Testing of a M5450B7 LED Driver

Thread Starter

k5 blazer

Joined Dec 1, 2019
26
I am trying to figure out the issue with my display circuit board. After pulling up the cut sheet for the M5450B7 I am looking for some clarification.

Can I just jump 5v to pin 20 and should that illuminate the display? What pin or how does the driver get reference to ground? Is there a way to put power on the chip and then manual trigger one of the outputs?

Really I am just trying to figure out how to trouble shoot this circuit board. I have a high-level of understanding on how it should work but a novice level of trouble shooting circuit boards.
 

Attachments

Ian0

Joined Aug 7, 2020
10,089
Can I just jump 5v to pin 20 and should that illuminate the display?
pin 20 should already be 5V. That's the positive supply to the IC
What pin or how does the driver get reference to ground?
Vss pin 1
Is there a way to put power on the chip and then manual trigger one of the outputs?
No, it's a serial datastream.
If there is a clock signal you should be able to illuminate all segments by forcing the DATA input high.
 

Thread Starter

k5 blazer

Joined Dec 1, 2019
26
So basically if I check and the chip has the 5v at pin 20 and Pin 1 has a connection to ground with out and clock signal this chip is just sitting there waiting for data change to then fire any of the other output pins if you will?
 

MisterBill2

Joined Jan 23, 2018
19,103
The data sheet implies, as I read it, that all of the control bits need to be sent before the data is displayed. And although it is not stated, it seems thst the data should be four bits of BCD for each character. And somehow the signals for pins 17 and 18 tell which set of four digits are being sent. So evidently there is a lot more information than is provided in the short data sheet.
 

Ian0

Joined Aug 7, 2020
10,089
The data sheet implies, as I read it, that all of the control bits need to be sent before the data is displayed. And although it is not stated, it seems thst the data should be four bits of BCD for each character. And somehow the signals for pins 17 and 18 tell which set of four digits are being sent. So evidently there is a lot more information than is provided in the short data sheet.
It's not like that at all. It's one bit per output. There's no decoding. It's very easy to use. Send a 1 followed by all the other output bits. When it gets the 36th bit, it latches them all into the output drivers, then waits for another 1.

It's quite easy to make a massive bargraph driver out of it: all you need is a pulse width that relates to the parameter you want to display, and a squarewave that gets 36 clocks in the pulsewidth.
 

MisterBill2

Joined Jan 23, 2018
19,103
It's not like that at all. It's one bit per output. There's no decoding. It's very easy to use. Send a 1 followed by all the other output bits. When it gets the 36th bit, it latches them all into the output drivers, then waits for another 1.

It's quite easy to make a massive bargraph driver out of it: all you need is a pulse width that relates to the parameter you want to display, and a squarewave that gets 36 clocks in the pulsewidth.
GOOD LUCK with that concept. What I got is that it takes all of the bits sent and then one more and the drivers latch on or off.
Consider that it is shown feeding an 8 digit display. Seven segments per digit. So there would be a bit of downstream decoding maybe?
 

Ian0

Joined Aug 7, 2020
10,089
Yes, the 8 digits takes some effort!
You have to decode the data from BCD to 7-segment, then you have to add the two multiplexing outputs to the datastream, and alternately send each pair of digits.
(I actually built the bargraph - long ago before there were microcontrollers, but the schematic is long lost.)
 
Top