Fakes Toshiba 2N3773 photos ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Is a 2N3055 and MJ2955 from toshiba manufactured in 2001 safe ?
I have a pair .
like this from jpg attachement?
check out also this:
http://www.diyaudio.com/forums/swap-meet/125248-fs-15-pcs-genuine-toshiba-2n3055.html

I am looking for an opened 2N3055 pic, first time launch from RCA between 1963 and 1968 so as the first RCA advertisement about the 2N3055 for the thread
http://www.diyaudio.com/forums/soli...-models-quasi-complementary-power-output.html

BTW - who was the developer of the 2N3773? First release from RCA or Motorola?
 

Attachments

  • Toshiba 2N3055.jpg
    Toshiba 2N3055.jpg
    59.1 KB · Views: 818
  • Toshiba 2N3055 open.jpg
    Toshiba 2N3055 open.jpg
    64.5 KB · Views: 814
Last edited:
like this from jpg attachement?
check out also this:
http://www.diyaudio.com/forums/swap-meet/125248-fs-15-pcs-genuine-toshiba-2n3055.html

I am looking for an opened 2N3055 pic, first time launch from RCA between 1963 and 1968 so as the first RCA advertisement about the 2N3055 for the thread
http://www.diyaudio.com/forums/soli...-models-quasi-complementary-power-output.html

BTW - who was the developer of the 2N3773? First release from RCA or Motorola?

It was nothing like those .
 

Attachments

  • C360_2011-05-12 13-44-52.jpg
    C360_2011-05-12 13-44-52.jpg
    830 KB · Views: 733
  • C360_2011-05-12 13-44-59.jpg
    C360_2011-05-12 13-44-59.jpg
    953.7 KB · Views: 648
Here's another look at this 'mystery'.
While it is entirely possible (per AndrewT) that Toshiba did make the old hometaxial 2N3773, they did not make the faster 4MHz version. Instead, they concentrated on their own designs and produced some rather desirable (for the time) audio power transistors. Still, the 'Toshiba' 2N3773 are around even today, and may be an unique meme, a fake of a fake so to speak. THe following might offer some clues:

I currently work for a small company that is nevertheless a successor of a larger one that has been in the PA business for 30+ years, and in the 80s sold a lot of amps based on a quasicomp construction with an op-amp front end, using 4-6 pairs of 2N3773 (the 4MHz version). Every once in a while some of these come back aven today, and there are still some 2N3773 from those days lying atound in the parts bins, mostly the Toshiba brand. How so? Well, these were the only ones that always worked, and worked very well. Here is how they look:
top.jpg bottom.jpg
Note the large heat spreader on the bottom. If this is not present, or there is a small round one, you are likely to have problems and the transistor is indeed a fake. So, is this one a fake? To answer this question, I did some measurements on a curve tracer, compared to a genuine ONsemi 2N3773.
First some low current tests, also at high voltages to test for breakdown voltage, original vs 'Toshiba'
Original_low.jpg Toshiba_low.jpg
It can be seen that the original 2N3773 has a huge variation of current gain both with current and voltage. At small currents (100mA full scale) beta ranges from 70 at very low Vce, to several hundred at high Vce. This particular part is actually one of the better ones and has a rather high breakdown voltage, probably because it's a newer production part (1998), so the process was very mature and optimized by that time. 140V Vce0 is guaranteed for 2N3773, here it's almost 200V. On the other hand, the 'Toshiba' part shows a completely different story. The breakdown voltage is lower but still over 160V, but the gain is very constant both with current and voltage. The gain is also much lower at around 60 or so.
Here is how things hold up at progressively higher currents, again original on the left, 'Toshiba' on the right:
Original_mid.jpg Toshiba_mid.jpg
Full scale current is 1A on these graphs. Pretty much the same story is revealed, the 'Toshiba' holds a fairly constant beta with negligible influence of Vce. Finally, a high current plot, 5A full scale (note lower Vce, up to 70V or so to limit disipation and avoid secondary breakdown):
Original_high.jpg Toshiba_high.jpg
The original has higher beta, therefore less lines in the display, however it drops drastically as current increases. The 'Toshiba' part however gets doem to it's low vurrent gain of 50-60 at 2.5A or so and at 4.5A it's still about 25, again very little change with Vce. Another feature that can be seen on the graph is that the lines are much narrower on the Toshiba graph, and doubled on the original one - this phenomenon is called 'looping and occurs due to the Cbc capacitance ('Miller' capacitance) influencing base current. The Toshiba part has much lower capacitances. In fact, it should be oted that this particular original 2N3773 is a very good one, as I have others that I have measured which show even higher capacitances (on the order of 3-4 times the Toshiba part) and lower and even less sustained beta (there is actually quite a variance, about 1:2.5 between the ten or so genuine 2N3773 I have). On the other hand, I have 16 'Toshiba' 2N3773 of the construction as shown in pictures above, and they show a variance of beta between 50 and 70 on average, and next to no variance in breakdown voltage.
The TOshiba part shows considerable looping at low currents and voltages in the first set of traces, however so does the original 2N3773, but it's not easily visible due to the slanted curves. The looping is present because of low base current combined with a large collector voltage swing, on a device that has non-trivial Cbc. That being said, it's definitely a much faster device compared to the original.

