Solectria PVI serial comms protocol [Archive]

View Full Version : Solectria PVI serial comms protocol

Dave Schwartz

27th August 2010, 13:50

I've got a microFIT roof-top PV system installed using a Solectria PVI-4000. I'm too cheap to pay over $500 for their communications gateway module (http://www.solrenview.com/downloads/Guide%20to%20SolrenView.pdf) (plus an annual fee for the privilege of being able to send the data only to their site). Although there is the free PVIDAQ program for Windows that can be downloaded and it has the capability of logging the data, it is a bit limited and I don't/won't have an always-on Windows machine within serial-cable distance of the inverter. Fat Spaniel (http://www.fatspaniel.com) seems to have figured it out as well but they want $650 for their module and its still a closed system (they say they have a public API but I can't find it).

I've asked (twice) Solectria support if I can get the data capture interface spec and so far have been totally ignored - haven't even had the courtesy of a reply. I take that to mean they aren't going to make it available.

I'm a fair hand at comms programming and have purchased a computing plug (http://computingplugs.com) specifically for the task which runs a Debian Linux O/S and uses less than 6 watts. So far, I've built the serial cable and have a USB-to-serial adapter to drive it from the plug (and have verified that PVIDAQ runs fine when the USB dongle is plugged into a Windows box).

I've done some preliminary reverse engineering of the protocol and have figured out, I think, the protocol for the parameters PVIDAQ gets from the inverter but this leads me to believe there are gaps and that there are more commands that could get at least the values that keep cycling on the LCD display of the inverter itself (of which PVIDAQ only gets a subset).

Before I go much further I thought I'd ask if anybody has a complete reference to the data acquisition spec or knows where I could get one?

Alternatively, would there be any interest in my sharing the comms protocol as I've reverse-engineered it and following my efforts to implement a C program to do the data logging? I eventually plan to implement real-time insertion into a database (postgresql) and possibly my own web page. I've written programs that interface with postgresql before as well as generating web pages from logged data.

Dave Farrell

15th September 2010, 18:44

Hi Dave,
I am will soon to be running 2 Solectris PVI5000 inverters and am very interested in following your work on the communications to the units. Unforunately all I can do is follow as I have no experience at all in this field. I hope the lack of response to your post does not discourage your posting of the testing and results.

Best of luck.

Dave Farrell

Dave Schwartz

16th September 2010, 08:54

So far I've created a document which describes what I've found out through tracing the comm port activity of PVIDAQ on a Windows machine. It shows the commands and responses for the initial poll/getSN (gets the serial number in a format I mostly understand) and the loop where it gets about 10 parameters (DC Volts, AC Volts, AC Current, AC Energy, Total Energy, Total Hours and 4 more that return zeros - probably for error codes, which I don't have). I'll upload that document tonight.

I've also written the skeleton of a C program that opens the comm port (/dev/ttyUSB0 on the plug computer) and starts to replay the sequence. I can see it send the poll/getSN command but I'm not receiving any response so I've still got a bit of work to do there. I know the USB-to-Serial adapter works because I can talk to the inverter through it with PVIDAQ when its plugged into the Windows machine, however I haven't tested whether a know good Linux program can talk to anything else through it from the plug computer.

The protocol seems to incorporate a message checksum and, since the unit ID is part of the message I presume that checksum will change for different IDs (each of your units will have to have its own ID). When my unit was installed, it was on ID 8 and I used PVIDAQ to change its ID to 1. This caused the checksums to change so that's how I'm sure its part of the calculation. So far I have just been treating the checksums as static info on the outgoing messages since they never change for a given message when running PVIDAQ and I don't yet know how to calculate them (and I'm ignoring them on received messages). If you're interested in using the program, if and when I get it working, I'll either have to figure out how to calculate the checksums or perform more PVIDAQ traces on all the other addresses and have static message tables per unit id (up to 16 it looks like).

Dave Schwartz

17th September 2010, 07:02

Here's the document that describes what I've figured out about the protocol.

Jim Lane

1st November 2010, 13:50

I cannot find a download site for the PVIDAQ software, could someone point me to a download site?

