from the WTF is this thing, and why does it exist department revisited

[pinkjimiphoton]

so i had picked up a couple sacks of random transistors a few years ago, and was going thru them looking for something "special <hah!!>" in a fuzz circuit, and i found a sack of
2sc2352's. did a google. found the following

Type Designator: 2SC2352

Material of Transistor: Si

Polarity: NPN

Maximum Collector Power Dissipation (Pc): 0.25 W

Maximum Collector-Base Voltage |Vcb|: 30 V

Maximum Emitter-Base Voltage |Veb|: 4 V

Maximum Collector Current |Ic max|: 0.03 A

Max. Operating Junction Temperature (Tj): 125 °C

Transition Frequency (ft): 850 MHz

Forward Current Transfer Ratio (hFE), MIN: 100

they appear to be around 100-130 hFE on average. cool by me.

but... on the tester, they show a diode between e and c, with the cathode band of the diode towards the collector.

i'm like... vass ist das?

never encountered this one before... what is it, why is it, what do you call it? any advantages to this for our diy purposes? any way to abuse it for fun and profit? 

education gratefully sought!

[R.G.]

By any chance, are the base and emitter shorted together?
There are a few possibilities. Maybe a very tidy person gathered up and bagged a whole bunch of similarly dead ones.

Does your ohmmeter, when set to diode test, show the base and emitter as having a diode-like characteristic?

Is the pinout what you think it is? The "emitter" that was really the base would cause something like this, as the base-collector of an NPN is supposed to look like a diode with cathode to the collector pin.

Possibly they are miss-marked and are not really that type transistor.

Just speculating.

[Rob Strand]

I don't know what's going on but from here (and your list),

http://www.datasheet.jp/pdf/1030916/2SC2352.html

VEBO is quite low. 

If your DMM diode test works at a high voltage than normal, for example it can test white LEDs, then perhaps the voltage is high enough to cause a reverse current.

If you have an ancient 3 1/2 digit DMM with a diode tester maybe try it out on that.  The voltage on those things was usually low.

The diode voltage you are seeing might be useful.

You can always try your luck and see how one of these biases up,
https://www.muzique.com/schem/multi.htm

[pinkjimiphoton]

By any chance, are the base and emitter shorted together?
There are a few possibilities. Maybe a very tidy person gathered up and bagged a whole bunch of similarly dead ones.

Does your ohmmeter, when set to diode test, show the base and emitter as having a diode-like characteristic?

Is the pinout what you think it is? The "emitter" that was really the base would cause something like this, as the base-collector of an NPN is supposed to look like a diode with cathode to the collector pin.

Possibly they are miss-marked and are not really that type transistor.

Just speculating.

from the little i could glean, it was used in the vhf amps on televisions quite a bit...
i didn't test the junctions with my multimeter, just one of them little chinese component testers.
i'd never seen a diode internally like this before show up on the little led screen!

i tried one in one of the swingerbox fuzzes i've been messing with, and it actually sounded fairly good and fuzzed ok, had an hfe around 130 or so.

i will try and hit it with the meter and report back later on today.

the mystery deepens, when RG hasn't seen it, ya KNOW its a different kettle of fish!

on a personal note, i hope all is well with you and yours, man. trying times we live in!

[pinkjimiphoton]

I don't know what's going on but from here (and your list),

http://www.datasheet.jp/pdf/1030916/2SC2352.html

VEBO is quite low. 

If your DMM diode test works at a high voltage than normal, for example it can test white LEDs, then perhaps the voltage is high enough to cause a reverse current.

If you have an ancient 3 1/2 digit DMM with a diode tester maybe try it out on that.  The voltage on those things was usually low.

The diode voltage you are seeing might be useful.

You can always try your luck and see how one of these biases up,
https://www.muzique.com/schem/multi.htm

thanks rob, hope you've been well too...

so weird. why would they want it so low? seems like they may have designed these to be fuses or something?

i will get the stats off my little tester and post them later, and try hitting it with my meter and see if i can find some info that maybe will help.

man! i didn't think this was gonna stump anyone but me!

this is what it looks like on the tester... its in the self contained database, so its gotta be a somewhat "normal" part.