Now, the $1M question is, which one would you rather have in your amplifier?
The Toshiba shows far better beta sustain, better speed, no discernible Early effect. So what actual transistor is this? Well - it measures suspiciously like a 2SD551 or similar. It is also very robust, took 1.5A at 100V for a second without a problem (heated up like crazy though).It is also definitely a genuine Toshiba style TO3 case, rounded top and large heat spreader.

What can be surmised from this? It is very possible that at some point someone re-labeled genuine Japanese transistors that were overstock, and obviously these could very much be a success. So, as stocks dwindled, it's plausible that fakes of the fake appeared. In fact, i know for a fact there is an epidemic of re-labeling overstock transistors in China, tons of these can be had on eBay. The sad thing is, some are genuine high-spec parts from Sanken, NEC and Toshiba, as can be seen from the typical (and expensive) case construction (Heat slughs, spreaders, or thick copper base Sanken style), but the original markings have been grinded off with no hope of finding out what these actually are. I've recently had 13 parts shipped with the same marking, but 2 completely different case styles and 7 different actual parts shown by the curve tracer (one was a high voltage low beta part with 3x the Vceo, another a darlington with 20x the gain).
That being said,if I was offered more of this particular part, I'd gladly use it. It must be the worlds first case of sticking with the real fakes rather than the fake fakes :p
 
That is the 2SD873. Re-branded as a 2N3773, presumably to sell into markets which called for the Motorola device. And as you've found out, meets the 2N3773 spec. The batch you have even came from the same lot as the ones I had (8714) and they worked perfectly. This was back when I used D424's and 3773's pretty much interchangeably, depending on which one MCM had in stock that day. And oh yeah - the B554 was SOOOO much easier to find than the 2N6609 and cheaper than the MJ15025 (another big plus for Toshiba).
 
That is the 2SD873. Re-branded as a 2N3773, presumably to sell into markets which called for the Motorola device. And as you've found out, meets the 2N3773 spec. The batch you have even came from the same lot as the ones I had (8714) and they worked perfectly. This was back when I used D424's and 3773's pretty much interchangeably, depending on which one MCM had in stock that day. And oh yeah - the B554 was SOOOO much easier to find than the 2N6609 and cheaper than the MJ15025 (another big plus for Toshiba).

Possible but not sure - I have 16 of them with different date codes. It doesn't look like a 200kHz Ft device on the tracer, though...
 
It's not. My data book said 4 MHz. Anything above 200KHz passes the 2N3773 fT spec (about the only thing that won't is germanium). Remember - changing a spec on a JEDEC device - or worse yet - JAN - takes an act of God to get the paperwork through.

