Shop OBEX P1 Docs P2 Docs Learn Events
ULN 2803 and Shift Register — Parallax Forums

ULN 2803 and Shift Register

jcfergusonjcferguson Posts: 86
edited 2007-03-22 07:14 in BASIC Stamp
Hi All,

Could someone take a look at this schematic that I am using to drive some leds?

I am not sure what voltage should go to the 2803 - should the uln2803 have 5v to pin 10, the same 5v that goes to the shift register and then all grounds tied together? Or should it get the 7.5 volts that are driving the leds with the grounds tied together?

It seems to work the way it have it, but if I take off the 5v to the 2803 - the leds still light but don't go off all the way...


I just wonder if someone could check my work as I am unsure about the power connections, despite having looked at the data sheet...


Thank You!

Carlos Ferguson
shift-register.gif
888 x 648 - 10K

Comments

  • T ChapT Chap Posts: 4,223
    edited 2007-03-22 06:19
    Try pulling out the 470uF and see what hapens. A ULN doesn't need 5 volts into it for an LED, lose that too. Make sure the grounds are connected on all the supplies. The outputs of the 595 may need pullups, try putting a 4.7k or 10k to 5V as a pullup on the ULN inputs, that should do it.

    Post Edited (originator) : 3/22/2007 6:29:53 AM GMT
  • jcfergusonjcferguson Posts: 86
    edited 2007-03-22 06:37
    The 470uF is to fade the leds - it works ok I think,


    Reading your post I get the sense that the 5v to pin 10 might be a problem? I don't need anything on this pin? - or should I instead tie it to the 7.5 that powers the leds? All grounds should be tied together as a general rule?

    The ULN's are actually driving 16 leds, I just didn't want to draw them all in...


    Anyway, it SEEMS to work right as drawn in that circuit, but I am just unsure of the pin 10 connection...

    Thanks,

    Carlos
  • T ChapT Chap Posts: 4,223
    edited 2007-03-22 06:42
    It is not a problem, it just isn't needed. The ULN2803 doesn't need a "+V", that "input" is there in cases of sinking a relay or other inductive device, not required for an LED. You said the LEDs don't go all the way off, that sounds like a slight issue, maybe the ULN needs more juice on the input to turn it on, thats why I suggested a 10k Pullup to hit it harder. Verify that the 595 is an open collector output, if so, use pullups.

    Post Edited (originator) : 3/22/2007 6:54:43 AM GMT
  • jcfergusonjcferguson Posts: 86
    edited 2007-03-22 06:58
    Thanks again for the help.


    I will let pin 10 float then...

    the uln2803 sinks current to the output pins when it is brought high on the input pin ( I think ). So wouldn't I want a pull-down resistor on the inputs to make sure it goes off all the way? I'm not sure this is necessary but I just want to understand why.

    Carlos
  • T ChapT Chap Posts: 4,223
    edited 2007-03-22 07:03
    Yes, you are correct, I am doing too many thing at once. The 595 is tristate output. This means in certain cases it can FLOAT the output.


    I am not very familiar with the 595, but you may need to check what state it is in when you want the LEDs off, it may be that the IC is floating when you want the IC off, which may account for the ULN not fully turning off the LED. Just for a test, look at an LED that is staying partially on, pull down that channel ULN input with a jumper to GND, see if it kills the LED. If it does, try a 10k, or 4.7k as pulldown. I think you should test it without the 470 to see if it goes out completely, as it doesn't really add up yet what is causing the LED to stay on..

    Post Edited (originator) : 3/22/2007 8:33:34 PM GMT
  • jcfergusonjcferguson Posts: 86
    edited 2007-03-22 07:14
    Ok, I'll try these things out tomorrow. Thanks for the help.

    Carlos
Sign In or Register to comment.