This is the adapter.



You could wire that up however you wanted in terms of the pins to the DB9 connector. As long as you matched it on the other end you would have a straight serial cable.



I keep a sheet in onenote that reminds me how ive done it so every job is the same.

RJ45 Pin 1-blue of DB9
RJ45 Pin 2-orange of DB9
RJ45 Pin 3-black of DB9
RJ45 Pin 4-red of DB9
RJ45 Pin 5-green of DB9
RJ45 Pin 6-yellow of DB9
RJ45 Pin 7-brown of DB9
RJ45 Pin 8-white of DB9

So at this point blue is pin1 at each end orange is pin 2 at each end and so on and so forth. So you could take those anywhere you want into the DB9 adapter as long as your 2,3&5 (and anything else) match at each end.

Stamp where do you get those?

Thanks for all the info people have contributed to this thread.

Stamps got it! that is all we use as well. i get mine from custom install supply but they can be had from anywhere. i do keep 1' 3' 6' black standard preterminated serial cables for racks and such but anything longer we cat5 it and let it rip.

I usually get them from MCM [Link: mcmelectronics.com]

As edizzle says.. you can pretty much get them anywhere. Search for "DB9 to RJ45 adapter".

We buy BAGS of these from monoprice:

[Link: monoprice.com]

Almost 10x cheaper than the MCM ones - BUT - I like the length of the screws on the MCM ones much better. You can unscrew those with your fingers. You almost always need a little screwdriver for the monoprice ones. After a few times of climbing behind a rack and realizing I need to get back up to go get a screwdriver, I think I would have rather paid the extra $3. :)

thanks all! this really helps. Ill let you all know how it turns out

Sorry Stamp but I have to call this like I see it. RJ45 to DB9 adapter = lazy technician. I cannot tell you how many times I have had techs blame my code for a device not communicating only to find out that they grabbed an already terminated adapter off the van and it was pinned wrong. If you solder you will always get it right. If your engineering department or programmer haven't provided you with schematics for pin-outs on the DB9 connectors then they are not doing their job correctly. Our techs used those damned adapters on most of the jobs when I first started with the company. Even with pin-out schematics there would always be at least one problem with the wiring on every job. We work in the electronics field and yet I meet many people who cannot solder to save their lives. How is that possible? Learn to use a portable soldering iron, build a few soldering jigs for male and female connectors (with a custom made heatsink on the back so you dont melt the pins) and get the job done. It's cleaner and more professional looking in the long run.

There are many reasons NOT to use CAT cables for serial communications. The primary reason is the noise cancelling effect of the twisted pairs. Yeah yeah, you've never had a problem. Congrats on having an awesome streak of luck. Serial communications is limited in distance to 50' at 9600 Baud, by specification. The length drops with lower wire gauge, higher Baud rates and/or sheilded cable. Yes, sheilded cable actually shortens the length the signal can travel. I don't expect you to believe me but you might want to pick up a copy of Serial Port Complete by Jan Axelson for confirmation. Amazon usually has it available for not a lot of money. It is an in depth study of serial communications from spec standards to wire lengths to UART differences. It won't help you with the protocols but it will help you understand how serial communications works. The good news is that Belden sells actual cables designed for RS-232/RS-422 spec and other cables for RS-485 spec. The Belden 9533, 9535, 9537 and 9539 cables are what you would want to use. You will rarely use a 9539 except to make a full 9 conductor cable and the only reason to buy this cable is to make a 50' serial cable for loading processors from the comfort of the couch. The 9533 is going to be the most common cable you use as it is TX, RX and GND only. For stocking purposes you could keep the 9535 on hand at all times. This would cover 98% of your installs as it gives you TX, RX, GND, CTS, RTS.

I have seen many instances where the installers have only pulled a CAT5 cable for serial communications and it didnt work. The length was about 35 to 40 feet on each of those problem jobs. The units had to be switched to IR control and that sucked from my standpoint as I lost all real feedback from the device so all my time programming (and counting on that feedback) was wasted. It will happen at some point in your career and inevitably it will happen when you must have 2 way comm and cannot get a new wire there. Pull the right wire and you won't have that problem.

