faulty transistors

[mstechca]

I bought a 25-pack of 2N3904 NPN's and after using alot of them, I realized that all of them were DEAD, coming out of the package. There were two left. I thought I will use them as my varactors. What happened, they DID NOT WORK LIKE 2N3904's. These 2N3904's had black faces

I bought an 8-pack before it and the face of the transistor was coloured white. They worked beautifully in every case, including the varactors. Unfortunately I ran out, and I need more.

Before I buy transistors, can someone tell me the difference between a working transistor and a faulty transistor just by looking at it? because once I buy it, it is not refundable, and they don't allow testing in the store.

 

[Nigel Goodwin]

Before I buy transistors, can someone tell me the difference between a working transistor and a faulty transistor just by looking at it? because once I buy it, it is not refundable, and they don't allow testing in the store.

There's no visual way to tell - although I would have thought it's EXTREMELY rare to buy a transistor and find it faulty (I've never done so in almost 40 years of buying transistors).

If you're buying a mixed bag of cheap junk it's a different matter! - probably everything in it has already failed testing at the factory.

 

[Someone Electro]

You meaby got fakes?

 

[audioguru]

The 2N3904 transistors I got from Digikey all have Fairchild's logo, all have the same fairly new date code and all work fine with just about exactly the same spec's.
The 100 1N4148 surplus diodes I bought at the same chain as MStechca bought his transistors were all on a factory tape and also all work fine.
MStechca is an unlucky dude!

Here's the solution to his problems:

 

[Russlk]

2N3904 at Mouser are 5 cents US each, why buy junk? Part number 511-2N3904.

 

[mstechca]

I like to see the product before I buy it :wink:

Would you like someone from the other side of the earth to ship an "orange" to you after you paid in advance for it? :lol: :shock:

 

[audioguru]

MStechca,
Why do you think your new transistors are dead?
What tests did you do with them?

 

[JimB]

I like to see the product before I buy it :wink:

Would you like someone from the other side of the earth to ship an "orange" to you after you paid in advance for it? :lol: :shock:

A 2N3904 is a 2N3904 is a 2N3904. Once you have seen one you have seen them all.
I think we have already estabished that there is no way to tell visually if a transistor is good.
Obviously if the case is burned and cracked open it is duff, but good case is not the same as working correctly.

On the subject of oranges I will agree with you, however, in North East Scotland there are many shops selling oranges, the same is not true for a 2N3904.
So the next time I want a transistor, I guess I will just have to send away for it.

JimB

 

[DigiTan]

25 bad transistors is pretty darn unlucky. I'd definately stay away from the dog track if I were you; and gambling in general.

Are you 100% totally certain the transistors are bad? I had a batch of "grab bag" 170C's that I thought were broken. Turned out the β gain was just way lower than expected (my fault for not having a datasheet and using analog meters).

 

[audioguru]

Hi Digitan,
He is using the 2N3904 transistors as a varactor vari-cap diode. How can he possibly measure only 3pF???
Isn't 3pF about the capacitance between the wings of a bee?
His connecting wires must have at least 20pF between them that changes whenever someone walks by.

 

[DigiTan]

Good point. I guess the capacitance plot would also look at lot different up in the commercial broadcast bands. This is probably the times you'd need some super-expensive RLC meter.

 

[audioguru]

The only difference that frequency would make is in the inductance of the connecting wires at high frequencies. The factory uses only 1Mhz when testing capacitance to avoid the inductance of their test jig from affecting the results.

Just think about how high is the inductance of MStechca's long wires. The transistor-varactor will resonate with them and not be a capacitor anymore. Then he thinks the transistors are baaaad.
Maybe the leads on his "bad" transistors have a different inductance from the other brand he was using. They are guaranteed to operate as ordinary transistors aren't they?

 

[mstechca]

I'm not using the breadboard for my receiver. I'm using a circuit board with ridiculously short leads (< 2mm between the track and the case of the transistor).

 

[audioguru]

Then how did you test the "bad" transistors?
What is wrong with them?

 

[mstechca]

Then how did you test the "bad" transistors?
What is wrong with them?

I tested them by trying both transistors with different colours of faces.

I used two 2n3904's of the same face colour for my two varactor diodes. Both 2n3904's that were white faced worked normally. The 2n3904's that were black faced (even though they were connected EXACTLY the same way) failed to work.

 

[audioguru]

The capacitance of some 2N3904 transistors is too low to use as a varactor diode. The minimum capacitance isn't guaranteed. You wouldn't have this problem if you used a diode that is made to be and is guaranteed to be a varactor diode.

As I showed before, a 2N4401 transistor has much more capacitance with a greater range and I showed a circuit using them. Maybe most of them would work as a varactor diode.

 

[audioguru]

MStechca is re-inventing the wheel. This circuit uses three 2N4401 transistors in parallel to get enough capacitance for an 8Mhz range: