RS232 - Does it require a termination?

[poynton]

Hi.

I have a data logging system which takes several serial inputs.

We had a lightning strike nearby a couple of days ago and that's when my problem started !

It is only with one channel of data. As far as I can tell, the other 3 are fine.

It's a rather ancient system with a backplane, 386 one-board cpu card and a 4-channel serial card.

I have changed the card, cpu board and hard drive to no avail.

First problem. The card was receiving data at 19200baud - no problem. After the strike, it has refused, even with the changes above. It works at 9600baud. This I can live with.

Second problem. One channel (as above) occaisionally refuses to record the previous 10secs of data flagging an error. This occurs about once per hour , randomly.

The data cable feeding this port has a branch (daisy-chain). If I terminate this with a RS232 tester/indicator the error stops !!

Does RS232 require a termination to work correctly? I cannot find this out.


Andy

.

[DF96]

It is over 10 years since I last played with RS-232. It is a point-to-point system, so a branch is not supported although you obviously got away with it before the lightning strike. Maybe the load of the tester is suppressing some reflections? Or is it changing the grounding?

[pinkmouse]

A long time here as well! I last had RS232 on a Hawkwind tour to drive some custom lighting effects, (before DMX!), and the builder insisted on daisy chain linking, and provided some termination plugs, (one of which I threw at an Austrian traffic warden, but that's another story! ), and these were needed to prevent false triggering.

[Pano]

I used to run very long RS-232 lines around basketball courts for the scoring and stats system used by the officials. Never used termination, never had a problem. Maybe the devices were already terminated?

I still use RS-232 to drive the DCX2496.

[poynton]

Thanks for the replies, guys.

Yes, RS232 is vintage but so is the equipment I'm using ! 396SX with 387SX maths coprocessor, 1m ram and DOS 5. It's slow but it works. (The techs did try migrating the system to a more modern platform but it was not as reliable!)

Just to satisfy myself, yesterday I disconnected the RS232 tester. The errors returned ! After 4 hrs, I reconnected the thing and have not had one since.

So no more messing about, it stays connected. Only another 3 weeks of the project left ......


Andy

.

[davidsrsb]

RS-232 is not terminated. It is a high impedance load system.
It is not intended for cables longer than 50 feet (but is does work way beyond this)
Data levels are typically +/-10V

Has a 0V got disconnected?