You might want to invest in a cable tester like this one:

[Link: google.com]

Serial communication, per se, has no single spec and rs232 has no distance limitation specified. What spec are you referring to?

When i see guys pulling stranded cable for serial connections i think "outdated technician". I also think "one more cable spool to keep in stock, order, carry around, more termination tools etc". Pulling Cat5 to a control point means that you can use serial today and you have the necessary cable in place for ethernet later.

Pinning adapters wrong is why i take a bag of 20 and pin 15 of them straight and 5 null and mark an S or N on the outside of them. But this really doesn't have anything to do with solder over rj45-db9 adapters. The same tech who pinned the adapter wrong is equally capable of soldering incorrectly. And.. its much faster to correct an RJ45-db9 mistake than it is to correct a soldered connector.

The engineering department provides lots of documents but im the type of technician who always tries to write a System Builder file to test control & AV zones before the programmer arrives so there aren't any issues with incorrectly pinned connectors. Most lead technicians can use System Builder and i strongly suggesting having them create a test file with econtrol for basic AV routing to each zone and control of AV gear and and subsystems that are in and done.

19 years ago i sat at a soldering station for 2 hours a day making straightwire and monster cable interconnects/speaker cables for almost 2 years. Then i spent another however many years screwing around with soldering stations @ dirty job sites or finished homes wasting countless hours trying not to burn the house down or myself.

The day i am told that i have to solder a serial connector is the day i pack up my things and go somewhere else. Im done with it. I keep a cheapo battery operated soldering iron in the far reaches of my tool kit for the rare occasion i need to 'doctor' an unfinished product that a manufacture shipped me. Thats they only time it sees the light of day.


Ive never had a problem in 10+ years of using adaptors and Cat5. Are you guys using Crestron? Pro2's? This is our typical combo and no issues.


My AV gear is usually within 20' of a room box or processor but subsystems have been @100' away over Cat5e. Pool/spa control, lighting processors,security systems. I haven't had any communication issues. I would wager that a Cat5 is a better cable to have anywhere vs the typical serial cable. Much more flexible IMO now and tomorrow.

The 50' thing is allover the web. Search for "rs-232 max distance" or "max distance serial cable". At one point i knew what all of it meant. The only thing i remember now is countless responses/posts/accounts of that distance being easily doubled/tripled/quadrupled over Cat5.

I can't tell you how many times a programmer has told me the problem is with wiring only to have me disconnect and test everything from the rack to the room end only to report "no issues with hardware" and have him say "oh.. there it is.. my bad". :)

Yes, lots of claims about the max length, but the actual eia standards for rs232 (or eia232) do not actually specify such, afaik.

Jan Axelrod'd book is heavy on technical, short on practical.

I don't recall a mention of max distance in the book, but I know that a good 100+ feet over twisted pair is usual.

I dug out the book.

Axelrod says 66 ft for STP, 106 for ribbon cable, and 133 for UTP. These are based on typical cable capacitances, receiver input capacitance of 100 pF, and a baud of 20 kbps or less.

These are calculated, based on TIA-232 recs.

I don't know if 100 pF is usual or not, for an 'input capacitance'.

Which version of the book? I believe there was an updated version released a couple years ago. The one I am referring to had lengths of 35 feet for shielded cable and 50 feet for unshielded cable. I am in the process of moving so I cannot find my copy at the moment. It is heavy on technical but the background information has actually helped me diagnose some problems in the past. I also enjoy reading about technology so I didn't find it as dry as some might. I don't remember my copy having any discussion of UTP or STP cables.

You are right that the spec doesn't state actual cable lengths but a formula for calculating cable lenghts. Newer UARTs can transmit further it seems but I still try to stick with the 50 foot marker. I also try to avoid twisted pair cable when possible. Ethernet has taken a lot of the serial control away so wire length becomes less of an issue assuming you stay within allowable wire lengths for Ethernet communications, which I think is 328 feet or so on copper. I haven't had enough coffee today to pull specs out of my head and I don't feel like Googleing for an answer. Regardless, I still prefer a soldered connection with proper serial wiring. It looks cleaner in the rack and more professional all around.