Only works half the time, transistor switching on by itself

Discussion in 'The Projects Forum' started by thieaux, Jan 20, 2014.

  1. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    Hey guys, im using a ic to control a latching relay, it works most of the time then all of sudden it would stop... Its not a code issue, since im bench testing it right now.

    Parts List:
    1. ne720-1a-z latching relay,http://www.relaymall.co.kr/pdf/NE720.pdf.

    2. Couple of darlington npn. http://www.nteinc.com/specs/200to299/pdf/nte261.pdf

    Basically the circuit is: each darlington drives the set and reset pin of the relay, and the common is sourcing 12v through a 10ohm 5w resistance since the case of the relay says "7v" and the measure resistence between the pins was 12 ohm, that gives me 416mA rounded it up to 600ma. That gives me ic of 600ma, the hfe of the nte261 is 1000, Ib=0.6/1000=>0.6mA, since im driving a darlington voltage drop is 1.4 from a 5v ic output is 5-1.4=3.6v, base resistor used is a 3.9kohm...3.6/3900 gives me 0.9mA to saturate the transistor. Im asssuming my math is correct, everything works as expected.

    Here is the problem all of a sudden the circuit will stop working and one of the npn will switch on without any voltage to the base, i even disconnected the base pin to make sure and the transistor will stay on, at first i thought a bad transistor and replaced the transistor array with new ones and the same thing happened. Than it will work again for no particular reason, i checked the B-E voltage and was getting 1.4v without nothing connected to the base.

    Heres a pic of the circuit (i know the npn are not correctly drawn, this was a sketch up).

    [​IMG]
     
  2. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,804
    1,105
    If you have the base floating, leakage current or even interference could turn the transistor on. Connect a pull-down resistor (say 10k-33k) between base and emitter.
     
    thieaux likes this.
  3. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    Thanks, but the base wasnt floating at the time its was tied to ic, i disconnected it from the ic once the condition presented itself assuming it was a code issue, pin was low and yet the transistor was conducting. Now that you mentioned it, i will ad one, if you dont mind, what would be the thevernin equivalent? ......Never quite figure out how to calculate a pull down base resistor.
     
  4. wayneh

    Expert

    Sep 9, 2010
    12,151
    3,058
    Try your base resistor calcs again assuming a gain of 100X instead of 1000X.

    A rule-of-thumb for saturating a transistor is base current = 10% of collector current, not 1%. High gain is in linear mode, not in saturation.
     
    thieaux likes this.
  5. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    Ill give it a shot, i tried it with 2n2222 last night same thing happened..Funny thing is that in the breadboard is fine, but on soldered to the pcb its when problems start, if it was a layout issue it would show up right away instead its totally random, it could go for hours and be fine and all of a sudden switch and stay on causing the resistor and npn to get hot
     
  6. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,394
    1,606
    As long as the Arduino outputs driving the base resistors are either hi or low (meaning not changing to inputs) then you don't need "base" resistors: the inline single base resistor acts as both.

    I do NOT see suppressor diodes across the coils. That can breakdown transistor, usually permanently shorted, but intermittent isn't unknown.

    Add diodes and replace the transistors.

    (Note I have no comment about how you drive the latching relay, I never used them.)
     
    thieaux likes this.
  7. tindel

    Active Member

    Sep 16, 2012
    568
    193
    Also, one of the downsides of a darlington is that the Vce,sat voltage is kind of high... in this case it's spec'd at 2.0V at Ic=3A, Ib=12mA... I agree... I'd drive it with a gain of 100 and maybe lower the source resistance with the Vce,sat in mind as well.
     
    thieaux likes this.
  8. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    According to the arduino forum and the atmega specsheet there is no internal pull down, wouldnt it be advisable to use it on the transistor?
     
  9. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    So far so good with the gain at 100
     
  10. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,804
    1,105
    Don't know. I just use something of the same order as the base drive resistor. No calculation.
     
    thieaux likes this.
  11. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    Nope, didnt work, left it alone for a couple of hours, and smell something came back and the transistors were switched on
     
  12. BobTPH

    Active Member

    Jun 5, 2013
    785
    114
    Internal pullups are for inputs. The output is a push-pull driver.

    Bob
     
  13. BobTPH

    Active Member

    Jun 5, 2013
    785
    114
    Did you put in the diodes that were suggested? A back EMF pulse could cause the Arduino to latch up.

    Bob
     
    thieaux likes this.
  14. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    Yep, used a couple of 4007
     
  15. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,394
    1,606
    No. If the Arduino output is low is has an active device to make that low, and a saturated transistor is a crazy good pull down element.


    1N4007 is as good a choice as any. Overkill for production BOM, but fine when in your parts bin.
     
  16. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0
    UPDATE, so i unsoldered the transistor and i soldered a 12v led between source and the pin were the collector used to be and it lid up, that was mind boggling, how can that be? the pin isnt connected to anything, the first time the problem occured i had a thought that maybe something was shorting but could find anything wrong, so i gave that idea up, now that the led lit up with nothing connected to it again the idea came up so i turn over the pcb and the only thing i see is flux residue, can that be the problem? I took a toothbrush to it and on the first brush the led turned off.....WOW that cant be, flux isnt conductive....i dont know if it is or not, i just know that afteri brushed it the problem went away
     
  17. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,804
    1,105
    If the flux is acidic it could conduct. But you shouldn't use that type anyway! Resin-based flux shouldn't conduct significantly. Perhaps your brushing action removed a solder splash which was shorting something?
     
    thieaux likes this.
  18. thieaux

    Thread Starter New Member

    Jan 31, 2013
    13
    0

    Thats what i thought, maybe a bridge was created while soldering, thats probably the reason it workded on the breadboard only.

    Anyway i used some acetone to remove the residue with a toothbrush eo far so good.

    Thanks for your time and patience with newbies....

    P.s. any recomendations on a flux and sodder?
     
  19. RichardO

    Well-Known Member

    May 4, 2013
    1,238
    385
    You should not use acetone to clean printed circuit boards. It is very flammable and too strong a solvent which can harm both you and the circuit.

    It is better to use isopropyl alcohol. You can get it as rubbing alcohol at your local drug store. Get the highest concentration they have -- typically 70 percent.
     
  20. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,804
    1,105
    Use resin-cored solder (e.g. Ersin. Don't know what's available wherever you are).
     
Loading...