LM3914 Cct Just wont work

Discussion in 'The Projects Forum' started by transducer, Jul 28, 2007.

  1. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    Hello to all from a newbie member.

    I have a circuit dilema that I am hoping a thousand minds are better than one.

    I have put together an LED driver circuit to show voltage from 0 - 4V using the LM3914 (Please see attached for schematic)
    I have made a prototype on breadboard and the thing just does not work corrrectly. I get all the LED's lit up all the time, without ANY I/P on pin 5

    Just to make sure I have costructed this correctly can you peeps please confirm my pin outs & cct please

    Pin 1 to LED #1
    Pin 2 to 0V (V-)
    Pin 3 to 8V supply from LM317T (V+)
    Pin 4 to 0V (Reference Low)
    Pin 5 I/P (Signal)
    Pin 6 10K pot divider to give 4V FSD (Reference Hi)
    Pin 7 1K2 to 0V for If of LED's (Reference Out)
    Pin 8 to 0V (Reference Adjust)
    Pin 9 to 8V supply for 'Bar' mode (Mode)

    I have redone this circuit at least ten times and it is the same every time.
    Please can someone help me as it is driving me mental

    Thanks everyone

    Transducer
     
  2. JoeJester

    AAC Fanatic!

    Apr 26, 2005
    3,373
    1,158
    With Signal at ground, what are you reading on pin 1 and the other pins connected to the LEDs?

    What is Vf for your LED?

    What does V+ measure ? ... I know you said 8 volts.

    What does R Hi measure?

    Are you sure the LM3914 isn't defective?
     
  3. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    Thanks for the help JoeJester

    With Signal at ground, what are you reading on pin 1 and the other pins connected to the LEDs?

    with pin 5 connected to 0v I have

    Pin 1 9.72
    Pin 18 9.35
    Pin 17 9.38
    Pin 16 9.46
    Pin 15 9.43
    Pin 14 9.41
    Pin 13 9.50
    Pin 12 9.53
    Pin 11 7.06
    Pin 10 9.71

    (NO LED's lit)
    Why is pin 11 low compared to the rest & how am I getting 9 ish volts when I am only supply 8 for the circuit, very confused :confused:

    With Pin 5 open I have

    Pin 1 6.34
    Pin 18 6.35
    Pin 17 6.21
    Pin 16 6.21
    Pin 15 6.17
    Pin 14 6.16
    Pin 13 6.20
    Pin 12 6.20
    Pin 11 6.34
    Pin 10 6.33
    (All LED's lit)


    What is Vf for your LED?

    1.6V

    What does V+ measure ? ... I know you said 8 volts.

    8.02V

    What does R Hi measure?

    8.02V

    Are you sure the LM3914 isn't defective?

    I have tried three different ones, along with different LED arrays and they are all the same.

    Could I possibly have a defective breadboard. It is the first time I have used it:eek:

    Thanks again for your help
     
  4. GonzoEngineer

    New Member

    Jul 8, 2007
    46
    0
    Ahhh.....breadboard huh? And you never used it before.......

    I quit using those things years ago......point to point wiring is better.

    If it is a used one, someone may have forced a large wire into a contact and spread the contacts out. That could account for funny voltage measurements because things will float around without good grounds.

    Also, some solderless breadboards with common power busses on the top and bottom have a break in the center seperating the busses on each side of the breadboard. That one burned me many years ago.

    Looks to me like the most likely culprit is the breadboard.
     
  5. JoeJester

    AAC Fanatic!

    Apr 26, 2005
    3,373
    1,158
    With pin 5 OPEN, you probably have some voltage there [making the breadboard part of the problem] causing all the LEDs being lit.

    It certainly appears from the zero test, that your circuit appears normally. Move it to a different portion of the breadboard ... and if it works, you'll know the area near pin 5 on the previous layout was bad .... I would mark it with a pencil so when it is clearly identified as defective, you can report it to whoever you report such things to.

    You could have taken readings on all pins [I know I didn't ask for them all ... but you would have seen the error on pin 5 from the open test] ... and yes, everyone first assumes they made the mistake when breadboarding a circuit ... you did part one of troubleshooting, the visual check. Follow that with voltage checks and waveform checks.

    I'm sure this is one experience you won't forget. Happy future breadboarding.
     
  6. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    I have moved it all over the breadboard with the results the same. The board is brand new

    I have checked the voltage on pin 5 and it is reading 5.52V
    Where is this voltage comming from? because obviously it will light the whole bar that is set up 4Vfsd with 5.52 on the rail

    Are there any other checks I can carryout although to be honest it will probably be quicker rebuilding point to point
     
  7. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    :mad:Arrrrrgggghhhhh:mad:

    Just soldered out my circuit point to point to prove the thery of the breadboard being dodgy and.............ITS STILL THE SAME
    All the LED's light up even with no voltage on PIN 5.

    So it must be something I am doing, or something in the circuit that is not correct.

    Can you generous people please prevent me from smashing my very nice office to pieces and check my circuit diagram. There must be something I am doing that is just 'incorrect' either taking something to the wrong rail or connecting points up that shouldnt be.

    Pleeeaaasseee can someone save me
    Thanks
     
  8. thingmaker3

    Retired Moderator

    May 16, 2005
    5,072
    6
    The voltage is coming from the buffer. Have a look at the datasheet: http://www.jaycar.com.au/images_uploaded/LM3914.PDF

    With pin 5 open, both buffer i/p's will be the same. (Op amps always do that.) I'll wager a box of donuts that an open pin 5 will always measure roughly 1.5v above pin 6. (I know you said pin six was 8.02v, but you also said it was 4v. All those pins and levels get confusing after a bit, don't they?)