Jim:)

Dave Schwartz

1st November 2010, 15:42

Very interesting... all mention of it is now gone from the SolRenView website. It looks like they're only interested in holding up people for their own version of the interface module ($500) and the annual 'monitoring' fee ($??). Solectria should probably stop advertising 'Free PC Software' on their residential inverters page. Maybe they're afraid someone will crack the 'tinkering' password and jeopardize their certifications.

Honest, I will get back to debugging my app sometime!

Steve Johnson

22nd November 2010, 21:13

Dave Schwartz

23rd November 2010, 08:29

You only need TXD, RXD and ground. TX A and TX B are only used for the RS485 used to link inverters.

The RJ45 end is just a normal ethernet configuration with about 50' of UTP CAT5 cable. The data rate is low enough that the length still works for RS-232. I initially left the connector open so I could swap TX and RX if I had to cross over but it worked first time (using PVIDAQ to a real serial port). PVIDAQ also works with the cable and the serial-to-USB dongle on my laptop, just not yet with my own program on the plug computer.

The problem I have now is that I could only work on the program on weekends... since the inverter is totally solar powered, and I am now at work during all weekday daylight hours, it is dead to me except for daylight on the weekend. Should have thought of that earlier :laugh:

If you don't have PVIDAQ (its still AWOL from the SolRenView website) shoot me a PM with your email and I'll send it.

Steve Johnson

23rd November 2010, 19:03

Thanks for the information, it beats the trail and error technique. I will give it a try this weekend. I did get a copy of the PVIDAQ before it went MIA.

The problem I have excluding work and the inverter is totally powered by solar (I didn’t see that coming) is I will be getting a second inverter installed and the distance to the inverters is 170 feet, so it looks like I will have to use RS485. But before I go there I want to try the RS232 connection with the cable I am building.

I will let you know how it went.

Steve

Steve Johnson

30th November 2010, 17:35

I finally got around to hooking up my laptop to the invert and ran the PVIDAQ program.
PVIDAQ went through the inverter search, 1 to 16 and found my inverter at ID 1.
When I click on the inverter in the inverter list, the ID box was populated with ID 40 and the PVIDQA crashed. I think I know now why they removed the PVIDAQ program, I think they changed the mapping now and PVIDAQ is reading the wrong values.

It looks like I will have to do some more investigation on how to read the values off the inverter.

Steve

Dave Farrell

4th December 2010, 07:29

Solectria models ~3000-5300 are made for them (or pieces packaged) by Motech.
Have a look at the Motech software PVMate Data Manager 1.0.1 I think it may satisfy all your wishes. I'm using it now after playing with what I consider the substandard PVIDAQ monitoring software.

Dave

Dave Schwartz

6th December 2010, 10:39

Thanks for that. Its sometimes hard to find out who is the OEM and who is just rebranding. I have run across several other companies who sell identical looking inverters with their own branding on them.

I've downloaded the software and run it on the same machine from which I ran PVIDAQ. I used the portmon utility to capture the serial port I/O and it does have quite a difference from what I observed with PVIDAQ. Its using the same format of command so it should help in filling in some of the unknowns from looking at what PVIDAQ did. Since it is obviously different, it seems that the two companies probably did not use the same source code but at least were looking at the same protocol spec. Maybe Motech will not be so allergic to my asking for that spec!

Dave Farrell

7th December 2010, 08:21

Dave
There is another version of that software (version 1.4.1 I believe) that you can download. It is approx 200 mb and contains software to communicate with TI devices, you may want to have a look at that. Both versions seem to support all Motech inverters which include their 3 phase units so we will see a lot of information that is not applicable. Also that 220mb download has some instructions as to how to use the software.

Dave Schwartz

7th December 2010, 09:54

Would you have a link to that version? I can't find anything matching that description on the website.

Dave Farrell

7th December 2010, 17:32

http://www.motech.com.cn/products/inverters_detail.aspx?id=b47e0735-c9d4-423b-88cb-1758da96cdbe

On the left side of the page 1/2 way down. 201mb

Dave Schwartz

8th December 2010, 11:14

Thanks - I was looking at the Taiwanese website which has English menu items. Its the only one that Googles when you search for 'motech pv system monitor' which are actual keywords on the page from the Chinese site. Maybe Google still bears a grudge.

Dave Farrell

8th December 2010, 18:14

Dave,
If you figure out the password for this software please be nice and let me know.
Also if you figure out the commands to get the daily production out of the inverter that also would be great.

