RS232 over CAT6
alexsquared
Posts: 166
Has anyone done an install doing RS232 control over CAT6? We have an install where from the rack location to the floorbox, there were numerous Augmented CAT6 runs. We used a couple for baluns and planned to use one for RS232 control. It's only about a 30 foot run. CAT6 is terminated onto jacks on both ends. I have done a cable test to verify pinout. What I did was take a patch cable and lop it in half, then solder DB9 connectors on each end. I used my breakout box to verify that the solder pinout is correct ( 2-3, 3-2, 5-5 for an Infocus LP640). I tried switching RS232 ports on the NI and still no go. As an FYI, the jacks/mod plugs are not shielded. At this point, I'm half tempted to just use the CAT6 as a pull string for a 22/2 shielded and be done with it. Any other suggestions first?
Thanks.
Thanks.
0
Comments
http://www.easyadapters.com/products.php?cat=8&PHPSESSID=60ce6d68b0b62ffc69b4796c92528f2a
We have also gone through patch panels in the past with out any troubles. Maybe I've just been lucky?
DB9 to RJ45
They work well, are quick to terminate and can be reconfigured quickly with a pin extraction tool.
--John
Obviously you can use these with any type of wire, 4 con shielded or CAT-5/6 depending the enviroment you running the cables in and when the pin out is wrong which I do alot I don't need to cut off the ends and toss a connector, I just loosen a terminal and move the wires. Much less aggravation.
Again, I would love to use these, but would really need a hood to enclose them. Has anyone found a hood that works for these?
Thanks.
http://www.amx.com/products/STS.asp
Kevin D.
Thanks.
Technically, twisted pair is not ideal for RS-232. Practically, it rarely matters, but when it does, you are right, you have to drop back to shielded, non-twisted. One thing about the RJ->DB-9 type adapters is that you don't have to use twisted pair cabling, you could use the flat line-cord stuff ... but I wouldn't do it for long runs.
This is one of the many examples of wire being used that is not spec for the protocol. The spec might be a bit tight since I've seen all kinds of cases of running rs232 long distances over cat 5 or cat6 just fine. I have one install right now where the designer tried to run two ipod docks over cat 5 and have found that it's just a bit too long. the ipods drop offline all the time and the meta data has all kinds of erros and whatnot.
I've quit yelling at them about rs232 over cat5. I just warn them that if it doesn't work, don't come crying to me. And in return, I don't beat them up when it doesn't work.
We're both happy when it does work.
The run is short enough. I would suggest making sure the TX and RX are not sharing a twisted pair. Use a twisted pair for TX and a twisted pair for RX and a twisted pair for Ground. 6 wires total. EMF shouldn't be a problem unless the wire is running close to an unshielded dimmer or florescent ballast.
There is actually a standard which prescribes what 'pins' on an RJ45 are supposed to be used for what purpose. If you look at the standard for ethernet cables (Or/w,Or,Gr/w,Bl,Bl/W,Gr,Br/w,Br) and compare to the RJ45/RS232 standard, you can see how the pairs work out.
Recently I bought some of those RJ45 to DB9 adapters and pinned them up (5 wire, just in case) so that I could reduce the number of cables that I carry around with me -- use Cat5E UTP cables for the RS232 connections. Just today, I was working on an old antique Cre$tron system with a 100' Cat5E UTP cable and had absolutely no problems whatsoever. In the recent past, we've used the same setup to communicate with some biamp, ClearOne, and LifeSize gear and have had no problems. Highest baud rate was 38400 -- maybe there would be problems at higher rates, I don't know.
Also, when communicating with standard RS232, the limiting factor is capacitance and not resistance. Current levels are so low that voltage loss over any reasonable cable length, even with Cat5 or Cat6, isn't going to hurt. The problem is the capacitance which will tend to "smear" the signals -- no longer get nice, sharp pseudo square waves. I'm thinking that using multiple pairs will probably exacerbate the problem because capacitance in parallel is additive.
So, I've had no problem using Cat5E for RS232 and our installer has been trending towards using it recently and we've had no difficulty. I suppose for long runs it may be necessary to spec a lower capacitance cable, but for most cases, it shouldn't be a problem. That's my opinion, anyway. Free advice, and all that.
My experience has been that this causes problems, and I have definitely had runs that worked fine if RX/TX were on the same pair, but not if they were combined. It probably varies with cable length.