairs of 2SJ74BL's from ebay seller QUA-CO who had a good story about these being the last of the genuine Toshiba JFET's from his known good stock. I bought these last winter and sat them on the shelf. Recently I pulled all of the Jfet's out that I have gathered and the ones from Qua-co did not look right. The stenciling was wrong and one of the fets had silver paint on the backside of it. I built up the little JFET test rig that Erno Borbely outlined in his 1999 Audio Electronics article and tested them all. The ones bought from Qua-co had a pinch off voltage of 1.25V and a transconductance of 11 mS, FAKES!
In comparison I also purchased a set of matched 74gr's from ebay seller Alweit and they tested with a Vp of .33V and a transconductance of 27 mS. I also have a few old J44's that I pulled from a mid-80's piece of electronics (without a doubt genuine NEC's) a few years back and they also had a Vp of .42 and a transcon. of 25 mS. I can highly recommend Ebay seller Alweit as I also tested the 170BL's and 170GR's that I bought from him and they have proper gain numbers in the 23 - 25 mS range with a low Vp of approx. .4V.
Lesson I learned, just because someone is from the US and gives you a good story about genuine doesn't mean that they are.
It is time for an update of my answer in post #382 of this thread in which I wrote:
Some of the transistors manufactured by OnSemi have a different appearance with datecodes of 2012 and on. Genuine MJL1302/3281 (and possibly many other OnSemi transistors of the same shape) should now look like the MJL3281 in the attachment.
Compared to the "old" version, almost everything looks subtly different. They also seem to weigh a little less.
Some of the transistors manufactured by OnSemi have a different appearance with datecodes of 2012 and on. Genuine MJL1302/3281 (and possibly many other OnSemi transistors of the same shape) should now look like the MJL3281 in the attachment.
Compared to the "old" version, almost everything looks subtly different. They also seem to weigh a little less.
Attachments
Mechanically it conforms to the Mechanical Case Outline document. That document hasn't changed, though, it's from December 2004.
The datecoding system has changed from YYWW to YWW, where in the latter case the year is now indicated by a letter. C26 presumably means week 26 of 2013. This does not conform to the document.
The most striking differences are narrower leads (but spec allows for quite some variation: 0.93-1.48 mm) and the plastic body is no longer continuous near the leads but has "air gaps". The latter feature is not specified in the drawing.
A little correction to my previous post: 2012 can still be found "old style", the "new style" were on the market around the end of 2012. The very first I've come across had datecode B50 (week 50, 2012).
According to OnSemi, the MJL3281A is becoming obsolete (status: "last shipments", it's replaced by this MJL3281AG (status: "active").
The datecoding system has changed from YYWW to YWW, where in the latter case the year is now indicated by a letter. C26 presumably means week 26 of 2013. This does not conform to the document.
The most striking differences are narrower leads (but spec allows for quite some variation: 0.93-1.48 mm) and the plastic body is no longer continuous near the leads but has "air gaps". The latter feature is not specified in the drawing.
A little correction to my previous post: 2012 can still be found "old style", the "new style" were on the market around the end of 2012. The very first I've come across had datecode B50 (week 50, 2012).
According to OnSemi, the MJL3281A is becoming obsolete (status: "last shipments", it's replaced by this MJL3281AG (status: "active").
Last edited:
i would be suspicious of any devices not conforming to the official marking specs. counterfeiters take advantage of the ambiguities...... that's why so many counterfeits are labeled with the last valid date code of a particular part number..... find out directly from ON if you can what date code transitioned from the "A" series, and the "AG" series so others won't get burned.
I have no doubt that the new style MJL1302AG/MJL3281AG we've used over the past year (several hundreds) are genuine as they were acquired from an authorized dealer.
The last datecodes of the A-version we received were 1225. The first of the AG-version were B50.
The "new" datecoding system is now the same as seen on the smaller transistors (e.g. BD140, MJE15030 etc.), where there was already a single letter indicating the year.
These weren't the only transistors to change appearance, the MJE15030 and 15031 changed too.
The last datecodes of the A-version we received were 1225. The first of the AG-version were B50.
The "new" datecoding system is now the same as seen on the smaller transistors (e.g. BD140, MJE15030 etc.), where there was already a single letter indicating the year.
These weren't the only transistors to change appearance, the MJE15030 and 15031 changed too.
I recently bought 4x "2SA1124 PNP MATSUSHITA TRANSISTOR SC-51 -150V -.05A 1W ECB Transistor, Type: SI-P, Voltage: 150 V, Current: 50 mA, Power: 1 W, Frequency: 200 MHz"
from MCM Electronics. 3 of them yet correctly marked were NPN's!! Called customer service and they were happy to credit me for them, all $0.72, LOL! Of course, they don't carry them anymore. They were xrefs for c912 differential pair drivers in an old Kenwood amp I repaired. I know I should have replaced them both, but only one was noisy.
from MCM Electronics. 3 of them yet correctly marked were NPN's!! Called customer service and they were happy to credit me for them, all $0.72, LOL! Of course, they don't carry them anymore. They were xrefs for c912 differential pair drivers in an old Kenwood amp I repaired. I know I should have replaced them both, but only one was noisy.
that's rare, because MCM's parent company is Newark/Farnell, and usually they buy direct from the semi manufacturers.... unless they had those devices in some dark corner of the warehouse from before they were bought by Newark.... that was quite a long time ago, and before the time that counterfeits became a well known problem. there were lots of counterfeits floating around, but the problem had not yet been clearly identified.
I just hope the amp keeps working, cuz these transistors aren't anywhere.Well, 2sc3381 is 1.5$ each and 2sk389 is sold at 7-8$ each (or even more with a V suffix).
This dude saw my movies, one with the testing of a real K389V and one with his K389BL (which showed a NPN) and he still insists that he sells only tested stuff. Even if you are right, you are talking to walls, there is nobody listening to you.
The other side of the story is that although there exists ebay protection, you have to return the items to get a refund. If you are in western/northern Europe the prices are prohibitive for such a returned shipment. Meaning, lost money anyway, meaning... better avoid!
Well, I am not here to complain, but to open the eyes of others that may be hoping.
I've just been through this with my IXYS devices, lost the case on eBay, tried hard arguing with everyone that I should not send fakes back so they could be sold again. Of course I expected a refund all the same.
eBay makes that much money through some of the HK/China sellers that the fakes that do get sold are not worth bothering about no matter how hard you justify.
I even emailed IXYS about the problem expecting them to have some concern, they don't seem to care either!
So my fake IXYS devices that I did not want to send back for moral reasons, are going to end up for sale on eBay again, only I'll specify they are fakes... We'll see how eBay reacts to that... they have been so pathetic on this issue I'm having second thoughts about being honest myself on their website.