Trace/Xantrex SWCA protocol

[debequechute]

So far no response from Xantrex.. I'd like to datalog the available parameters from my SW Plus 4048 onto a PIC microcontroller circuit w/ attached flash card. Not enough power for a full-time PC to run any of the available software.

[conohawk]

I may be able to assist.  My setup is based on Outback equipment.  As you may know, this company is mainly composed of former Xantrex employees.  These fine folks have very kindly provided a communication protocol on their web site (see link, below).

It is possible that the Xantrex SWCA protocol will be similar; especially considering that a third party company makes a monitor and control product (Winverter) that can talk to both Xantrex and Outback equipment.

As with the Outback equipment, you can connect a simple com program (like hyperterminal) to the SWCA (page 16 of the SWCA Users Guide).  If you can do this, logging some of it to a file, you might find similarities to the Outback protocol.

Outback  Protocol

Bon chance,

-Mark

markc@conotech.com

[TomW]

mark;

Very nice bit of information! Be nice if more companies released this type data rather than hide it away behind NDA paperwork. Any reasonably proficient programmer type could convert this data to usable form for plotting and graphing in a computer over long periods. Also allow some interesting control scenarios.

I don't have any of this equipment but wanted to comment on how nice it was to see it made available.

Cheers.

TomW

[conohawk]

Tom,

Dealing with Outback (which is located in Arlington, WA -- about 45 minutes from me) has been a pleasure.  After-purchase customer support has been very good.  

-Mark

[debequechute]

The guy that develops this program sent me some specifications that suggest the hardware interface is proprietary, but the actual communication at the serial-port end of the cable is more of an interactive, menu-based terminal program. You have to send keystrokes and parse the raw output to get the data. I didn't ask if I could post it, so you might ask directly if you want it.

http://www.senecass.com/abelmon

So that's workable, but I'd now like to identify and decode the electrical signals and make my own cable. Something about $150 for a cable and maybe a PIC microcontroller really gets to me.

Anyone willing to break one open and post some close-up photos of the guts?

[RatOmeter]

I someone would give me one of those Trace units, I'd gladly reverse-engineer it and make the information available to all.

With that said, I now return to my regularly scheduled wishful thinking.

[RobNCD2]

Any luck with this?

[(unknown)]

I have reverse engineered the SWCA interface.  I didn't want to spend $150 for something that should have been designed into the inverter to start with.   I opened the inverter and had a look at the control board and the front display.  The DB25 connector was designed initially as a remote display / keypad and the connections seemed identical to the front panel on the SW inverter, with data and power and ground on the same pins.  I took a chance and connected the front panel to the header that connects to the DB25 and it worked! The SWCA is basically an emulator for the front panel with a serial interface instead of the switches and LCD.

I reverse engineered the front panel and put together a schematic. basically two 8-bit latches and the LCD which is one module.  To talk to this thing all is needed is to emulate the LCD interface chip to read the data output and set bits on an 8 bit latch to emulate the push button switches.   I'm in the process of breadboarding a test circuit.

If you would like more information I would be glad to share my notes with you.  Just post a reply in this board.   I hope to put together an ethernet interface to the SW inverter.

[(unknown)]

I'd like to look over your notes, one of my SWCA's quit and I'm not buying another... solar AT J-com.net

[(unknown)]

I would be interested in your information regarding the reverse engineering of the SWCA, specifically the signals on the DB25 and the front panel.  There is other information on the net regarding the reverse engineering but it doesn't seem to be consistent with my measurements.  Please let me know if you can share your reverse engineering of the SWCA/front panel.

The Ethernet interface sounds interesting but you may want to consider one of the peer-to-peer links instead like Bluetooth.  They carry much less baggage in their protocol stack and are lower in cost in their implementation.

Thanks