Troubleshooting Three Fives, soldering and testing transistors
  • Hi all -

    I just assembled the Three Fives kit, and connected the suggested test circuit.  The test circuit failed (no light), and I'm looking for troubleshooting advice.  With the circuit energized, I found voltage drops across most but not all components.  I also disconnected the circuit, set my meter to diode testing mode, and probed each transistor, first with Com on B and V on each outer lead, then V on B and Com on each outer lead, expecting the meter to read about 0.660 with V on B for the NPNs (2N3904) and with Com on B for the PNPs (2N3906), and 0L in all other configurations.  Unfortunately, only 14 out of the 26 transistors passed this test completely.

    Re: transistor testing, I saw the following (presumably) failing tests:

    * Q5 (PNP), Q8 (PNP), Q18 (NPN), Q19B (PNP), Q23 (PNP) : The BC connection was a glitchy 0 when it should have been 0.660.  I say glitchy because the meter would jump when I'd first touch the probes.  I'm pretty confident I made proper contact with the probes, and retried each one.

    * Q13 (PNP), Q14 (NPN), Q16 (NPN), Q17 (NPN), Q21 (NPN), Q22 (NPN), Q24 (NPN) all showed 0.660 the right way around, but showed high (1.2 or 1.8) with the probes inverted on one or both outer pins.  (I recorded more detail if the distinctions matter.)

    Are these both symptoms of transistor failure?  Would the diode-meter test produce an incorrect result in-circuit under some circumstances with Three Fives?  (I was sure to disconnect the test circuit before doing the test with the meter.)  Are there other testing techniques you recommend for in-circuit transistors that have already been through-hole soldered and leads clipped?

    Re: soldering transistors, I was wary of overheating the transistors, so I tried to be careful.  My Weller iron was set to 520 F, and I methodically avoided spending more than 10 seconds on each lead, and paused 10 seconds between each lead.  I did not use a heat sync clip, and I did all of each transistor's pins in a row; clips and longer pauses (such as doing one pin on a bunch then the next pin on a bunch and so on) would have been sensible additional precautions that I didn't take.  How likely is it that I damaged the transistors during soldering?  Do my test failures above indicate which transistors need to be replaced?  Which precautions against overheating transistors during soldering do you recommend?

    Any other troubleshooting steps you recommend?  I'm eager to treat this as an opportunity to learn more about the 555's design based on the excellent educational supplement that accompanies the kit, assuming I've done what I can to verify the transistors individually.  I have a simple scope (the Rigol DS1052E) if that's worth using in this case.

    Thanks for any help you can provide!

    -- Dan

  • Oh, and a quick question while I'm here: The test circuit included in the instructions calls for a 1uF cap.  Is it OK to substitute an electrolytic here?  I assume it's usually OK as long as the behavior of the circuit doesn't depend on storing charge of different polarities at different times.  I just don't have a 1uF ceramic at the moment.  (I ordered more stock recently.)

    Thanks again!
    -- Dan
  • Hi Dan,
    The first thing that you should do is take a look at the circuit diagram, which is available in the datasheet. 
    ( See our documentation page, here: http://wiki.evilmadscientist.com/555 )

    You will note that some of the transistors (Q5, Q8, Q18, Q19B, Q23) have pins connected together. This likely explains your "glitchy zeroes."

    Others are connected in more complicated ways, and you need to remember that the diode tester tests *everything* connected to the two points that you are measuring. In most cases, there is not just a single diode connecting any two given points.  

    Most errors with the 555 are due to a misplaced or backwards component, or possibly a bad solder joint somewhere. It is also possible to damage a component while installing it, but isn't very common.  (We generally recommend heating a solder joint for about one second, with a hotter iron.)

    What I'd recommend to look further is to open up our "principles of operation" guide, and see if you can work through the circuit to see what is (and is not) working. Can you see a change in the output of the threshold comparator when you change its inputs?  Can you see the change in the output of the trigger comparator when you change its inputs? It sounds like you've got the right equipment and the inclination to work through it, and really see it works inside.  We've gone through this once-- we intentionally let the magic smoke out of one (by putting enough voltage on-- way outside of its limits), and were able to trace the damage all the way through to one of the transistors in the output stage.

    You can use an electrolytic cap, just watch the polarity please.

  • Hi Windell -

    Thanks for the advice!  Good news: I think I just had the test circuit hooked up incorrectly.  :)  I verified the placement and orientation of all of the parts, examined the joints, and went back to my breadboard and started to wonder if I misread the pin assignments.  I hooked it up fresh and it works now.  The diagram looks fine, and I'm not sure where I messed it up, so no further feedback.

    I'm very excited about this kit and the documentation.  High quality, easy build, and very educational.  Thanks again!

    -- Dan

  • Great, glad to hear it!