how do I furnish a low going TTL pulse for a sonic anemometer using only a serial port

[_jp_]

To perform some syncronized measurements with an anemometer I need to furnish a low going TTL pulse or a low going MARK RS-232 pulse through a serial port, which is the anemometer's only means of communication.  The pulse must be at least 10 micro-seconds wide.  A similar question was posted earlier but for triggering a laser instead of an anemometer.  That solution wasn't completely adequate since he was able to use a parallel port; I do not have that option.  From the other postings I've read it would seem that the best solution would be to create a high-low-high pulse using a three sequence frames and writing this to the port using VISA write.  Unfortunately that hasn't worked nor have I been able to find a way to use the DTR state with the VISA property node, that was also suggested with the laser question.  As of now I don't know if I'm using VISA write incorrectly or some other aspect of the code.  Any suggestions would be very helpful, thanks.

[pflores]

Hello _jp_,

Check out this KB.  This talks about using the serial port to output a constant signal with a small example and may be helpful in your application.

Does the device need to see this pulse on a specific line (Rx, Tx, DTR, etc.) on the serial port?

[Coq_Rouge]

You also have to remember that the RS 232 output may swing from -12 volt to + 12 volt. So if interfacing to TTL type circuits you will need a level translator. Else you may might cause damage on your anemometer

[_jp_]

"Hello _jp_,

Check out this KB.  This talks about using the serial port to output a constant signal with a small example and may be helpful in your application.

Does the device need to see this pulse on a specific line (Rx, Tx, DTR, etc.) on the serial port?"

Yes the device needs to see the pulse on the Tx line, since that is the only output line connected to the device.  (There is no pin connection for the DTR, DSR, RTS, CTS, and RI pins). 

[pflores]

_jp_,

What device are you using? It seems strange that it would be triggered off of a low pulse of at least 10 usec on a serial line. If so, then almost any byte that you send will have a mark bit in the message (except for 0x00 which will have a mark stop bit that). 10 usec is rather fast for serial communication. For example, a generous baud rate of 9600 transmit each byte for 104 usec.

Here is an article on serial that shows an example serial frame.

If this is not what your device is looking for, post some more information on what type of signal it expects on what line.

Cheers

[_jp_]

Pflores,

The device I'm using in an Applied Technologies Inc. Ultrasonic Anemometer.  It requires a 9600 baud rate, 7 bit, 1 stop bit, even polarity, and "none" for flow control.  All that has been properly configured using VISA Configure Serial Port.  All the other information I mentioned earlier came directly from the Operator's Manual provided by the manufactorers.  So from what I understand about the article, the device is expecting that signal.

[Dennis_Knutson]

Do you have an electronic copy of the manual that you can post? Or a link to the manual?

[Henrik_Volkers]

Are you shure about the 10µs?

For me it sounds like a 10ms Break signal (at 9600 baud, it is worth a try)  , when used with an ordinary rs232 cable.  

Send a break and read back the value....

Another pitfall: 

This seems to be an older device, and the old RS232 standard worked with higher voltages (+-12 (15?) Volt) , new(er) Computer provide only 9 to 5V levels , (recent PC none 😞  )  

If that is the case, maybe you can harvest the needed potentials from the Anemometer RS232 output  and use a transistor or a optocoupler to boost your PC signal....