I was comparing the attributes of a Darlington Pair Vs. Szlikai pair and grabbed a 2N3906 from my small bag of "working" stock. I was having problems and in frustration (thinking I had maybe blown the xstr) put it in my component tester. First thing I noticed was the pin out was backward. OK, that explains a lot. The, I notice that it had a diode from collector to emitter! Huh? Then I saw that it has an hFE of 1! What? Unity gain/buffer! Pulled out my magnifying glass and checked the front and found all the markings had been sandblasted off (OK, I got it from AliX). Just what the heck is this thing. Don't remember ever seeing a BJT with an internal diode and unity gain before.

 

I think the tester is confused.

If the transistor is biased in inverted mode, beta would be low. You can verify if this is the case by swapping the collector and emitter terminals.

 

I think the tester is confused.

If the transistor is biased in inverted mode, beta would be low. You can verify if this is the case by swapping the collector and emitter terminals.

The tester doesn’t have dedicated terminals, that can’t have any effect.

 

I just checked the small bag of ~50 xstrs. About 1/3 marked KSL, 2/3 simply marked with 2N3906 and 3 number batch code. 9 had no markings or markings so scratched off as to be unreadable. All had hFE from 230s to 260s with Ic and Vbe in expected ranges with expected pin out. Looks like it must be a defective xstr unless it's something I never heard of. Anyway, it was a surprise to see it. Oh, and NONE of them are laser engraved so I expect all are home grown chinesium made. But jellybeans are jellybeans as long as they work. I WILL start checking transistors from AliX a lot closer from now on.

 

Just put it in my PEAK atlas DCA Pro tester and get the same plus quite a bit more data. It also says it is an CBE PNP with shunt diode but gives it an hFE of 2 along with many more parameters. But, it doesn't mark it as defective! So, I am still a bit puzzled as to just what it really is? Defective or...

 

Just in case, I tried it with a TL431CP. It reported an NPN transistor, with a diode, Cathode to collector, a gain of 1, and Vbe of 1.37V!

 

Could it be a MOSFET? They have a diode from drain to source.

 

Yeah, the only xstrs that I know of with a diode shunt are FETs but this IS (or at least it tests as on 2 different testers) a BJT? Hmmm... Let me throw it in a simple FET circuit and see what it does. I really suspect it is some kind of mfg defect... Problem is that the shunt diode appears between the collector and emitter which does correspond with (for example) the 2N700 Drain and Source so let me try it as a 2N7000 switch and see what happens. Nope, it simply acts as a diode (which the shunt is) and nothing on the base/gate except when I put + on the emitter with a red LED and 560Ω feeding it and the collector to ground, there is 115mV on the base/gate and the LED on. But, turned around, the LED is off and 4.5V is on the gate/base. So the gate/base IS active. But why is there a shunt diode across the collector to the emitter? It must be a defect?

 

I tried it with a TL431CP

That's getting close, except the "collector" has the anode side of the diode attached. That is a component I am not familiar with but is very close to what I am seeing here.

 

Well crap, my TC1 Multi-Function Tester died after years of faithful service. I guess it's telling me it's time to upgrade to the TC7 with the OLED screen and they are even cheaper than the TC1 now... On order!

 

I would never buy anything from a third world website.

 

I would never buy anything from a third world website.

I’ve bought a lot of excellent things from AliExpress. It’s also the case that AE hosts the official Espressif and LilyGo stores. The good to bad product ratio that I’ve encountered is about 50 to 1, and since none of my purchases to unknown sellers much exceeds $10, the risk is low.

Many of the things I bought would simply not be available to me if I didn’t by from AE. I am very happy with the results.

 

I only buy for hobby/educational use. When I do a project for something durable or a repair it's time to shop the major distributors. Jellybean transistors are cheap and work well. This is an ODD item that typically isn't found. In the last month, I've found FETs that were etched with the part number but were in fact BJTs. Now this, although checking the rest of the small parts bag, they are all BJT PNPs as they should be except this crazy unmarked anomaly.

 

I have seen Darlington's with diodes.

 

The testers recognize Darlingtons. Even my old tester says it's a PNP and shows the shunt diode.

 

Actually, I was referring to your comment that the only transistors that you are aware of with shunt diodes were FETs.

 

whovever programmed the testers recognized that they make PNP transistors with shunt diodes or it wouldn't be showing it or in the program's database of graphic figures to be shown. Just how it is or would be used is a mystery to me.

 

whovever programmed the testers recognized that they make PNP transistors with shunt diodes or it wouldn't be showing it or in the program's database of graphic figures to be shown. Just how it is or would be used is a mystery to me.

Protection diode when driving inductive loads such as relays.

 

Keeping in mind that a diode across a transistor can only protect that transistor and anything in parallel with it, unlike a diode at the inductor that can protect the entire circuit including the inductor.

 

Protection diode when driving inductive loads such as relays.

Typical part number so I can check their PDF?