DMX on 250000 baud receive

[syscoder]

yes i now it.

but i want to know how to use "hsein" command to receive for dmxsginal

is everything true here in the code?can i use it for dimming a led ?

[skimask]

yes i now it.

but i want to know how to use "hsein" command to receive for dmxsginal

is everything true here in the code?can i use it for dimming a led ?

PBP manual, page 75-81. Whether or not the code is accurate or correct for DMX512...well, quite frankly, this is posted in the "MELabs PICBASICPro" forums, not the "DMX512 Questions" forum.

[JEC]

Heh...

That's my code snipped from a year or so ago.

It does work, and it's dead stable. But you need to understand that it's only a subroutine, called from the main program loop. You get to write the main loop.

The reason for not using HSERIN is that I found there was a slight startup delay when using the command. Using PULSIN to find the break, then calling HSERIN took too long and I was losing data. As you may know, the make-after-break signal can in some cases only be a few microseconds long.

At 20 MHz, you get 5 assembly instructions per microsecond, which may not be enough.

When I first wrote the code I didn't dig too deeply in PBP's internals to see what exactly was happening when HSERIN was invoked. But it seemed like an overhead issue within the command.

And note that the 'skip' function which can be used as part of HSERIN is only byte sized. So 'skip 254' works but 'skip 257' doesn't.

It was easier to access the registers directly and be assured that everything was being received properly.

Note that there's absolutely no guarantee that each and every packet coming down the wire is exclusively dimmer data.

JEC

[DynamoBen]

Like most things there is more than one way to accomplish a task. Last year I saw the post from JEC about DMX. While it was nice to hear that DMX was possible with PICBasic I really wanted something that was interrupt based.

After doing some research I found that all you need to do is look for a framing error, after which should be the start code (there may be several framing errors in a row). Over the last few days I have created the interrupt code snip below. While I haven't tested it exhaustively it seems to work.

DimmerCount	VAR WORD  ; Location in DMX string
DMXStartAddr	VAR WORD  ; DMX start address
DMXStartCode	VAR BYTE  ; DMX Start code (Dimmer=0)
Dummy		VAR BYTE
DMX_State	VAR BYTE  ; Stage of DMX reception     

Valid_Break     CON 1
Valid_SC        CON 2
Valid_Data     CON 3

Disable

INTERRUPT_SUB:
	IF OERR Then 						
		CREN = 0
		CREN = 1    			; Clear Overrun Errors
		High Err_LED			; Notify user of overrun error
		DMX_State=0
		EndIF

	IF FERR Then
        	FERR = 0
		DMX_State=Valid_Break
	Else
		GoSub RX_DMX_Data							
     	EndIF
     	
    	Dummy=RCREG				; Empty UART of junk data
	Resume	

RX_DMX_Data: 
	IF DMX_State=Valid_Break Then
		HSerin 20,NoData,[DMXStartCode]	; Start code of 0 = dimmer data
		
		IF DMXStartCode=0 Then 
			DMX_State=Valid_SC
		Else
			High Err_LED		; Notify user of Error in Start Code
			DMX_State=0 
			EndIF
		EndIF
		
	IF DMX_State=Valid_SC Then
		For DimmerCount=0 TO DMXStartAddr; Count incoming data bytes (frames) 
			Dummy=RCREG		; Read unwanted values
			Next DimmerCount
		DMX_State=Valid_Data
		EndIF
		
	IF DMX_State=Valid_Data Then
		HSerin 20,NoData,[STR DataBuffer\Number_Of_Channels]
		DMX_State=0
		EndIF
		
	IF DMX_State=0 Then
		Resume
		EndIF
	
GoTo RX_DMX_Data
	
NoData:
	DMX_State=0
	Resume
		
Enable
End

[JEC]

Very elegant. Nice job, DynamoBen.

[Darrel Taylor]

Well, I can see some potential problems with that program.

So be ready for some more questions from programmer07, oops, I mean syscoder.
<br>

[mister_e]

Assuming it's a snip and a ISR... first observation, something like a L?CALL have to be change but it's a first & fast observation.