Zwave Home Automation, Leviton RS232 Serial Interface RZC0P basic wiring for USB Serial Adapter

I've started to work with a ZWave home automation control product from Leviton, the RZC0P-1LW. This device allows the control of Zwave based wireless home automation devices via a RS-232 interface.

The Zwave system is a proprietary system requiring a licensing agreement with Zwave Alliance group. Joining this group and paying some level of fees gives a developer access to programming and related information. The lowest level of membership in 2009 appears to be the Affiliate Member with an annual fee of USD300 and perhaps the requirement to purchase a USD500 hardware development kit.

There are a couple of other ways to programmatically work with Zwave devices, there are several home automation software and hardware systems that put one or more layers on top of the Zwave proprietary protocol. Doing a Google search of home automation and zwave will give you a list of these product. There are even a couple of open source projects that have reverse engineered parts of the Zwave control protocols and devices.

And a third way, in a sense a mini home automation layer, is to use one of these RZC0P devices in an existing Zwave network. As far as I can tell so far, the RZC0P cannot be the primary controller of a Zwave network. And every Zwave network requires one of these primary controllers to add, delete and manage the devices in a Zwave network. However, the RZC0P can be included as what is called a secondary controller. And a small amount of documentation has been created by Leviton to show you how to do basic functions to Zwave devices via ASCII commands to the RZC0P.

I've studied the Zwave products, vendors and public information for a couple of years. More on what I have found about it and my opinions later.

But for now, I wanted to share some tech work I did to get the RZC0P running on a small Zwave network I have set up to explores of of the uses of these devices for both Aging In Place and Energy Management.

I am not a RS-232 expert nor an electrical engineering guru. But I have spend enough time in both of these areas that I knew that I had a problem talking to the RZC0P pretty quickly. I had the documentation on the ASCII commands and the RS-232 configuration requirements. Using these, I hooked the RZC0P up to a IBM Thinkpad with a build in RS-232 port and was able to start communicating with the device right a way. The problems came with I tried to move my testing to a Apple OS X computer with a KeySpan USA-19HS USB serial adapter. I wanted do my testing using the Python language and tools and was more comfortable with using these tools on Linux and OS X.

The problem I encountered was that I could not get the RZC0P to respond to commands when attached via the KeySpan USB serial port. I did not try the KeySpan USB adapter on the Windows machine, but I suspect from reading some posts on the web that I would have found a similar problem. I found several other people that indicated they were only successful in communicating with the RZC0P using a serial port directly attached to the Windows machine.

At first I suspected that I was not correctly toggling some of the RS-232 control signals, however the devices cable and documentation point to, but do not directly spell out, that only TX, RX and Signal Ground pins are required on the cable and no software or hardware handshaking is done.

A further interesting fact appear as I played around with various terminal emulators on OS X and cable combinations. At various points of these changes, the RZC0P would start communicating. I struggled to find the pattern that made it work.


Well bottom, apologizes for my long route to my conclusion here.. I suspect that the RZC0P has some bug or non-standard implementation of RS-232 electrical interface. I found that if I used a RS-232 break out box between the Keyspan USB serial adapter and the RZC0P, with just the three pins, TX, RX and Signal Ground connected, the unit would work consistently. I could plug and unplug it, power it off and it would always come right up and communicate with the Mac software. So what was special about this RS-232 break out box?

The breakout box connects a set of LED from the TX and RX lines to signal ground via pull up resistors. This allows you to visually see the signal states change on these and other lines on the RS-232 specification. The breakout box I was using is a totally passive unit that adds no power or logic to the RS-232 signals it monitors.

So my conclusion is that the signal ground and RX/TX lines on the RZC0P are wired internally in some way that caused the logic circuit to not correctly start RS-232 communications without some electrical connection/kick between the RX/TX lines ad signal ground line.

My solution to the problem was to build a mini 9 pin DB-9 Male to Female adapter that recreated the passive monitoring circuit I found in the RS-232 breakout box.

I am guessing there is a better and simpler solution to this problem, but this seems to be working, and you get some visual feed back of the communications occurring via the two LEDs. Below is a picture of the finished adapter.



Here are the parts and steps:

Parts:
Male 9 pin DB-9
Female 9 pin DB-9
2 - 560 ohm 1/4 watt resistors
2 - LED
wire to connect pin 2 to pin 2, male to female DB 9
wire to connect pin 3 to pin 3, male to female DB 9
wire to connect pin 5 to pin 5, male to female DB 9

Steps:
Connect pin 2 to pin 2, male to female DB 9
Connect pin 3 to pin 3, male to female DB 9
Connect pin 5 to pin 5, male to female DB 9

Connect pin 2 on the male DB-9 connector to one lead of a 560 ohm 1/4 watt resistor
Connect the second lead of the 1st 560 ohm resistor to the anode lead of the 1st LED
Connect the cathode lead of the 1st LED to pin 5 on the male DB-9 connector

Connect pin 3 on the female DB-9 connector to one lead of the 2nd 560 ohm 1/4 watt resistor
Connect the second lead of the 2nd 560 ohm resistor to the anode lead of the 2nd LED
Connect the cathode lead of the 2nd LED to pin 5 on the female DB-9 connector