    Try putting an actual signal on pin 5 instead of leaving it open. That will make the buffer do something for you instead of idling at just above your reference level.
     
  9. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    Pin 5 = 1.6V
    Pin 6 = 4.0V
    I like custard filled donuts :)
    The LED's dim slightly when I measure on pin 5 :confused:


    Sorry if I got it wrong but what I thought I'd said was that I had a 10K pot set up to put 4V, from an 8V supply onto pin 6 as this is what I want the max voltage measurement to be, ie 4V = FSD = all LED's lit

    Just tried, but it is still the same with all LED's lit

    Can I please ask you have a look at my circuit to ensure I am not making some simple error. Many thanks for your help on this, greatly appreciated
     
  10. JoeJester

    AAC Fanatic!

    Apr 26, 2005
    3,373
    1,158
    What have been the results when you inputted zero volts, 1 volt, 2 volts, 3 volts, and 4 volts into the signal input? Is it behaving normally?

    Your not in some high energy enviroment are you? What does the scope show when you connect between pin 5 and ground? At both the x10 and x1 settings.

    You said the LEDs dimmed when you loaded the pin with the DMM. You might end up putting a resistor between pin 5 and ground to solve this.

    I didn't see anything wrong with your circuit the other night.
     
  11. JoeJester

    AAC Fanatic!

    Apr 26, 2005
    3,373
    1,158
    Transducer ...

    Take a look at these two ...
     
  12. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    Hi Joe,
    The results are the same regardless of the voltage I apply to pin 5. The LED's are on all the time as soon as I power up the circuit. The only time the LED's go out is if I connect pin 5 to ground

    I am trying to solve this in my office at home, hence no wierd EMF or the comfort of a scope to use :)

    When you say drop a resistor between 5 and ground, do you mean to solve the DMM dimming the array, or to solve this overall problem?

    Many thanks for the pdf's you attached. I will compare voltages between circuits to see if there are any differences between yours and mine.

    I am beginning to think this is a simple error on my part because:

    I have changed both the driver and array with no diference
    I have moved the circuit around the breadboard (numerous times), again with no difference
    I soldered the circuit together point to point with....you guessed it, no difference.
    I have put this circuit together soooooo many times, surely if I had placed something incorrectly I wouldn't have done it numerous times ??

    I am that desperate I even sat down with my 7 year old daughter and explained my problem to her to see if she could shed some innocent light into my situation. She failed so I sent her to bed

    Thanks
    Transducer
     
  13. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    I wonder if it is possible that the use of a 10K potentiometer to supply the top of the reference resistor network R(HI) is causing problems.

    I suspect that the use of a low impedance voltage source at R(HI) will be beneficial. The device contains an internal voltage regulator that can be used to generate a voltage reference.

    You might try disconnecting pins 6, 7, and 8 from their present connections and connect pin 6 and 7 together then connect a 1K resistor from the junction of pin 6 and 7 to pin 8. Then connect a 2.2K resistor from pin 8 to ground. This will set the voltage at R(HI) to 4 volts.

    If you need the voltage at R(HI) to be adjustable that can be arranged.

    hgmjr
     
  14. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    Hi hgmjr,
    Many thanks for the pointer.
    I just tried setting the circuit up as you described but the result is the same. All LED's are lit with no I/P on pin 5
    The only difference before is the LED's seem to be a bit dimmer, due to the added resistance on pin 7?

    Here are the voltage levels on the pins with 8V supply and pin 5 open

    pin 1 5.66
    pin 2 0
    pin 3 7.70
    pin 4 0
    pin 5 0.9
    pin 6 4.01
    pin 7 4.01
    pin 8 2.14
    pin 9 7.67
    pin 10-18 5.64


    It just dont make sense to me:confused:

    Thanks

    Transducer
     
  15. techroomt

    Senior Member

    May 19, 2004
    198
    1
    first pass -
    is it possible that "pin 5 open" is not an operational option here, as obviously the comaparators are going to compare your input signal on pin 5 to ref hi or 4vdc.

    pin 7 must be properly loaded, don't overlook it in your checkout. is pin 7 tied to pin 6 in your circuit?
     
  16. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    isn't pin 5 open the same as 0v to the comparators? ie when the circuit is in operation but there is no voltage input on pin 5

    In my original cct pin 7 was connected to 0v via 1K2
    I have also tied it to pin 6 and this makes no difference
    Am i correct in thinking pin 7 is for LED current, to control brightness
     
  17. techroomt

    Senior Member

    May 19, 2004
    198
    1
    no, an open input is not the same as 0 volts. your connection for 0 volts would be when it is tied to the circuit ground. you should only be applying an input voltage between ref low (gnd) and ref hi (4v).

    pin 7 is for led current, did you tie pin 7 to pin 6 (wiper of 10k pot), or did you tie pin 6 to pin 7 (1.2k rsistor)?
     
  18. techroomt

    Senior Member

    May 19, 2004
    198
    1
    also, have you tried dot mode? tie pin 9 to (V+ - 100mv), or data says you can leave it open (when specified) for dot mode.
     
  19. transducer

    Thread Starter Member

    Jul 28, 2007
    25
    0
    techroomt,
    My ref LO is 0v so with nothing on pin 5, effectively this is 0v no?
    I have tried pin 7 tied to pin 6 as well as to pin 7

    Nothing I do seems to make any difference
     
  20. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    OK transducer,

    Can you take a digital picture of your breadboard and post it here. The more detail the picture the better. If you have camera with a macro setting then that would be best.

    A closeup glamour-shot would be perfect.

    hgmjr
     
Loading...