Recording and re-generating RS 232 signal

[LAVA 1.0 Content]

Hi all,

I'd like to record a RS 232 signal to disk so as to be able to regenerate it (to simulate a device), the RS 232 speed is 57600 Baud.

I read that I should over-sample at least 16x if I go for a Analog Input board because RS needs steep rising edges.

Has anybody ever done something similar ? Does it work ?

I have never used HS DIO (655x or 656x series), maybe this would be more appropriated to record and regenerate an RS 232 signal ?

Any comment, ideas, though or orientation will be very appriciated, thanks in advance.

[Grampa_of_Oliva_n_Eden]

QUOTE(TiT @ May 29 2007, 11:09 AM)

Hi all,

I'd like to record a RS 232 signal to disk so as to be able to regenerate it (to simulate a device), the RS 232 speed is 57600 Baud.

I read that I should over-sample at least 16x if I go for a Analog Input board because RS needs steep rising edges.

Has anybody ever done something similar ? Does it work ?

I have never used HS DIO (655x or 656x series), maybe this would be more appropriated to record and regenerate an RS 232 signal ?

Any comment, ideas, though or orientation will be very appriciated, thanks in advance.

It has been years and NI no longer makes the board ...

Re:over-sampling...

The chips (when last I looked) do X10 over-sampling themselves so to do it correctly would required even more over-sampling. If they want to look a tht rise time response then ingore the data transfer rates and look at it as an anlog signal and do the Nyquist math.

Ben

QUOTE(Ben @ May 29 2007, 11:49 AM)

It has been years and NI no longer makes the board ...

Re:over-sampling...

The chips (when last I looked) do X10 over-sampling themselves so to do it correctly would required even more over-sampling. If they want to look a tht rise time response then ingore the data transfer rates and look at it as an anlog signal and do the Nyquist math.

Ben

I have also used a "ESCC-PCI" from Commtech to do this type of work. It gave me the flexibility do send bad packets.

Of course the FPGA stuff should allow similar results these days.

Ben

[peteski]

QUOTE(TiT @ May 29 2007, 11:09 AM)

Hi all,

I'd like to record a RS 232 signal to disk so as to be able to regenerate it (to simulate a device), the RS 232 speed is 57600 Baud.

I read that I should over-sample at least 16x if I go for a Analog Input board because RS needs steep rising edges.

Has anybody ever done something similar ? Does it work ?

I have never used HS DIO (655x or 656x series), maybe this would be more appropriated to record and regenerate an RS 232 signal ?

Any comment, ideas, though or orientation will be very appriciated, thanks in advance.

TiT,

I have a couple of quick comments (I've been swamped recently, not much time to contribute to this forum, unfortunately!)

Beware of the voltage levels on the RS-232, if this is a "standard" RS-232 device! (and standard can be a very difficult term to apply to RS-232!) I've measured voltages of +/- 11 volts using an oscilloscope. If you use a standard DIO at TTL levels, you may need some circuitry or a box to properly handle the mismatch between TTL and RS-232 levels.

BTW, why not just listen using the computer's RS-232 port and copy down the byte stream to a file, then spit that file back through the RS-232 port when you want to simulate? It might be faster, better, and cheaper that way. Is there something particularly "interesting" you expect to need to do that an RS-232 port may not quite be able to facilitate?

-Pete Liiva