Dave Schwartz

12th December 2010, 09:37

That shouldn't be too difficult. Anything that PVMate displays, either on the main screen or on the detailed device data page is retrieved one at a time from the inverter. My method of figuring them out is to trap the serial comms using portmon and then match up the data in the replies (usually in hex) to the data from screenshots taken at the same time. Here's a cycle of the comms from PVMate:

1023 0.00013521 PVMateDataManag IRP_MJ_CREATE Serial0 SUCCESS Options: Open
1043 0.00007543 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 85 00 01 95 E3 0D
1047 0.01615261 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 00 CB F9 D3 0D
1052 0.00007655 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 03 06 00 08 A4 49 0D
1056 0.02535099 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 23: 0A 01 03 10 53 6F 6C 65 63 74 72 69 61 20 20 20 20 20 20 20 7E
1061 0.00007319 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 03 0E 00 08 25 8B 0D
1065 0.02522918 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 23: 0A 01 03 10 50 56 49 20 34 30 30 30 20 20 20 20 20 20 20 20 94
1070 0.00006789 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 03 16 00 08 A5 8C 0D
1074 0.02544653 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 23: 0A 01 03 10 53 2F 4E 20 32 34 30 31 30 39 31 30 30 30 38 38 3F
1079 0.00007152 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 17 00 02 74 0F 0D
1083 0.01884485 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 11: 0A 01 03 04 03 8E 00 58 9B A6 0D
1088 0.00006621 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 12 00 01 24 0F 0D
1092 0.01586738 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 00 04 B9 87 0D
1097 0.00007459 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 56 00 01 64 1A 0D
1101 0.01678062 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 01 86 38 76 0D
1106 0.00007431 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 86 00 01 65 E3 0D
1110 0.01764386 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 00 75 79 A3 0D
1115 0.00006844 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 83 00 01 75 E2 0D
1119 0.01587995 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 00 8B F8 23 0D
1124 0.00006928 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 B5 00 0F 14 28 0D
1128 0.03081760 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 37: 0A 01 03 1E 00 32 00 00 00 00 00 00 00 00 0A F3 04 BA 04 B9 03
1133 0.00007319 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 C4 00 0F 44 33 0D
1137 0.03035078 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 37: 0A 01 03 1E 00 01 18 81 00 01 1B C3 00 00 00 00 00 00 00 00 1F
1142 0.00007347 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 D3 00 0F F4 37 0D
1146 0.03212615 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 37: 0A 01 03 1E 00 31 00 00 01 64 00 00 00 00 00 00 00 17 00 00 00
1151 0.00007319 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 E2 00 0F A5 F8 0D
1155 0.03125174 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 37: 0A 01 03 1E FF C7 03 E6 09 25 00 00 00 00 00 01 00 00 00 00 00
1160 0.00013940 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 F1 00 0F 54 3D 0D
1164 0.03238512 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 37: 0A 01 03 1E 00 00 00 00 00 00 00 00 00 00 00 00 03 1B 0F A0 00
1169 0.00009247 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 01 00 00 0E C5 F2 0D
1173 0.03035330 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 35: 0A 01 03 1C 00 00 00 28 00 00 00 00 00 00 00 00 00 00 00 00 00
1178 0.00012013 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 C1 00 01 D5 F6 0D
1182 0.01493514 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 02 F0 B9 60 0D
1187 0.00011594 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 F7 00 01 35 F8 0D
1191 0.01654456 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 03 1C B9 7D 0D
1196 0.00011342 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 01 00 0B 55 CD 0D
1200 0.02741913 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 29: 0A 01 03 16 00 31 00 09 00 45 00 09 2A F8 00 3B 22 60 00 77 00
1205 0.00011286 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 0C 00 0B C4 0E 0D
1209 0.02787449 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 29: 0A 01 03 16 22 60 01 2C 09 2E 00 14 2A 94 00 00 00 04 00 01 00
1214 0.00009191 PVMateDataManag IRP_MJ_WRITE Serial0 SUCCESS Length 10: 0A 01 03 00 15 00 01 95 CE 0D
1218 0.01759469 PVMateDataManag IRP_MJ_READ Serial0 SUCCESS Length 9: 0A 01 03 02 00 00 B8 44 0D
1220 0.00000391 PVMateDataManag IRP_MJ_CLEANUP Serial0 SUCCESS
1221 0.02238581 PVMateDataManag IRP_MJ_CLOSE Serial0 SUCCESS

During this cycle, its retrieving 20 parameter values, which is more then I previously had seen using PVIDAQ. I think I remember seeing the daily numbers on the screens, which is what you're looking for, plus a bunch of others (e.g. total DC energy, heatsink temp, AC frequency, etc).

