Real Motorola 2N3055 or Fake?

[stockrock]

I acquired a small batch of 2N3055s from a reputable US seller, and they all test good on the DMM. But I'm not sure if they are real.  Any help appreciated.  I understand the date code is for either 84 or 74, week 43?  When I bought them, I was told they were 1974.

This is a pic - I sacrificed one and peeled the aluminum cap off to see the inside.

 

[scott2000]

Here's another thread talking about these......

https://groupdiy.com/index.php?topic=42602.0

 

[JohnRoberts]

They'll be harder to sell cut open like that...  :

Back in the 60s as a technician one of my jobs was cutting open to-3s like that to see why they failed (I would grind off the top with a grinding wheel). You can see visible failures like melted base wires or collector voltage punch through that would leave a small melted indentation in the collector region.

That die looks a little small (to me)  for an old old 3055, but if they say it's a 3055 sell it fast.. they're old and getting older every day (newer versions were physically smaller, and better). I do not understand the attraction for old, slow, inferior semiconductors.

JR

 

[stockrock]

I did see that other thread - I searched images and threads all over the web, but couldn't come to any definitive images or characteristics of the real thing, inside and out.

I've seen some saying an inside like this looks like the real thing, and others saying no, but maybe it's just a matter of what era. Not sure if this is an 84 or a 74, and I guess they would have slightly different characteristics, though I'm not sure what they would be.

The lettering does not rub off at all.

 

[scott2000]

I repaired an old Altec amp I have with what I'm sure were fake MJ transistors..... the lettering rubbed off with my finger :-\

It's still working years later....It's not mission critical in my scenario though.... 

In That thread about these, PRR  mentions a way to stress them to failure and there are other ways to check if you are really concerned....

Darn Counterfeits..... The bummer is that they'll sometimes work fine but not the same on some levels ....Sometimes you'll never know unless you test them properly and not just in the gear....

 

[zamproject]

Hello

little of topic

JR today

Back in the 60s as a technician one of my jobs was cutting open to-3s like that to see why they failed (I would grind off the top with a grinding wheel). You can see visible failures like melted base wires or collector voltage punch through that would leave a small melted indentation in the collector region.

JR in 2011 in the topic linked above

PS: Looking at those photos make me flash back to my first technician job back in the 60's where I often had to grind the top off of blown T0-3s to determine the failure mode. Typically a melted base wire, or voltage punch throughs between collector/emitter, that would leave a small melt mark where the short occurred. The 3055s were robust so harder to kill, but too slow to use in a switching supply like I was working on.  I don't miss the old days.

THIS  is consistency  and great memory 8)

I like when you start sentence with "back in the..." as most of your PS:...
always something to learn !

Best
Zam

 

[JohnRoberts]

Hello

little of topic

JR today

JR in 2011 in the topic linked above

THIS  is consistency  and great memory 8)

I like when you start sentence with "back in the..." as most of your PS:...
always something to learn !

Best
Zam

Sad that I'm repeating myself... but that happens a lot, more in the political threads.  :

Good that I keep saying the same things.

JR

 

[scott2000]

THIS  is consistency  and great memory 8)

I like when you start sentence with "back in the..." as most of your PS:...
always something to learn !

Best
Zam

[/quote

Only in trouble when it happens too close together....lol

For those of you who are just tuning in..........

 

[zamproject]

Sad that I'm repeating myself... but that happens a lot, more in the political threads.  :

Yep !!!

I enforced myself to not notice this side of the (hi)story, trying to show the bright side 
you put it front, self mockery is goodness 

Cheers !!
Zam

 

[Gene Pink]

I acquired a small batch of 2N3055s from a reputable US seller, and they all test good on the DMM. But I'm not sure if they are real.  Any help appreciated.  I understand the date code is for either 84 or 74, week 43?  When I bought them, I was told they were 1974.

This is a pic - I sacrificed one and peeled the aluminum cap off to see the inside.

I would bet "fake".

Like JR says, that die is tiny, and the lead-in wires seem too small to handle 15 amps.

I just went through my 2N3055 box, there were six motorola's, pic attached. These are originals, as I have been stripping stuff since the '60's.

Note 4-digit date codes, and the "3" in the number is a different font than yours. Also, stamped numbers around the mounting holes.

Your topless one is actually a good mod, more surface area for air cooling. 

Gene

 

[stockrock]

What part is the "die" that is tiny?  The small part with the lines on it?

Does anyone have a picture of the insides of a known real Motorola 2N3055?  There's that one on Wikipedia that's been cited on many threads, but it seems that's a very old one, and not necessarily representative.

I have one true Motorola 2n3055 still in original packaging from 1976 - pic attached - you can see the font is different, and it does not have gold pins. But it is definitely a Motorola. I won't be sacrificing this one.  Mine are from 64 or 74 or 84, so not sure how that would be different from a 1976.

 

[Tubetec]

I found an excellent document on the venerable 2n3055 in IEEE ,gave full details of the different die's used and the approximate dates for production starts and finishes of the various kinds. It also explains in depth the different processes used to make them .
I see four digit date codes on any motorolas I have here too . From memory sometime in the mid 70's they moved to the smaller die although some smaller producers may have continued with that a bit longer .There seems to be some suggestion that the older types are a little more robust and less prone to meltdown due to lower current density ,technically the performance of the later ones is better in frequency terms at least.
I didnt officially have access to IEEE ,but some person posted login details for the site online ,and I was able to nip in and out to get the document. I'll have a look later ,I should be able to find it and post it up here. Probably worth posting in the documents section too actually .

 

[stockrock]

That would be great, thank you - I found "The 2N3055: A Case History" from 2001 on IEEE, but it does not have this information.

 

[JohnRoberts]

What part is the "die" that is tiny?  The small part with the lines on it?

