Driveing LED displays without constiant output from MCU

[Ferroto]

This diagram uses 74HCT373N's to drive 2 7 segment LED displays.


74HCT373N: (Mouser, Datasheet)
   When LE is HIGH values are read from D0-7 when OE is low values are transmitted to Q0-7. Thus since LE and OE are connected if Dig1 is HIGH it receives it's value from D0-7, if it's low then it transmits it's values continuously to the 7 segment LED display it's driving this frees up the 8-bit data bus to talk to additional digits.


LDQ-M282RI: (Digikey, Datasheet)
   7 Segment LED display with common Cathode.



Note one thing I've noticed is the more digits there are the more control pins (Ex: Dig1, Dig2 etc...) there are I'm working on a way around this so you'll only use 1 control pin for all digits.


I was just wondering weather or not this would work I haven't dealt with discreet logic such as flipflops much.

[jimmc]

I think you're a bit confused here, The /OE line is purely an output enable bar.
When it is high the outputs are tri-state but the state of the latches is not affected.
The LE line controls the latches, when it is high the latch is transparent (the output Q follows the input D) when it drops low output D holds the state of the input Q (when LE was last High).
If you put an inverter in one LE line and the other, one '373 will be latched when the other is transparent.
(If you want to avoid the transparent state use the edge triggered 74HCT374 this device only samples the D input on the rising edge of the clock.)

You can't really use a single resistor in the common cathode line; if a '1' is displayed then 2 segments are on, for an '8' all seven are on. So the brightness changes depending on what is displayed.

Jim

[Ferroto]

I think you're a bit confused here, The /OE line is purely an output enable bar.
When it is high the outputs are tri-state but the state of the latches is not affected.
The LE line controls the latches, when it is high the latch is transparent (the output Q follows the input D) when it drops low output D holds the state of the input Q (when LE was last High).
If you put an inverter in one LE line and the other, one '373 will be latched when the other is transparent.
(If you want to avoid the transparent state use the edge triggered 74HCT374 this device only samples the D input on the rising edge of the clock.)

You can't really use a single resistor in the common cathode line; if a '1' is displayed then 2 segments are on, for an '8' all seven are on. So the brightness changes depending on what is displayed.

Jim

I see what your saying, I was under the impression that it couldn't input and output at the same time. So basicly i can just pull OE to GND and have LE high when transmitting data to the '373. Also i noticed my first project the counter had the bright\dark problem too. also the inverter can't be used because that would have atleast 1 '373 with a high LE at all times and i need to use D0-7 for other things. Basically J1 is a header that connects this board to another. That could contain anything including a shared bus.

[jimmc]

That's it.

Since you only want to control one latch at a time you can use a '1 of n' decoder such as the 74HCT238 to cut down the number of control lines once you get past two digits.

Jim

[KuchateK]

I've recently stumbled on this project. It has one interesting thing:
http://www.elektronika.ba/714/darkroom-timer-for-pcb-exposure/
Schematic direct link:
http://www.imagesforme.com/upload/6713c44a.gif?id=6d4f88f2a9e1044d343436383637
This guy is using 4094 shift registers to controll four 7 segment LED displays (32 LEDs) with only two pins of MCU.

[GeekGirl]

This guy is using 4094 shift registers to controll four 7 segment LED displays (32 LEDs) with only two pins of MCU.

I have done this before with 74HC164, these are very fast, so I can clock at full speed from say an AVR

Regards,

Kat.

[qno]

I would tie the /OE to Vss (GND) or else the LED's would only light up when you are writing to the 373.

Also take a look at the 4511.