From a brief comparison of the PVMate comms vs the PVIDAQ comss, its not very similar in the request code/response formats. The structure of the request looks the same (confirmed by the 'send raw command' window which identifies the fields by name) but only one of what I thought were the command codes is in common so there must be several ways to request similar parameters. Responses are also differently formatted so it will take some more work to match them up.

David Gladstone

22nd February 2011, 10:41

PVIDAQ software is available at http://www.solrenview.com/downloads/pvidaq.zip

David

Very interesting... all mention of it is now gone from the SolRenView website. It looks like they're only interested in holding up people for their own version of the interface module ($500) and the annual 'monitoring' fee ($??). Solectria should probably stop advertising 'Free PC Software' on their residential inverters page. Maybe they're afraid someone will crack the 'tinkering' password and jeopardize their certifications.

Honest, I will get back to debugging my app sometime!

Erik Henrikson

2nd October 2011, 19:44

Greetings Dave - I know this is an old thread, but I'm new to this whole thing. :)

At any rate - I am wondering how you made out in the end. What came of this project? Naturally I have a vested self-interest here as I'm about to go live with a pair of Solentria inverters (2500 and 4000) and would very much like to find a way to monitor them (I too refuse to pay for their monitoring service).

So - any update? If you hit a brick wall I'd like to know before I too run into it.

Thanks in advance...

David Gladstone

2nd October 2011, 20:26

Hi Erik
The PVIDAQ software previously noted on the thread works fine. I got an adaptor kit to go from 9 pin serial to RJ45 and used a standard ethernet RJ45 cable to go from the inverter to a serial -> ethernet adapter. There's some virtual serial port software that comes with the serial -> ethernet adapter that must run so the PVIDAQ software can see the inverter without being physically wired. I can now monitor my solar output over my home LAN and can tunnel in via VPN from outside the house. I never did figure out how to screen scrape the PVIDAQ software but ended up not caring since I can access my own data as described. If you are running a pair of inverters you will need to use RS422 in a daisy chain configuration rather than a simple RS232 as I did. The instructions for this are included in the Solectria manuals.

So would I do it again? No, I'd go with Enphase micro inverters for a number of reasons:
1. The software is now free from Enphase.
2. The Enlighten web based interface is much more sophisticated than PVIDAQ
3. The micro inverters are more efficient under partial shade conditions (one shaded panel does not bring down the whole string).
4. The wiring from the micro-inverter array is standard romex, not heavy gauge DC.

Hope this helps,

David

Erik Henrikson

2nd October 2011, 22:04

Hi Erik
The PVIDAQ software previously noted on the thread works fine. I got an adaptor kit to go from 9 pin serial to RJ45 and used a standard ethernet RJ45 cable to go from the inverter to a serial -> ethernet adapter. There's some virtual serial port software that comes with the serial -> ethernet adapter that must run so the PVIDAQ software can see the inverter without being physically wired. I can now monitor my solar output over my home LAN and can tunnel in via VPN from outside the house.

Thanks for the quick reply - I'll have to look up the components in your solution as that sounds like a reasonable approach for me as well.

I never did figure out how to screen scrape the PVIDAQ software but ended up not caring since I can access my own data as described. If you are running a pair of inverters you will need to use RS422 in a daisy chain configuration rather than a simple RS232 as I did. The instructions for this are included in the Solectria manuals.

So would I do it again? No, I'd go with Enphase micro inverters for a number of reasons:

As to the choice of central inverters - I may have jumped the gun on putting this array up when I did as I didn't find an installer that seemed comfortable with anything other than the central inverter design. When I asked about micro inverters the general attitude was along the line of "too untested in the field" and "adds too much cost up front" etc... I believe it really came down to the installers not having enough experience with the design, but I digress into a different discussion...

Hope this helps,

David

Thank you - it does!

