Need help sourcing current for quad 7-segment LED display

Thread Starter

hspalm

Joined Feb 17, 2010
201
Hello again.
First I need to explain my circuit.

The problem I am showing you is part of a bigger circuit, which will eventually be a dual egg timer. Attached is a circuit showing one quadruple 7 segment display, which eventually will be doubled.

My microcontroller is pulsing the clock input on a 4017 decade counter, each output from 0-3 is connected to common anode of each of the 4 digits on the display. Since all digits share catodes a-g + punctuation mark the circuit needs to be multiplex. Each time the 4017 is clocked, the microcontroller outputs a different BCD code fed through a SN74LS47 BCD to 7 segment decoder, then further to a-g inputs on quadruple 7 segment display. I have this circuit right, see video, problem comes next.
http://www.youtube.com/watch?v=Z0-M1H9PTqs

These logic devices provides next to no current to my LEDs, as shown by the dim/non-bright digits on the video. I need to amplify this, but do I need both PNP on LS47 side AND NPN on 4017 side? As the LS47 can hardly sink anything, and the 4017 can hardly source anything.

As always, I am thankful for any advise here :)

EDIT 1: Also, one more think I want to know. Since I will double the amount of displays, I will face the problem of one segment being lit either none, one or two at the same time. In the case of one of the displays lighting say "g" segment, I must limit the current the half of what is needed for when same segment "g" is lit on both displays. Is there a workaround for this? Do I just add two separate but identical current limiting resistors in parallel from transistor on whatever side it is placed?
 

Attachments

Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
The 74LS47 can sink up to 24mA per open-collector output.
You need to use current limiting resistors between the 74LS47 and the 7-segment display, one per "letter"; eg "a", 'b', etc.

The problem is that the 74HCT4017 can't source much current.

You'll need to use a pair of NPN and PNP transistors per output (with appropriate resistors) to make it source more current.

[eta]
Like this:

 

Attachments

Last edited:

Thread Starter

hspalm

Joined Feb 17, 2010
201
Thank you!

How is this helping the 4017 source more current? I don't see an extra 5v supply anywhere, so from all I can see the current is still drawn from 4017 output. Maybe you have a name for this setup, so I can google it or something? Only similarity to the transistor write-ups in the AAC wiki/book might be the cascode configuration.

Thank you for your help!

Edit: One more thing, what about an inverter and a PNP switch?
 
Last edited:

Thread Starter

hspalm

Joined Feb 17, 2010
201
I am sorry to bring this up again, but it's just the idea of finishing a circuit without really knowing what I've done... I just want to know how this external circuit can make the 4017 source more current by "magic" (as it seems to me).
 

SgtWookie

Joined Jul 17, 2007
22,230
Gosh, I'm sorry - I have a lot of stuff going on, and I forgot to add the +5v supply at the R4 and Q2 emitter node.

Do you understand what I just wrote? I mean for you to connect +5v to the horizontal wire above R4 and Q2.
 

Thread Starter

hspalm

Joined Feb 17, 2010
201
No problem SgtWookie, just greatfull that you're willing to help me on this.

Yes, now I understand the circuit. I see that the first NPN is only to turn on the PNP which is capable of sourcing current as a switch, and the NPN can only sink current?
However, if I could use an 8-input inverter to substitute for the NPN, that would suite me good. And I guess if that would not work, I could use a transistor array like the ULN2003 or similar for the NPNs? Just for convenience, since it's an ic package.
 

SgtWookie

Joined Jul 17, 2007
22,230
Yes, you can replace R1, R2, and Q1 with a ULN driver. I suggest the ULN2004 or ULN2804.
If you can't find those easily, the 2003 or 2803 will work.
 

Thread Starter

hspalm

Joined Feb 17, 2010
201
There is something which is really bugging me. As I was browsing ebay for a cheap lot of ULN2803's I saw this chip, MAX7219CNG.

This chip replaces
1 x HCT4017
1 x 74LS47
16 x transistors
46 x Resistors

But I felt like I was on a roll here! It was fun getting the multiplexing/timing circuit working, now it was only a matter of amplifying current to the LEDs and adding another display. On second thought, it was no fun at all making a PCB with those components. I think I'll just act like an engineer and go for the fastest and easiest solution. It's actually the cheapest one also. Besides, I've had my fun already.

What would you've done?
 
Top