# Consideration when switching from 2N3904 to MMBT3904

I created a (very) basic circuit with:

simulate this circuit – Schematic created using CircuitLab

On the test board, with 2N3904 transistor, VM1 show ~12V and ~6V when SW is pushed (as expected).

When trying to go to SMD components, I am using MMBT3904, but it seem to behave differently: Showing about 6V indifferently of the state of SW1.Additionally, the collector voltage seem to contaminate the base voltage.

I tried several options:

• Minimized soldering temperature up to minimum (270º) about a few seconds to solder. Is it still to much?
• Created a mini pcb only for the transistor, so all the other circuit is exactly the same.

The question is: is there special considerations to switch between 2N3904 to MMBT3904, either in the circuit or the soldering process? What am I doing wrong?

EDITED:

On request, here is a a picture. I created the transistor PCB about 1h ago.

EDITED 2:

• Guessing the pinout isn't as expected? Commented Oct 22, 2016 at 1:45
• The circuit SHOULD work perfectly. That is why people are guessing that you may have made a mistaken assumption regarding pinout. Commented Oct 22, 2016 at 3:02
• Isn't 12V too high for the 2N3904? The datasheet states that the absolute maximum Vbe is 6V. Commented Oct 22, 2016 at 3:07
• @electrophile that's without a series resistor. With one, the bulk of the voltage should be across it and not the vbe junction. Commented Oct 22, 2016 at 3:29
• Just to clarify something, Adrian. When you read the datasheet as saying that the absolute maximum of $V_{BE}$ is $6\:\textrm{V}$, I suspect you may have not noticed that the datasheet parameter is $V_{EBO}$ there. Note that E comes before B, there. This is the reverse voltage across the BE junction. Not the forward voltage. The junction will start to zener and avalanche near that spec.
– jonk
Commented Oct 22, 2016 at 5:05

Here is a comparison of the pinouts:

Maybe you assumed the pinout was E-B-C on the SOT-23 package. Sounds like you have a forward biased junction between what you think is the collector and what you think is the emitter, so the base must be connected where the collector should be and either the emitter or collector where the emitter should be.

So either base and collector are swapped, or all three are mixed up.

• Actually, this is one of the first thing I checked, so, in your example, I follow B=B, E=E, C=C equivalence while connecting the transistors. +1 for the nice answer. Commented Oct 22, 2016 at 11:11
• How about showing us a photo of the circuit? This has to be something rather simple. Open resistor or connection maybe. Commented Oct 22, 2016 at 11:32
• @AdrianMaire Have you checked that the transistor is OK before soldering? Using the diode function of a DMM you can easily detect if the junctions are OK and verify the pinout. Commented Oct 23, 2016 at 19:46
• @AdrianMaire If the transistor is OK before soldering and you fail that test when you probe it after soldering, you've got a soldering failure. Otherwise, the culprit is something else. Commented Oct 23, 2016 at 19:48
• @LorenzoDonati: I have difficulties to check with so small device, so even if tried, I could not test it. On the other hand, I would imply I destroyed now 12 transistors.. which really make those device pretty sensitive. I will try further. Commented Oct 23, 2016 at 19:48

After some more research, it seem the Base require a pull-down:

When SW1 is open, the Base is not connected to anything and get pulled up enough by the transistor to make Collector-Emitter to conduct.

This seem not a problem with the 2N3904 for any reason that I do not know.

EDITED:

It seem I had also an unrelated problem with inappropriate flux.

• Your setup with the SMD part mounted on the breaker board could be more prone to noise pickup. You could try with the 2N3904 connected to the breadboard using long(er) wires and see if it the fault reappears. Commented Oct 27, 2016 at 18:32