Dot matrix display help

[simonden]

What file did you load? I changed the message, should be legible, read the 93c46 again and show me what it says

 

[Ronnie Dent]

I must be doing something wrong! when I select read it shows what you
see below but it can't be what's on the chip as I just get the same
stuff as before? it says in the top data buffer, not device

 

[Ronnie Dent]

I think this is actually what's on the chip even though it verifies ok with your hex file?

 

[Ronnie Dent]

Right I think I have figured it out! when I burn your file Simon it is indeed working and is how you wrote it, but when placed back in the display it must be getting written to by the processor chip? does that make sense? as each time I remove it from the display it reads as above!

Thanks Ronnie

 

[Muppz]

EDIT: as a side note a 93c46 is a 'Serial' EEPROM which means that you can either program with a programmer or via a serial interface hence the RS232 connectors,

Sorry but I have to correct you here, serial in the case of a 93c46 means the data is accessed serially one bit at a time from the memory cells the chip has an i2c interface that is basically clocked data using a clock pin and datain and dataout pins.

Normal eeproms are known as 'parallel' because you access the whole width of the bus in one go.

Muppz

 

[Ronnie Dent]

EDIT: as a side note a 93c46 is a 'Serial' EEPROM which means that you can either program with a programmer or via a serial interface hence the RS232 connectors,

Sorry but I have to correct you here, serial in the case of a 93c46 means the data is accessed serially one bit at a time from the memory cells the chip has an i2c interface that is basically clocked data using a clock pin and datain and dataout pins.

Normal eeproms are known as 'parallel' because you access the whole width of the bus in one go.

Muppz

Thanks Muppz all alien to me! but all help appreciated.

I'm still wondering how this device alters once it has been run in the board?

 

[Ronnie Dent]

A reply from UDC-

"Hi Ronnie regarding your cash kong the rom image we don’t have as i say we only have
dot matrix with 10p scrolling msg which might not be a good idea to use."

I wonder if he means that's all it displays just scrolls 10p 10p 10p......

I also wonder if they would lend me either Winband or the 93c46 whichever holds the image I need? I doubt it somehow!

 

[obcd]

You will need the 93C46 or it's image.

Just say you are fine with the 10p message.

The 93C46 is not using the IIC protocol. IIC only uses a clock line and a bidirectional data line.

I suspect that there is a sort of check on the message in the 93C46. (Like a crc or a hash) If the check doesn't match, the chip is reprogrammed by the microcontroller with that default text you see scrolling. For that reason, the '10p' 93C46 would be handy as it might show where the checksum is.

 

[obcd]

Could you tell us what you see as first symbol of the scrolling message and what the last might be?

 

[Ronnie Dent]

Hi obcd, thanks once again, I think we are making progress, I'm fairly sure it starts with !" and ends with "i as in the picture above but never shows UKVAC rules as Simon has set it, but will need to to double check that.

Thanks Ronnie

 

[simonden]

Sorry but I have to correct you here

Please do - and you are absolutely right

The original code was a complete character set and just looked and there does appear to be a checksum that I didn't put in, I can redo with the checksum.

The first character is a space (not an exclamation mark - that is second character) and the last is a tilde.

The character set showing is:

 !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[]^_`abcdefghijklmnopqrstuvwxyz{|}~

I am at a different computer just now, so will look at putting the checksum in the file later so you can see if that makes a difference.

 

[Ronnie Dent]

Thanks Simon, that would be great if you could do that for me.

If that goes well does that mean it could be programmed to display whatever we want? if so that would be cool

Thanks Ronnie

 

[Muppz]

The 93C46 is not using the IIC protocol.

you're correct, my mistake. I have a nasty habit of calling all serial protocols i2c as when I started in the electronics game it was the only one around

 

[Ronnie Dent]

Sorry but I have to correct you here

Please do - and you are absolutely right

The original code was a complete character set and just looked and there does appear to be a checksum that I didn't put in, I can redo with the checksum.

The first character is a space (not an exclamation mark - that is second character) and the last is a tilde.

The character set showing is:

 !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[]^_`abcdefghijklmnopqrstuvwxyz{|}~

I am at a different computer just now, so will look at putting the checksum in the file later so you can see if that makes a difference.

It displays exactly as you listed above Simon, but I can't tell the space! it displays in 4 blocks stops at 7 and starts at 8 then at 0 starts at P then stops at g starts at h. does not show the 2 lines of UKVAC

Thanks Ronnie
Ronnie Dent2016-06-24 10:16:07

 

[Muppz]

if the controller software was written in C maybe the strings should be 0 delimited?

Just a thought.

Another thought if it is using straight ascii then maybe bit 7 denotes red or green, try a character value or'd with 128.

Muppz2016-06-24 14:15:57

 

[Ronnie Dent]

Thanks Muppz, the display seems to flash red green all the time like sparkle, even the boot logo does this so I think the hardware controls that?

Tried a few test files with Simon but results were the same, also tried lifting pin 3 DI of the 93c46 but it failed to display anything apart from one random burst of part of the alphabet, but it only done that once?

Thanks again to all who have responded with ideas it is very much appreciated

Thanks Ronnie

 

[obcd]

I understand the idea of disabling writing to the 93C46 by lifting the DI pin, but this also disables the abillity to send command bytes to the eeprom chip which would explain why it isn't working. Even if you could somehow write protect the chip, it's unknown if it will display the message when it's detecting a bad checksum. it might as well jump to a routine that initialises the eeprom with the text we are seeing, and jump again to it's display routine afterwards that calculates the checksum.

So basically, it shows 4 lines of 24 characters.

Does the demo text you see now also switch between red and green?

Did someone figure out how it's calculating that checksum? With only one challenge - result, it might be difficult to know if the calculation is correct.

Maybe you could try to read the microcontroller flash contents. I haven't checked if your programmer supports it. The code could be protected against reading, but it's also possible they simply didn't protect. The only way to figure out is by reading it. Just be carefull that you don't erase it.

 

[Ronnie Dent]

Thanks obcd, I know this fault is common to most of these displays, even the distributor have a few that do this!

Yes the display flashes red green right from the boot logo, so might be hardware controlled?

I will need to check my programmer list but sure win bond is not on the device list unless they come from another manfacturer also.

Cheers Ronnie

 

[obcd]

Looks like the microcontroller has features to program it in circuit. It has a 4K flash area for a bootloader and a 32K flash area for the user code. If P2.6 and P2.7 are both grounded, it should go into bootloader mode after a reset. I found a website with a bootloader for that chip, but their might be others and it will be hard to figure out which one they have used.

This is the link:

http://www.marthel.pl/en/tools.php?id=3

Maybe Winbond is delivering a bootloader framework as well since it wouldn't make much sense to use a controller with bootloader options, and ignore that feature.

It's possible that the bootloader produces a message on the serial port to identify it. You will need to connect the device to a pc serial port to figure that out.

For the moment, the 10 pence eeprom contents are the easiest way to attempt to fix it. If we can figure out the checksum, it might be possible to alter the message it displays.

 

[Ronnie Dent]

Thanks obcd, the 10p message one might be off limits as the guy has not even replied to verify if that one works correctly!

A friend of mine has offered to read the contents edit he meant the 93c46 which I can do

Thanks RonnieRonnie Dent2016-06-26 11:29:46