My Thermostat stopped working a couple of days ago and I tried to debug it. The first thing I check is if there is any component that gives a strange result with the multimeter and I found these three components which look a lot like transistors (maybe BJT?): Suspects

I'm sorry for the bad quality but I couldn't get any better picture with my phone. Anyway, the package is like this one (only the package is the same, the component is different from those on my board!:


Two are the same component (5Bp or 58p is what is written on the package), and I can measure 0.6V (with the diode option of a multimeter) across any pins. This is strange for a BJT, but it may be related to how they are interconnected.

And the third one, pointed by the green arrow, with a horizontal "P" and a vertical "D3" written on the package, is what I would assume is a PNP BJT. Here is a sketch of the component (really bad proportion!):

Sketch of suspected component

My assumption is also that this is broken: I can't measure any diode for this component. Two pins are not connected (multimeter shows an open) and the other two couple of pins are connected but the multimeter measures a short with the diode option, and the resistance measures approximately 27 Ohm and 19 Ohm, regardless the polarity of the multimeter.

With measure

So, my question is if anyone knows by heart what this component is exactly, or if he can confirm that this is at least a pnp.

Unfortunately, I don't have any way to understand what type of bjt (assuming it is a bjt) without knowing the pin out and measuring the orientation of the diodes...

I may infer this from how the components are interconnected and especially if any terminal is connected to Vcc/ground, but without knowing the pin out of the component it's still quite bothersome, I think!

PS: I said I'm debugging my old thermostat, so I'll briefly describe the issue I had, maybe someone can help with the debug in general and not only recognizing the bad component! There may be some inaccuracies/errors in my explanation though, as it's not my business to work with this kind of component, but this is what I could understand from my trials. Please feel free to ask/correct me.

The relay that should give the start signal to the heater works (or the relay that should switch when the thermostat tells the heater to start). I can tell this because when I set the temperature so that the heater should start, I heard the typical noise of a relay that switches. This also means that the temperature sensor works, because it correctly switches the relay when I set a higher/lower temperature on the thermostat. Regardless of the relay that switches, the signal to the heater doesn't arrive, so the heater is stuck off.

There is also a quite big IC (NEC), which I suppose is a microcontroller, and I suppose it works correctly because: it drives correctly the relay, the LCD on the front panel works correctly, etc. There is also a quarz oscillator, which should work because the microcontroller works (I assume it's the clock generator for the microcontroller).

All the other components (15 components approximately in addition to these I already mentioned) are just passive components (capacitors and resistors), mainly SMD.



JRE pointed out that it could be the realy, and it really was. Now I'm trying to find a good replacement and asked a new question there Suitable Replacement for a Dual Coil Latching Relay. Feel free to help/see the end of this story! :)


2 Answers 2


If the relay clicks but doesn't deliver power to the heater, then the contacts of the relay are either burned out or there's a broken wire between the heater and the relay - or maybe the heater is broken.

You can't really expect to measure anything reasonable when the parts are still on the board. The measurements of those SOT-3 parts are really questionable. They connect to so many other things that there's no way for you to tell what you are really measuring.

Check the relay. A relay that clicks but doesn't switch is probably broken. If it is OK then check the wiring to the heater.

  • \$\begingroup\$ D<Censored> you just beat me to it :-) \$\endgroup\$
    – Oldfart
    Apr 3, 2020 at 20:03
  • 1
    \$\begingroup\$ @Oldfart why did you have to censor "ude"? \$\endgroup\$ Apr 3, 2020 at 20:31
  • \$\begingroup\$ I'll check the relay! Anyway, I still find it strange to measure low resistances on both terminals of a transistor, especially on this kind of low power board, where I would only expect high values. And if it was only across one couple of terminals it might have been ok, but on two couple of terminals is really strange to me. \$\endgroup\$
    – Tripola
    Apr 4, 2020 at 8:47
  • \$\begingroup\$ 1. Measurements on board are usually wrong. 2. They may be dual diodes rather than transistors. \$\endgroup\$
    – JRE
    Apr 4, 2020 at 8:51
  • \$\begingroup\$ It really was the relay! It's a dual coil latching relay. The relay clicks on both ways but connects only two of the three output poles. I also asked another question to find a replacement as I'm finding it hard to find the same component, maybe you can help again there! electronics.stackexchange.com/questions/490920/… \$\endgroup\$
    – Tripola
    Apr 4, 2020 at 10:11

I figured I would answer this question as posted, even though others have already answered a "deeper" question.

The device marked "5Bp" (yes it is an upper-case letter B) is a PNP transistor, type BC807-25 (or BC807-25W). Datasheet here.

Your other device marked "P" is probably (not 100% sure, best guess) a dual diode, type DAP202U, datasheet here.

The usual way to find these is to google the code, together with any/all of the magic keywords "smd" and/or "case" and/or "mark" or "marking" and/or "code". So, "smd code 5Bp" was my 1st query w.r.t. your 1st device.

It's a bit of an "art" as these parts often have extraneous markings (date codes, production indicators, etc.) and different manufacturers sometimes use different codes for the same type# (i.e. the DAP202U above is marked "P9" by ONSEMI but "P" by Rohm). SMD markings are not standardized...sigh.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.