Hum, I only found a spec for a clone (Inchange Semi) that says 1.5MHz and a databook excerpt that says 200kHz... In nay case, those curves are NOT a hometaxial device and the lack of severe beta drop is the clue. Also, transistors used for any sort of switching do not have a pronounced beta drop at low Vce as this would require an ever greater base current to saturate the device as Vce dops, pumping more charge that also later needs to be sucked out = slow and inefficient switching. Looks more like a regular epitaxial diffused with ft in the MHz as you state. Certainly far more usable than the original 2N3773.
 
Real or Fake ??

I recently bought a few 2N3773 with Toshiba branding.

Being suspicious, I cut one open.

The die size looks correct.
The heat spreader is also a good size.

The die was covered in a white semi-solid goo.

Question :- Real or Fake ?


Andy
 

Attachments

  • 2n3773.jpg
    2n3773.jpg
    128.7 KB · Views: 403
That is a "real" 2SD873. The innards look exactly like the Toshiba "2N3773's" I have. And they work perfectly. That is, like an epitaxial 2N3773. I've used them in pairs on +/-55V and threes on +/-70 and they survive. The +/-70 was with a 4 ohm sub. That amp is still in service at a friend's house in Florida. The white goo is what they used before they perfected the passivation process. Most all real Toshibas from that era used it - including the D424's. Fakes still use it because the junk dies they use in fakes may or may not have any passivation coating at all.
 
I used 2N3773 in my 300W/4r Bass amps since late 70's (more than fast enough in that use ;) ) had to absolutely stop in the early 2000 because of absolutely chaotic fake situation.

The breaking point was when an until then reputed serious seller offered me a pack of the triple thickness iron base ones ... no coopper heat spreader die inside of course.

Had to redesign with IRFP240/250 which being used in Industrial stuff, unglamorous and cheap, are bad business for fakers.
 
TSL 2N3773 transistors

I am always on the hunt for genuine 2N3773 transistors to build or repair amplifiers. I won't buy the obvious Chinese fakes on eBay, its just not worth it no matter how cheap and will almost certainly end in tears or more likely smoke. With "genuine" 2N3773 available costing anywhere between £4 and £20 each, I also don't want to get ripped off on price either. Recently, I bought a batch of 12 TSL 2N3773 from a private seller on eBay which are also available elsewhere on eBay priced at around £1.60 each. Maybe the seller thought these might be fakes and just flipped them back on the bay.


The transistors look to be NOS, with an oxidised, aged appearance, somewhat consistent with their 1996 data code. I tested them with an Atlas DSA55 semiconductor tester and within the limitations of the tester, the Hfe and Vbe seemed to be reasonable, with 4 indicating Hfe less than 10 and the rest up to 25. Next, I tried to rub-off the markings but only had methylated and white spirit to hand and as the picture shows, after some seriously hard rubbing, the markings did start to disappear but this was not conclusive.

There was only one thing left to do, I picked a low gain transistor and removed the top with a metal saw (see pics). It did appear genuine based on 1)the large die size 2)the double bond wire on the emitter 3)the alloy paste die mounting onto the plug 4)the clear epoxy die encapsulation layer. These are things that a fake transistor manufacturer would probably not bother to do, as the assembly costs would be so high they may as well make the genuine ones;) As a side note, I have an early hand-built Crimson Elektrik amp from the late seventies and one channel was repaired around 1993 with similar TSL 2N3773 transistors, albeit with different marking style. I have yet to properly test the transistors in a working amplifier but welcome any comments from others who may have knowledge of them or used them already.
 

Attachments

  • RIMG0629.JPG
    RIMG0629.JPG
    671.5 KB · Views: 212
  • RIMG0628.JPG
    RIMG0628.JPG
    655.4 KB · Views: 196
  • RIMG0627.JPG
    RIMG0627.JPG
    647 KB · Views: 215
  • RIMG0626.JPG
    RIMG0626.JPG
    654 KB · Views: 210
  • RIMG0631.JPG
    RIMG0631.JPG
    647.8 KB · Views: 195
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.