Yes the die, is the actual transistor device. The entire bottom is the collector contact that bonds to the header (or case). The interdigitated pattern on top is the outline of the base region, located inside the emitter region.  I can clearly see the base wire connected to the top of the base region, the emitter wire looks a little odd but it has been decades since worked with these. In fact not 2n3055s even back in the 60s because they were too slow for the new DC to DC switcher I was working on. 

Does anyone have a picture of the insides of a known real Motorola 2N3055?  There's that one on Wikipedia that's been cited on many threads, but it seems that's a very old one, and not necessarily representative.

Huh.... What do you mean "real".  If you are talking about original technology 3055 it must be an old picture because they stopped using the old technology in the mid 70s. Since then it has probably been upgraded multiple times.  Modern 3055s are smaller because they can... the size of a transistor die is the major cost factor so die shrinks are always desirable to reduce manufacturing cost.

I repeat the old technology 3055 is slower than dirt. Perhaps a candidate for some semiconductor museum, but not for use in modern circuits.

The newer versions are not fakes, they are actually better parts, but the iconic 2N3055 part number remained popular for years (decades?) after there were newer better options. I have even seen a MOSFET power device use the 3055 number hoping to sell more units (to old engineers). 

JR

 

[stockrock]

I jacked up one of the wires when I was peeling the top off.  So that may be why it looks off.

I'd be using these in Neve 1073 clones, so that's the reason for my interest. I have the correct vintage transformers, so trying to get this right too.

 

[JohnRoberts]

I jacked up one of the wires when I was peeling the top off.  So that may be why it looks off.

Thanks that explains it... should be bonded to the emitter region similar to the base wire.

I'd be using these in Neve 1073 clones, so that's the reason for my interest. I have the correct vintage transformers, so trying to get this right too.

This is pretty well explored...  I expect the slower part's transfer function  could be mimicked with other circuit compensation tweaks, but I still do not understand the powerful attraction. At the time these (old) units were designed, the design engineers were not using slow transistors on purpose for some specific target response. That was just the parts they had available at the time and they were most likely making the performance as good as they could.

Enjoy and do whatever floats your boat... This is not a new pursuit so I expect any original 2n3055s have been searched out, and resold for ridiculous prices. There may be a few isolated parts in old junk drawers (no I don't have any, but did back then).

JR

 

[rackmonkey]

Got me curious about this, so I ran some curve traces of Ic/Vbe for 3 batches of 2n3055s. One, a modern batch of ON Semi parts from Mouser; a batch of real, old Motorolas (some pulls from 70s gear and some that were from an ancient box of replacements from a retired tech I know); and a batch of suspected fake RCAs I bought from eBay several years ago. The RCAs look like brand new, shiny TO-3s with no oxidation on the case at all. I don't think RCA branded semiconductors were made after the late 80s, so it seems fishy.  Traces attached.

I'm confident in the genuine-ness of the On-Semis and Motorolas I have, so maybe compare your batch to those curves if you have a tracer.

 

[scott2000]

Make a solar cell with them??? Never thought of that......

just 5 of them in series will power a small 1.5 volt calculator in good indoor lighting conditions.

https://rimstar.org/renewnrg/make_solar_cell_using_transistor_2n3055.htm

 

[PRR]

> 2N3055s ... I'm not sure if they are real.

Real "what"?

What are you poking at? A Neve, or a power amp?

IMHO, a 3055 is a baked-rock from RCA. Always in steel. And all the Motorola "3055s" are wimpy planars. Which moreover are easily "duplicated" by any planar foundry with a silkscreen.

What did Neve use, the RCA or the Motorola?

> I ran some curve traces of Ic/Vbe

I don't even see a point in testing a 15 Amp 3055 to 10 or 11 mA. OK, you did find two "families".

On the original baked-rock process you would find 10mA Vce all over the place. The leakage could be some mA. Above that it would trace the exponential you show, but shifted by effective Area which was not well controlled.
Then it would rise Ohmic from all the parasitic resistance. The RCA datasheet shows obscene Vbe at high current, because nothing was well controlled and they just sorted-off the ones that sucked real bad. OTOH in the Planar process the area is controlled by the mask and planar rather demands good control of doping, so parts come out pretty consistent.

If you want the Motorola because Neve used Motorola, IMHO use any Silicon NPN in a TO-3 can. The heavy overall NFB and bias-trim will swamp-out all variation. But I'm not a Golden-Ear. And if I were, I'd start by finding Neve's old part-bins or known-unmolested examples for comparison.

 

[rackmonkey]

I don't even see a point in testing a 15 Amp 3055 to 10 or 11 mA. OK, you did find two "families".

Well, in the context of what the OP is asking about (vaguely, Neve line driver, I assume the BA-283), why would we care about testing at the upper 3/4 of the 3055’s capability? IIRC, we’re talking about a ~60 - 70mA bias point, and swings within the limits of a 24V power supply in the BA-283. Okay, we’re not even there with the curves I ran (which were at the limits of the tracer I used), but it’s not totally out of the ballpark of characteristics in that context. And as you acknowledge, it did show 2 distinct “families” of devices. Can’t draw conclusions beyond that, but if his goal is to reproduce that unit as closely as possible (notice I didn’t say “as closely as ‘necessary’), there’s utility in the exercise, is there not?

I can’t recall where I read it, but someone out on the interwebs  did some tests on the differences between the early homotaxial 3055 units and the later ones, and documented a 20-fold (ish - again, IIRC) higher base to collector capacitance in the old homotaxial units. It’s conceivable that in some circuits (no idea if the 283 is one of those) that one could end up with parasitic oscillations (the dreaded “fuzzy worm” sine wave) without external compensation. You can’t always hear those, so it might not be just a golden ear waste of time to look into this. Just a thought.

BT