[anotherjim]

Would an RF part maybe need some flyback diode protection from a coil? Or, there have been a few proprietory transistor fabrication methods and this could be an accidental body diode like MOSFET's have.

[Rob Strand]

thanks rob, hope you've been well too...

Yep, same old ...     Hope things are going OK for you too.

so weird. why would they want it so low? seems like they may have designed these to be fuses or something?

RF transistors tend to make VBEO (the VBE breakdown) lower.   I guess that transistor is moving towards RF.

i will get the stats off my little tester and post them later, and try hitting it with my meter and see if i can find some info that maybe will help.

man! i didn't think this was gonna stump anyone but me!

this is what it looks like on the tester... its in the self contained database, so its gotta be a somewhat "normal" part.

There's two possible cases:

- the CE diode is really there and the diode forward drop is 0.7V or so.
  Here a DMM diode test will show the diode.
  There are real parts like this but I wouldn't expect that part to be like it.

- The tester is showing  a false part pic.
   The low VBEO is causing the tester to see a CE test current,
  it then brain-farts and shows the closest pic it knows, which is
  a pic not representing what's happening.
  A DMM diode test should show no conduction between CE
  in both directions.

Here's a summary of what goes on,

[pic server not working maybe it will come back later.]

[pinkjimiphoton]

Would an RF part maybe need some flyback diode protection from a coil? Or, there have been a few proprietory transistor fabrication methods and this could be an accidental body diode like MOSFET's have.

yes, i think that does have something to do with it, on some page i found yesterday surfing looking for info

[pinkjimiphoton]

thanks rob, hope you've been well too...

Yep, same old ...     Hope things are going OK for you too.

so weird. why would they want it so low? seems like they may have designed these to be fuses or something?

RF transistors tend to make VBEO (the VBE breakdown) lower.   I guess that transistor is moving towards RF.

i will get the stats off my little tester and post them later, and try hitting it with my meter and see if i can find some info that maybe will help.

man! i didn't think this was gonna stump anyone but me!

this is what it looks like on the tester... its in the self contained database, so its gotta be a somewhat "normal" part.

There's two possible cases:

- the CE diode is really there and the diode forward drop is 0.7V or so.
  Here a DMM diode test will show the diode.
  There are real parts like this but I wouldn't expect that part to be like it.

- The tester is showing  a false part pic.
   The low VBEO is causing the tester to see a CE test current,
  it then brain-farts and shows the closest pic it knows, which is
  a pic not representing what's happening.
  A DMM diode test should show no conduction between CE
  in both directions.

Here's a summary of what goes on,

[pic server not working maybe it will come back later.]

i will try testing it with my meter when i hit my dungeon a little later... but i found this:

https://www.eevblog.com/forum/projects/npn-power-transistor-with-built-in-diode/

They are not uncommon in horizontal output stages of CRT type TV sets & Monitors,& in the SMPS of such devices.
It's too long since I've worked on them so I'll have to look up a reference.


High voltage NPN transistors with a C-E diode were fairly common in the horizontal output stages of TV sets in the last 20 years or so of the CRT era. They often included a 50-100 ohm resistor from B to E, as well.  Typical type would be 2SD2586, but there are MANY others, both TO-3 metal and various plastic packages.

The diode is referred to as a "damper", to protect the transistor from excessive voltage spikes when switching inductive loads, like transformer windings...


Quote from: Alex Eisenhut on February 16, 2014, 02:06:37 am

    but the big question is what is the job of these C-E diodes

With the two transistors stacked one above the other, the diodes stop the halfway point that drives the transformer going any higher than the top collector or lower than the bottom emitter. If the diodes were not present, if for example the voltage swung too low it would forward bias the lower C-B junction and consequently reverse bias the the lower B-E junction. The B-E junction will only tolerate about 5 volts or so before it begins to avalanche like a zener, and push it too hard and ->  :-- Short answer, it protects the transistors.

