RS-232 confusion

[Christopher4187]

This may be a case of a dog chasing his own tail but I'm trying to figure out if what I'm looking at is an RS-232 signal. I've tried various baud rates and it seems to be something around 9600 as that's the most consistent data I can see on the screen. Each piece of data is 500uS. If it's not RS-232, can anyone identify the communication protocol?

[richard]

definitely not rs232 and not a Manchester code , its a rtz(return to zero) stream of some sort . where does it come from and is there any identifiable start/stop sequence


bi-phase ?

[rsocor01]

Also, RS-232 signals go from -12V to +12V. What you have there is some kind of 1-wire serial communication going on.

[Christopher4187]

It is one wire communication. The voltage level is 18VDC. What you see in that picture the output from a RS-232 IC. I can identify the start/stop and know what each piece of data is. The problem is, I have to do it with my oscilloscope and eyes as opposed to using a PIC to interpret it.

Each bit is 500uS. The high/low within the 500uS bit is predetermined. In other words, it may be 200uS off and 300uS on or 400uS on and 100uS off but I know before the packets starts. The entire packet is 100mS long.

What is the best way to get this information to a PIC where I can log it instead of staring at my oscilloscope?

[richard]

if you can't identify the device then a complete transaction from line idle to send/receive and back to idle with a data translation would help . guessing seems pointless

[Demon]

If you know it's RS232, why not send into a PC?

Assuming you don't have a Saleae logic probe or other device capable of reading signal.

Robert

[Christopher4187]

If you know it's RS232, why not send into a PC?

Richard has confirmed it's not RS-232.

definitely not rs232

if you can't identify the device then a complete transaction from line idle to send/receive and back to idle with a data translation would help . guessing seems pointless

I can identify everything.

I don't really need to convert the data into a "readable" format, all I need to do is accurately measure the on/off period in each bit. For example, if I'm able to do something like this, my job is complete:

Bit 1 // On - 100uS / Off - 400uS
Bit 2 // On - 350uS / Off - 150uS
Bit 3 // On - 100uS / Off - 400uS
Bit 100..........

I'm using an 18F87J50 and tried using an ECCP input but was never able to get it working. Then I figured if it was some type of standard communication I'd be able to push the easy button but that's not a reality now.

[Demon]

I got confused by this I guess.

What you see in that picture the output from a RS-232 IC

If you only want to know ON/OFF duration of 18v signal, why not use the signal to control the base of a transistor that would send 5v to a PIC?

Then I would use interrupts on that pin and use a timer.

Robert

[richard]

getting info here is a bit like pulling teeth , but lets assume a 48mhz clock and that timer4 is free along with a pin like portb .2

some vars
bit_is var byte 
I_data var PORTB.2    'lets call this the input

a routine
bit_test:
bit_is =0
t4con=0   ' timer off
tmr4=0   ' clear timer 
pir3.3=0  'clear timer overflow flag

while ( ! i_data) : wend 'wait here till input goes high  to sync up
 while ( i_data) : wend  ' wait here till input goes low
t4con=6     ;  prescale 16 timer on
 while ( ! i_data) : wend  ' wait here till input goes high  
 if  pir3.3 then bit_is=1  ;  offtime > 340 uS  
else
  bit_is=0 '  the offtime < 340 uS 

endif
       
 return

it could be nicer if
1 does the line idle high or low
2 what does a start bit look like
3 how many bits in a session
4 100uS / Off - 400uS a 1 or a 0 ?

[Christopher4187]

If you only want to know ON/OFF duration of 18v signal, why not use the signal to control the base of a transistor that would send 5v to a PIC?

Then I would use interrupts on that pin and use a timer.

I'm a novice when it comes to the timers and interrupts. Seems like a good idea that's worth trying.

1 does the line idle high or low

Idles low.

2 what does a start bit look like

1ms on/1ms off

3 how many bits in a session

100

4 100uS / Off - 400uS a 1 or a 0 ?

It's a 1 but it can vary. Before the packet is received, I will know what is a 0 and what is a 1.

[richard]

this should get close to it

define OSC 48
some vars
bits var byte[13] ' an array to hold the 100 bits
I_data var PORTB.2    'lets call this the input
bit_count var byte
 
 
'a subroutine
bit_read:
t4con=0   ' timer off
tmr4=0   ' clear timer
bit_count = 0
while ( i_data) : wend  ' wait here till input goes idle
while ( ! i_data) : wend 'wait here till input goes high  to sync up
t4con=$7e  '256:1// count of 45 or so = 1mS
while ( i_data) : wend  ' wait here till input goes low
t4con=0   ' timer off
if (tmr4 >43 ) and (tmr4 < 47) then goto start_ok '  bit of latitude might need more
goto bit_read
start_ok:
while bit_count < 100
tmr4=0   ' clear timer 
pir3.3=0  'clear timer overflow flag
while ( ! i_data) : wend 'wait here till input goes high 
t4con=6     ;  prescale 16 timer on
while ( i_data) : wend  ' wait here till input goes low
t4con=0   ' timer off
 if  pir3.3 then bits.0[bit_count]=1  ;  ontime > 340 uS  
else
  bits.0[bit_count]=0 '  the ontime < 340 uS 
endif
 bit_count=bit_count+1
wend
       
 return

[Dave]

As far as the IDLE high or IDLE low state, looking at the signal with a scope, where is the bus signal level when NO data is being sent? I would imagine it to be high, as most single wire communications use a passive sourced current to stimulate the bus wire (usually a resistor). That way there can be more than 1 device listening as well as more than 1 device able to answer a request. This is done by pulling the line LOW thru an open collector transistor or N channel fet.