/*
My current logs (for one inverter) are a little over 4MB (yes that's mega)per day. But that is because Fishbowl currently logs data once per second. I beleive Henry is working on a a version that has selectable sampling intervals that should cut the size down quite a bit.
*/

Well, that's only around a gigabyte a year per inverter, and hard disks are cheap...

Does FishBowl close daily logs so they can be read out over (say) a network connection, or does it just create One Giant File? What happens if the computer crashes while the file is open?

/*
The port on the Xantrex does not supply any power.
*/

Yes it does. I'm not talking about the DB9, I'm talking about the port you plug the Remote Meter into, which is next to the DB9...

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote

DON'T EVEN THINK ABOUT IT , you don't know what the remote draws , you don't know the internal power buget and you don't know things like the fact that the startup has the lcd backlights off to minimize power-up loads ... you didn't design this product , don't pretend that you can do anything to it ! .. Strong words , but I need to get thru to you OK ?

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote

See the mini-itx , there perfect , I run 3 of them 24/7 and my home is net positive on grid interactive solar generated power , they use less than laptops since there headless (and I use laptop drives in them)

/*
Thake a look at mini-itx information to learn more
*/

Aha! Perfect, in fact http://www.mini-itx.com/store/hush.asp looks pretty durn close to what I want, plus or minus a cusomization...

Thanks!

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>
Use "perl" and sendmail ... it would be a 8 line script for your special needs

<HR></BLOCKQUOTE>

I already use perl to reduce the current data. It would just eliminate that step. A "mailbox file" is not about mailing it (besides I would use postfix instead of sendmail :-). It's just an agreed upon file that two cooperating programs can use to communicate with each other. One program drops data into it, the other one picks it up.

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>
Why would the program miss anything , if the inverter is online , its sending data once a second
<HR></BLOCKQUOTE>

It's the polling program that can miss a pickup of the data (although this is probably not an interesting problem in this scenario). We use cricket at work with thousands of datapoints and stuff happens.

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>
I don't understand what you are saying , all DC readings are values , like volts , amps , watts
<HR></BLOCKQUOTE>

Let me rephrase that. Make that a request for total DC (PV) output, similar to total AC output.

Hmm, that's only interesting when looking at the inverter efficiency. I guess I'm still used to the instrumentation around a battery based system were DC (PV) output is important to track.

Thanks for the info the mini-itx systems, another item for the project list.

arnold

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by William P N Smith:
....the environment the inverters and computer will be in ranges between 64 and 92 degrees F (average about 80), with an average humidity of 77 percent. <HR></BLOCKQUOTE>

Is there some reason the computer needs to be located by the inverters? I've got mine about 75' away and indoors where the environment is more hospitable.

/*
Is there some reason the computer needs to be located by the inverters? I've got mine about 75' away and indoors where the environment is more hospitable.
*/

Yeah, I suppose that's do-able, though with the USB opto-isolators you probably want to keep everything fairly close together.

Aha, perfect solution! http://www.vpi.us/usbc5.html

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by William P N Smith:
Yeah, I suppose that's do-able, though with the USB opto-isolators you probably want to keep everything fairly close together.

Aha, perfect solution! http://www.vpi.us/usbc5.html<HR></BLOCKQUOTE>

The USB opto-isolators can be located by the computer and all you need is a single pair of wires to run to each inverter. The link you provided is for extending the USB ports themselves which would probably be overkill. Depending on the distance, you could probably just use some cat5 cable which has 4 twisted pairs to run from the opto-isolators to the 4 inverters. I'm using 4 conductor round phone wire which works just fine. The baud rate is only 9600 so long runs would not be an issue.

I had forgotten that it's only a single pair of wires to each inverter (tx and gnd), and was thinking you'd want to keep the potentially hazardous voltages inside the SunTie case, but I suppose if you properly insulated the wires and ran them in a separate conduit you could use a single Cat5 run for everything.

Thanks!

Why extend the USB when you can extend the RS-232 interface by just using a longer cable? 75' is easy for serial.

arnold