Rob Beckers

3rd October 2011, 08:03

David, there is one thing I would like to clarify because this is often misunderstood: A partial shaded panel does not bring down the whole string.

Panels have bypass diodes, causing each panel to be divided in three sections: If you look at a panel in portrait orientation, standing it on its short side, most will have 6 columns of cells in it. Shade part or whole of two columns and its bypass diode will activate, taking that section of the panel out of the string. The rest of the panel, and the string, will continue to work normally as before though!

Where things get hairy is if multiple strings are connected in parallel: Take out a panel due to shading and the voltage of that string will drop (it will be lower by the voltage of the panel that is now bypassed). This will cause the voltage of the other string to follow, in effect causing the same one panel loss on the other string. If you have three strings in parallel this means one shaded panel causes the loss equivalent to three panels! A good installer will know this and configure strings to either minimize this or make it a non-issue (we use Power-One inverters where strings are rarely put in parallel, because each inverter has two independent inputs, so panel shading had very limited effect).

So where do micro inverters shine? A hard shade (some object projecting a clearly outlined shade) cannot be salvaged by any (micro) inverter. The shaded part of that panel is essentially dead for power production. No inverter can make something out of nothing. Lots of diffuse shade, and/or having multiple sections of panels facing different orientations/tilt-angles, so those panels still get a reasonable amount of light but not the same on each panel, that is where micro inverters can help. They run each panel at the panel's optimum power point, and those power points can be very different for different modules under those conditions.

Having seen (and done) many, many solar PV installations I can tell you that those latter conditions are pretty rare. Most shading issues involve hard shade. Diffuse shade due to far-away objects is uncommon (that only happens in winter when the sun is low on the mid-day horizon, and there's little energy to be harvested at that time of year anyway). With hard shade a regular string inverter could have done just as well as micro inverters if strings are configured properly. Having multiple roof orientations and not enough inverter inputs to give each their own string is the only real reason in practice where micro inverters are needed (and personally I like to use optimizers such as Tigos in those cases). Despite all the marketing, string and micro inverters will do just as well for regular roofs or situations involving hard shadows.

Truth is that what sells micro inverters (ie. Enphase) is their monitoring. People like their Web monitoring and feel more 'in control' when they can see what each panel is doing. At the same time you are adding many points of failure to the solar installation with micro inverters, and if this is on a roof I would think long and hard before doing that.

-RoB-

David Gladstone

3rd October 2011, 13:36

Rob,

Thanks for the thorough review. You obviously have thought this through from an installer's point of view.

As an end user of both types of systems, (Solectria and Enphase) I have to say that I like the Enphase solution better.

--As you point out, the Enlighten web software is very nice.
--It is easy to diagnose exactly which panel / inverter is performing poorly.
--No need for heavy DC cable
--Much less space in the basement where I now have many controllers (PV and Water)

As for power generation, you are correct, there is no measurable difference.

David

Paul Bailey

3rd October 2011, 19:50

I'm sure theres lots of others as well that outperform the PV Watts estimates as well. I'm sure the data is well scrutenized and possibly biased toward Enphase but just nice to see some are getting more power than expected from there system See Attached article

Paul

Ralph Day

4th October 2011, 06:33

My Enphase 10kw microFIT installation outperformed the expected too. 12kw of pv expecting 1200kwhr per year = 14400kwhr. Our first year saw a yield (as paid by Hydro One) of 16080kwhr. Are PVwatts numbers the same as the NRCAN numbers? I don't know, but when you're harvesting more than you're expecting it's a nice surprise.

Ralph

Rob Beckers

4th October 2011, 07:44

Everybody outperformed (reasonable) estimates over the past 12 months: We simply had more sun-hours than average. Everyone also had brand-new modules, and in case of poly modules manufacturers under-rate them because they have an accelerated power loss in the first months of use (it then stabilizes and goes down at the same rate as mono panels).

FYI, for roof mounted solar here in Ottawa that's perfectly south facing on a 30 - 40 degree tilt angle roof I use 1140 kWh/kW/year as a conservative estimate. For ground mounted I use 1180 kWh/kW/year since the panels stay a little cooler. Those are conservative numbers that will be outperformed almost any year for the first number of years (until the panels start to age). With that in mind, I would have given your array, Ralph, 14,160 kWh per year. In reality I would expect more the first 10 years, less the last 10 years of the contract, to about even out to this over 20 years.

