Sanity Check for RG Keen PNP Transistor Test

[wimacfee55]

Hello everyone!

I'm testing out Transistors for a Fuzz Face clone and I'm testing out some Russian Transistors before I lose my mind on why the Tagboard I made doesn't work. I am looking for a sanity check on the values I'm measuring for some MP39B and GT308V that I have.

I built the RGKeen Tester as listed here (http://www.geofex.com/Article_Folders/fuzzface/fffram.htm). After combing through the extensive bag log of Transistor tester posts here, I understand getting as close to possible to the values is key to good measurements. With that, here are the value measurements for each component.

2M2: measures at 2.21M on my Sperry DM-4400A
2.472K: a combination of 3 resistors, a 2.4k, a 33R, and a 48R that measures 2.47K on my Sperry DMM
A B1M Potentiometer is between the black lead of the battery and the common leg of the 2M2 and 2k4 resistors to get as close to 9.00V as measured.







For the MP39B, I am assuming the leg to the far right in this picture is collector, then base in the middle, and then emitter on the other side. This appears to be consistent with various datasheets online and in other form posts. Again, sanity check. Here is a photo of one of the transistors.




For the test itself: I verify the switch is open with a continuity check. I then test the voltage of the battery as the pot could shift during moving parts around. I measure by putting the red lead of the DMM on leg 2 of the Pot, and the black lead in the socket where the Emitter will go. I check for -9V to verify. I put the transistor in, then measure on the collector leg of the transistor and leg 2 of the pot. It's my understanding this measures "across" my 2.4k Resistor.

Black lead on the collector and red lead on Pot leg 2 I get -8mV with the switch closed.
I engage the switch, and measure -20mV.

Per the article, my gain should be 0.020V - 0.008V = 0.012V, multiplied by 100, I get a gain of 1.2

As per the wonderful post about measured gains here (https://www.diystompboxes.com/smfforum/index.php?topic=108054.0), this is laughably low. These values are consistent with the other 3 MP39B transistors I have.

Are there any glaring errors in my testing method above? Also, any additional tests I should run and if so what should they be? If not, I understand I have some garbage transistors and will order some from SmallBear to complete my JBF fuzz face as listed here (https://tagboardeffects.blogspot.com/2014/07/dunlop-jbf-3-joe-bonamassa-fuzz-face.html)

Thank you in advance!

[antonis]

For the MP39B, I am assuming the leg to the far right in this picture is collector..

Your assumption is wrong..



Extremely low current gain is an indication for reverse-active mode configuration..

P.S.

I understand getting as close to possible to the values is key to good measurements.

As long as your DMM is accurate enough..

[wimacfee55]

Agreed, always a question if the DMM is accurate!

With the picture above, the leg to the far right would be the collector, right?

[mozz]

Might be a dot or paint on the collector. Just reverse the leads, doubt you will fry it. Anyway, those transistors with the base connected to the case, may be noisy.

[Rob Strand]

I'm with Antonis.  (Maybe I'm not LOL! as I'm not sure if Antonis is assuming the view is from top or bottom)

[INCORRECT INFO]
The datasheet has this info.  I guess the thing to notice is the 3rd-angle projection implies the right view is from the top.

Possibly incorrect:


With my pic it makes the right lead in your pic the collector.  My understanding is Russian transistors put the dot on the emitter - also matching you pic.

I could have this all wrong!
[LIKELY CORRECT]
I did find some info implying C and E were flipped.

This one
https://www.cpearson.me.uk/2019/03/06/MP38A-transistor-pinout/

That means the datasheet pic isn't third-angle projection but first angle projection.

Corrected: (likely correct - see later posts)


The first time you build a tester the best thing you can do is verify it works by putting something known in there to see if it works correctly.  For example take a low-ish gain silicon PNP like a 2N3906 (or NPN if you know how to flip power) and put it in the tester.  You should get a ballpark gain of 100 to 300, and no leakage because it is silicon.  If you have a DMM that measures transistor gain you can get a better estimate of the gain with that.  If not you can set-up a test jig but unfortunately the simplest test jig is identical to RG's test jig!  There's other more devious tests you can do from outside the box but I won't go into them.

[antonis]

A DMM with diode test mode should unravel the mystery..

https://sparkrey.com/how-to-test-for-npn-and-pnp-transistors-using-a-multimeter/

P.S.
V_BE or V_EBfor Ge BJTs should lie in 150 to 300 mV range..

[Rob Strand]

A DMM with diode test mode should unravel the mystery..

That gives you the base but doesn't split between emitter and collector.   

However if you are take very careful measurements the emitter to base measurement will show a higher voltage than the collector to base voltage.   I actually don't know how reliable that method is for germanium.   It helps not to touch the part or leads as the heating will add uncertainty to the measurements.  If you are oblivious to the heating problem it can flip the identification of the pins.

In a normal universe the gain measurement is going to be higher when you pick correct collector and emitter  .

[antonis]

In a normal universe the gain measurement is going to be higher when you pick correct collector and emitter  .

In a normal universe the proper leakage test for Germs should be performed by using a hair-dryer..

