Capturing High Low Pulses Code Problem

[Suraj143]

I’m trying to capture hi low pulse timing & save them in a binary form.

I have attached my drawing that how the pulse streams are coming. In every stream the timing is T & 2T length of pulses.

If the high pulse is 2T then the binary must be 11
If the high pulse is T then the binary must be 1
If the low pulse is 2T then the binary must be 00
If the low pulse is T then the binary must be 0

I want to arrange like above format. Any idea how to do. No need code just tell me the idea. Do not suggest capture modules. Want to do it from a general I/O pin with a timer.

Thanks

 

[AtomSoft]

Seems simple.... Since T is the normL length the count every T if it's high the OR in a 1 and shift left if it's low AND in a 0 and shift left. If u do this in every T u will get everything correct

 

[AtomSoft]

to clarify... let's assume T= 10 us When u start counting check if high
if high then or in a one
if low and in a Zero

shift left by one

wait 1T (10us) then check if high
if so or a One
if low And a zero
shift left by one
repeat until complete and It should be good

 

[Mr RB]

(edit) deleted; sorry I posted wrong code, i didn't read your question clearly.

 

[colin55]

How long is "T" in uS?

We have to know how much time you have for housekeeping.

What PIC chip are you using?

Use Timer0 or Timer1

 

[Suraj143]

Hi for everybody thanks for the support you all giving.

The input pulses varying from 200uS to 2000uS.For example T=450uS,2T=900uS so the whole string goes with those two values.On next string it can start from T=550,2T=1100uS.

I'm using TMR0 I don't like TMR1.

I can prescale & give inputs.

 

[Suraj143]

Ok I'll tell whats my method.

First you wait for a mark pulse.When mark pulse detected turn on the TMR0.Wait until space pulse & when space pulse detect save the "Mark_Time" & reset TMR0.Wait until next mark pulse,when next mark pulse save the "Space_Time".

Now we have two values."Mark_Time" & "Space_Time".

Now subtract “Mark_Time” from "Space_Time". Check the carry bit status & arrange the result registers. Likewise on every after Period subtract “Mark_Time” from "Space_Time & arrange the result registers.

The problem with my method is when both the mark & spaces equal the result will be wrong.

 

[Mr RB]

Why does the timing change from one string to the next string???

 

[Pommie]

We need a little more information,

How many bits are involved in a transmission?
Is the line low between bytes/words?
Is there a change that signifies the start of a byte or could a zero (or two?) come first?
Does a byte/word always contain 00,11,01&10 sequences?

Mike.

 

[Suraj143]

We need a little more information,

How many bits are involved in a transmission?
Is the line low between bytes/words?
Is there a change that signifies the start of a byte or could a zero (or two?) come first?
Does a byte/word always contain 00,11,01&10 sequences?

Mike.

Very good points mike. You have come up closer.

*Per transmission they are about 25-30 bits.
*Those are not bytes those are just high low pulses.
*The start of a string always 1(T) or 11(2T)

To detect end of the frame I'm checking the "space time" has gone above 2000uS.because the maximum T is 2000uS.So I can identify the frame is end or not.

Mike as you understand lets assume you get 2T “Mark Time” & a 2T “Space Time” on start of the frame. According to my method if I just subtract “Mark Time” from “Space Time” It will be “10” format. But it should be “1100”.Now see the error. I think you understood now.

 

[Pommie]

I think you need to capture the whole data stream as a series of time values and look for the shortest pulse to work out the "baud" rate. If you are not guaranteed that the stream will contain a single T pulse then you will have to assume that the time for the shortest pulse that you received is a single T and see how many bits you have. If the bit count is less than 25 then recalculate with T=T/2.

Mike.

 

[Suraj143]

I think you need to capture the whole data stream as a series of time values and look for the shortest pulse to work out the "baud" rate. If you are not guaranteed that the stream will contain a single T pulse then you will have to assume that the time for the shortest pulse that you received is a single T and see how many bits you have. If the bit count is less than 25 then recalculate with T=T/2.

Mike.

Ok thanks for the input.I'll tell what I'm doing.

I'm trying to make a universal Infrared Remote Decoder.I need to store particular remote incoming data & compares the same remote code is coming on next time.

I'm ignoring start bit & the first space time.After that only I'm capturing incoming data.If you study all remote codes based on T & 2T format.

Many will tell that you can't make a UIR decoder.But at least I can make a one working for many remotes but not for all.

Any method to capture those stream.