With the two transistors stacked one above the other, the diodes stop the halfway point that drives the transformer going any higher than the top collector or lower than the bottom emitter. If the diodes were not present, if for example the voltage swung too low it would forward bias the lower C-B junction and consequently reverse bias the the lower B-E junction. The B-E junction will only tolerate about 5 volts or so before it begins to avalanche like a zener, and push it too hard and ->  Short answer, it protects the transistors.

Quite correct. The diodes are fast recovery and clamp any voltage over/undershoot to the rails. All MOSFETs have them inherent in their structure so no additional diodes are required.
The reason bipolars are used as apposed to MOSFETs is that they are a fraction of the price of a MOSFET since the amount of silicon used is much less.
Switchmode converters using bipolar transitors need to be carefully designed to ensure that the device doesn't experiance second breakdown under certain circuit conditions. Sometimes even when the device is used within its safe operating area some devices seem to fail without warning.
The mje13005D is very common so shouldn't be a problem obtaining.

so apparently, its there to protect the transistor from blowing out. wondering if it could be combined with a diode going the other way to make a clipping stage   

[Rob Strand]

so apparently, its there to protect the transistor from blowing out. wondering if it could be combined with a diode going the other way to make a clipping stage   

Yep, those are the cases where I wouldn't be surprised to see the diode.    I've got about 10 or 20 of those pulled from old monitors.  They are TO-220 package or larger.   Pretty common to see the diodes on TO-220 package or larger Darlingtons. 

For small stuff not common at all.   If you need protection in those cases you normally add your own diode.  Some of the earlier TVs added their own diode on the larger package parts before the built-in diode stuff was available.

That's why I'm thinking tester weirdness.   

RG's idea that a few parts could be fried is on the list as well.

[m7b52000]

This may have no relevance at all but my Russian IT308 PNP usually test like this

[pinkjimiphoton]

same thing, but pnp.... mine are npn, which is why the diode points the other way!

[pinkjimiphoton]

so i tested it last nite with my meter. i didn't write all the results down, but it DID INDEED work out to having a diode reading around .760 or so on my multimeters between e and c, and the rest of the diodes read the expected .716 or whatever. reversing polarity, no readings, so there's definitely a diode internally.

here's a shitty phone shot of it in my crappy little chinese tester showing it matches M7b52000's other than being npn rather than pnp

[Rob Strand]

so i tested it last nite with my meter. i didn't write all the results down, but it DID INDEED work out to having a diode reading around .760 or so on my multimeters between e and c, and the rest of the diodes read the expected .716 or whatever. reversing polarity, no readings, so there's definitely a diode internally.

That's pretty weird.   There's not a hint about a diode present on the NEC 2SC2352 datasheet.   The gain you are measuring doesn't look unreasonable for a 2SC2352.

Even if we suppose it's not a 2SC2352.  It's going to be tricky finding a small transistor with a built-in diode..

I was pondering it could be one of these "digital transistors".  All that means is it has resistors on the base to take digital signals directly.


Those will measure a diode in one direction between C and E.   The thing is though the base resistors probably won't give you the near equal (which is normal) BC and BE diode drops.  Also the BC & BE diode drops will be probably larger than the any fake CE diode drop.   These things will show a resistance across BE in the reverse direction instead of an open reverse diode.   They probably won't show a gain or VBE on your tester.

So none of your measurements give any evidence it's one of these.   So bin that idea.

Let's say there is a diode and it might not be a 2SC2352.   The diode isn't going to affect most amp stages so if it biases up OK and sounds OK (and perhaps not noisy) then it's still a usable transistor.  We don't need to care what the label says.

Here's an example of a TO-92 with a diode, STL73D,
https://www.st.com/resource/en/datasheet/CD00216340.pdf

Crazy low hFE on that one (it might be higher at low currents).