[duck_arse]

somebodies learnt me a few things: the resistance measure of the base-emitter junction reads HIGHER than the base-collector junction.

and - looking onto the top of a transistor means you cannot see the pins. therefore, when drawing the pinout, solid lines indicate the things you CAN see, and dotted indicates obscured.  Antonis' excellent image shows only solid lines, so must be viewing looking onto the pins from below.

also, his diagram agrees with the pinout I always use for those parts.

A B1M Potentiometer is between the black lead of the battery and the common leg of the 2M2 and 2k4 resistors to get as close to 9.00V as measured

when you push the read button, a current will flow. that will induce a voltage drop, and your measured 9.000V will no longer be so. the original scheme shows no pot.

[Rob Strand]

and - looking onto the top of a transistor means you cannot see the pins. therefore, when drawing the pinout, solid lines indicate the things you CAN see, and dotted indicates obscured.  Antonis' excellent image shows only solid lines, so must be viewing looking onto the pins from below.

The datasheets I found are slightly different so I read that line as the center line (partial dash) for the centers of the pins, not as an underside line for a physical feature (if it was an underside line, it's not clear what the line would be!)

At the end of the day all the datasheets and drawings are not precise enough to be useful.

So,
- plug in a silicon transistor to make sure the hFE unit works at all.
- play with the C and E pins to find the highest hFE then that's what C and E are.

[Rob Strand]

Well Smallbear have some mp42b's with a nice pic, which is also clear about top/bottom view,
https://smallbear-electronics.mybigcommerce.com/germanium-transistor-russian-mp42b/

Their pic matches my second pic, the corrected one.

[wimacfee55]

Thank you for all the replies!

Based on the discussion above, I believe the leg I have painter's tape on in this picture is the collector:




This yields the same amount as before. Flipping it around (painters tape leg goes into emitter position) gives the same 0.118V reading with switch open or close so I'm confident that painter's tape leg is the collector now.

On Rob's suggestion, I tested a NTE158 I have from a previous build. 0.046V with switch open, 0.115V with switch closed for a gain of 6.9 (nice). Datasheet for NTE says an hFE of 50 is the minimum (https://www.nteinc.com/specs/100to199/pdf/nte158.pdf).

the B1M pot is to drop the power supply from 9.16V down to 9V to the 2M2 and 2k4 resistors. Not in the spec, but I saw in other forum posts that getting as close to 9V on the nose is important to the test.

I'm also getting a MP39B from Smallbear to check my sanity as well. The ones I have currently are straight from Russia so who knows.

For testing a Silicon NPN transistor, I'm using a voltage inverter to connect to the battery clip and when the black leg of the DMM is on the emitter socket and the red lead on pin 2 of the pot (for voltage control) I read +9V.

I am testing a BC183C, and luckily for TO92 package, it's easy to tell which leg is the collector! With the switch open I get a 0V reading (as expected for silicon!). Switch closed with black lead on collector and red lead on leg 2 of the pot, 0.0237V. Based on the BC183C datasheet here (https://www.mouser.com/ds/2/149/BC183C-310111.pdf) that's really low. Starting to think something might be wrong with the DMM...

EDIT: I just remembered those BC183C I got from a random online source years ago so again I may be in bad batch territory. I'm switching to a 2N2222A that I got from Tayda. I get 0.103V with switch closed and 0V with switch open. This is in range based on the datasheet here: https://www.onsemi.com/pdf/datasheet/p2n2222a-d.pdf

Think I bought from bad sources. Hence why this was a sanity check!

[mozz]

Just tested a few MP39, Peak DCA55,  according to your picture i get CBE so the opposite of what you just said. Gains were 69 and 100.

[wimacfee55]

Interesting mozz, so the leg with painters tape in my picture would be the emitter?

[antonis]

so the leg with painters tape in my picture would be the emitter?

Yes..

[mozz]

Yes..

Turkey makes me slow to respond.
 I don't see a "b" on mine but should be the same pinout.

[antonis]

Turkey makes me slow to respond.

I presume transistor testing isn't indicated during Thanksgiving..

[mozz]

Well i don't cook, the wife takes care of that. To stay out of the kitchen, i am populating some fuzz pcb's. I have a large collection of Russian transistors, many types. I used to buy 100's of them direct from a guy, but i lost his info.

 The gt308 the OP has are different pinout again.

[Rob Strand]

Based on the discussion above, I believe the leg I have painter's tape on in this picture is the collector

No, looking from the bottom the smallbear pic puts the emitter on that pin.

Just tested a few MP39, Peak DCA55,  according to your picture i get CBE so the opposite of what you just said. Gains were 69 and 100

So that would match the smallbear pic.

DIT: I just remembered those BC183C I got from a random online source years ago so again I may be in bad batch territory. I'm switching to a 2N2222A that I got from Tayda. I get 0.103V with switch closed and 0V with switch open.

You need at least one *known* good part.  If everything is in doubt you can't go forward.
What if the jig is measuring 10 times lower than it should?

[wimacfee55]

Yes, the GT308V has the red dot above the emitter and the collector in the center. That one I can figure out!

I did measure with correct orientation, got 0.118V with the switch open AND closed. Does that mean complete leakage on these suckers?

Based on the 2N2222a test, I think the jig is working. The BC183Cs we're off a random website from about 10 years ago.