SHIFTOUT and BS2 on HomeWork board with MM5480N display driver?

pixlo

Hi there,

I've just hooked up a national semiconductor 3.5 digit display driver to my breadboard and plugged in the power and ground with several LEDs to see if I can get the first 11 pins or so to alternate on,off,on,off.

The MM5480N protocol is similar to the MM5481, 35 bits with a 1 at the beginning. I thought the datasheet said it latches when it encounters the 36th 1. In any case, my data out is pin 2 and my clock out is pin 3.

DO
SHIFTOUT 2, 3, MSBFIRST, [noparse][[/noparse]%1101010101010000\16, %0000000000000000\16, %0001\4]
LOOP

first off, can my pic on this homework board that i have do this?

second, is my snippet of code right?

third, has anyone worked with these M548x display drivers before using SHIFTOUT? Using something else?

Any suggestions? I'm using the display driver to control a set of electroluminescent wires arranged into 5 large (4'x2') 16 segment displays. I just got my optoisolators and triacs to switch and am now working on the data half.

Thanks in advance,

-nick

Chris Savage

Do you have the brightness control set up properly?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com

pixlo

I have a wire coming off Vdd that measures around 4V with current at .68mA

I do not have a cap hooked up to the brightness control.

The datasheet suggests you place a .0001uF cap around the controll to keep it from oscillating.

the datasheet said...
The MM5480 is specifically designed to operate 3(/2-digit
alphanumeric displays with minimal interface with the display
and the data source. Serial data transfer from the data
source to the display driver is accomplished with 2 signals,
serial data and clock. Using a format of a leading ``1'' followed
by the 35 data bits allows data transfer without an
additional load signal. The 35 data bits are latched after the
36th bit is complete, thus providing non-multiplexed, direct
drive to the display. Outputs change only if the serial data
bits differ from the previous time. Display brightness is determined
by control of the output current for LED displays. A
0.001 mF ceramic or mica disc capacitor should be connected
to brightness control, pin 13, to prevent possible oscillations.

I wonder if that means the 36th bit is the 1st bit of the next set of 35 bits?
My SHIFTOUT has 36 bits, not 35.

I'm at work and dont have the board in front of me...

more info from the datasheet (found at eshop.engineering.uiowa.edu/NI/pdfs/00/61/DS006138.pdf)

the datasheet said...
There must be a complete set of 36 clocks or the shift registers
will not clear.
When the chip first powers ON an internal power ON reset
signal is generated which resets all registers and all latches.
The START bit and the first clock return the chip to its normal
operation.
Figure 5 shows the Output Data Format for the 5480. Because
it uses only 23 of the possible 35 outputs, 12 of the
bits are `Don't Cares'.

Post Edited (pixlo) : 6/29/2006 3:41:38 PM GMT

pixlo

Chris Savage (Parallax) said...
Do you have the brightness control set up properly?

Is that a common problem with these m548x drivers? As most of these chips are obsolete there is very little
information on using them other than the datasheets.

A usenet post suggested the following:

Mark Scotford said...

I am using Picbasic Pro and I need help in understanding the Shiftout
Command. I am driving a two digit 7-Segment LED Display using a M5481
display chip. This uses Clock and Data and needs a 35 bit long data
stream. I have tried several times to get the Shiftout command to
work, with no joy. I think I am getting the Variable\Bits part wrong.

The only way I can drive this chip, is not to use the Shiftout
command, but write a routine which makes the clock input high then low
35 times, at the same time turning the data input on at the
appropriate moments, which is a very lengthy process. Anybody had any
experience with this command/driver chip?

then later:

Mark Scotford said...

My first reaction is that I was trying to get it to shiftout the whole
35 bits in one go (most probably an invalid option), I did not think
to shiftout in three goes. Also thank you for the tip on shifting the
last three bits.This I will experiment with, thanks for pointing me in
the right direction.

so for kicks I tried sending 36 individual high and lows to the data pin while alternating the clock pin high every other bit, like so...

high 3
high 2
pause 50
low 3
high 2
pause 50
high 3
low 2
pause 50
low 3
high 2
pause 50
high 3
low 2
pause 50

to make:

pin 3 (clock): 1 0 1 0 1
pin 2 (data) : 1 1 0 1 0

not that that would work... but i'll try anything once. Would that work?