B.t.w. the NRCan numbers are slightly higher in estimated yield vs. my numbers. As I said, I like to keep it on the conservative side. Underpromise & overdeliver is what I prefer. Keeps customers happy.

-RoB-

Dave Schwartz

5th October 2011, 11:58

Original poster here... and back on topic...

I did run into a problem simply with opening the comm port through the USB-to-serial adapter from the plug computer and I just haven't gone back to working on solving it. Linux is less forgiving that way. Seemingly better things to do anyway while the sun shines (and I would have to work while the sun shines because the inverter goes totally dead in the dark).

To be honest, the urge to monitor has lessened as well - probably because the novelty-factor has worn off. The darn thing just sits there, keeps on working and the cheques keep showing up. Since the inverter is right where I park when I come home at night, I just read the daily and total numbers from the readout once a day and that satisfies me for now. I probably will get back to it this fall because once its dark when I leave for work and dark when I come home, that strategy will no longer work because since the unit is totally powered by the PV, the display is dark when its dark (and, no, not even the brightest full moon is enough to wake it up).

Hans Buehler

10th December 2011, 14:33

Hi there,

I'm a new solar user and have a PVI 5000. PVIDAQ was not working well for me (can't find the inverter) and I had the same problem with PVMate DataManager.

My question is on the silly little "termination" switches inside. How should they be set? I have the left one "on" and the right one "off". I just have a single inverter and I'm trying to connect RS-232 so I just need to know where to have the switches. Installer left them "off" left and "On" right. I switched these to "On" left and "off" right, but I'm not sure if I have it right.

also, I'm using Cisco blue cable RJ-45 to serial plug. then serial to USB into my PC. I know the USB to serial cable works (on com2) because I've synced an old Palm V over that serial cable.

any hints on 1) what settings to have on termination switches or 2) cables to use?

thanks!

Hans Buehler

12th December 2011, 19:18

Dave Farrell

13th December 2011, 07:11

Only the left RJ45 connector supports RS232 (when facing the inverter) and this switch should be "ON". Right switch "OFF".
See Solectria manual pg 24-26.

Hans Buehler

15th December 2011, 20:51

Hi David,

I’m very interested in the work you are doing. I have a PVI5000 inverter installed and I wish to monitor it and I don’t want to pay Fat Spaniel or any other company.

I’m at the beginning stages of putting together my serial cable and I noticed in your posting you created one. Could you offer any advice on how the RJ45 connects to the RS232 connector?

So far I have the following:

RJ45 pin --------- RS232
1 TXD ----------- 3 TD
2 RXD ----------- 2 RD
3 Not used ------
4 GND ---------- 5 GND
5 GND ---------- 5 GND
6 Not used
7 TX A (RS485) - Not sure what pin
8 RX B (RS485) - Not sure what pin

Any help would be appreciated.

Steve

Steve, I just had to figure this out for my own PVI5000.

Here is what worked for me and PVIDAQ software and PVmate datamanager both could find the inverter and collect data!

RJ-45.... DB-9
Pin 1 Pin 2 - TX
Pin 2 Pin 3 - RX
Pin 4 Pin 5 - Ground

The solenview manual actually provides info on what pin is what pin. see page 8 figure 4.2
http://www.solren.com/SolrenView/downloads/SolrenViewManual.pdf

I also found this software was better than PVIDAQ... it's called PVmate data manager - get it here...
http://www.motechsolar.com/products/invertersdetail.aspx?view=5000E_4000E_3800E_2800E

Ben Chess

2nd October 2014, 16:28

Thanks Dave Schwartz for that incredibly helpful document. For anyone curious, the checksum algorithm used in requests is CRC16-Modbus, in big-endian notation. The checksum is calculated excluding the first "version" byte (0x0A).

Eddie Parra

3rd June 2016, 18:16

I wish I would have found this thread prior. I just posted this now:

Solectria PVI RS485 Port
http://www.greenpowertalk.org/showthread.php?p=30786#post30786

I have a PVI7600. All the references I have found to people reverse engineering the RS485 interface are all using smaller models. I presume the SW is the same, but if anyone can confirm this, I would appreciate it. Thanks.