Uh, CAT5 cable ? Have you tried with just the specified 3 wires, and no "adapters" some cat5 cables are crossover cables, most are "straight".

DB9 connectors, for GT reporting need only 3 wires,
pin 2 to 2 Data
pin 3 to 3 Data-
pin 5 to 5 Gnd

GT Series RS232 Commands :
http://www.xantrex.com/web/id/1626/docserve.asp page 6
3. Protocol
The GT inverter implements a standard RS232 protocol. It is designed to be a data communication
equipment (DCE); therefore, on the DB9, pin 2 is the transmit data pin and pin 3 is the receive pin. To
connect to a computer ( DTE ) a straight through cable is needed.
The settings for the RS232 are 9600-N-8-1.

Just a quick update, the system ended the day by generating the most killowatts ever and it didn't get started until 10:40. At this point it will be faster and easier to borrow a laptop from the neighbors that has a serial port and hook it up with serial cable to see if it has an effect. Baby steps, One step at a time. I'll let you know.

I have not been able to pinpoint the problem exactly, but there is a problem. At first I thought it might be the cat5e cable because I was able to duplicate the problem on both computers.
What I have done is borrowed a laptop with a serial port and hooked up a serial cable, no adapters and let it run for two days. It ran perfectly, in fact it finally made me proud of the system install because the data file was so staright forward and clean; highest two day production ever for me on file.
Then I hooked up to the same laptop a purchased 10' Cat5e cable between two new adapters,(serial to Cat5, male and female). After two days - no problem.
I messed up by trying to update the laptop with newer virus definitions, updated firewall, and windows update,and hooking the laptop up to the internet through my router. At the same time I added an extension or basically another 10' of cat5 cable with a straight through connector. Somewhere in that mix of mulitple reboots the laptop was able to shut down the inverter. When I say shut down I mean in GT View all the voltages read correctly, DC Amps to .16, but the MPPT goes to 600 and the production goes to 0. I verify by looking at the inverter screen which reads "Insufficient Solar Energy" As soon as I unhook the data cable, the inverter will start to produce. Hook it, up porduction to 0.
I went back with the laptop to square one with just the serial cable. no problems. I hooked up the two cat5e cables with the pass through connector and the serial to cat5 adapters. After 3 days, one of which was a rainy day session, never a shut down even for a minute, and I now have new record highest production days on file.
At 7:30 this morning I verified the system producing about 400watts and steady. I made sure the main computer and the laptop had the newest Gt View 1.9 beta and identical settings and both programs were running. All I did was unhook the cat5 cable from the laptop and within two feet plug it in to the main computer. The inverter immediately shut down. From the main computer I disabeled Anti Virus, firewall any other programs that were running, (but not some background processes). The inverter would not produce until I hooked the cat5 cable back into the laptop. The inverter immediately showed production. The laptop is not hooked up to the internet but has anti virus and firewall running.
I thought it could be the serial to cat5 adapter that I have in the back of the main computer so I unscrewed the one in the laptop serial port and screwed it into the main computer with the cat5 cable wtill attached. I made sure Gt View was not running and then went outside to read the inverter, which had shut down. I unscrewed the adapter and hooked it back up to the laptop and immediately went outside to read the inverter again. It read 1800 watts which was less than 10 seconds since I plugged the cable back into the laptop.
I did verify comm port settings on both computers earlier and I just confirmed they are still the same.

So this is what it is.

Like I said in an earlier post, this may have happened to me once, with a 3 wire db9 cable, while connecting to laptop. I did not try to repeat it, but maybe there is some sort of ground loop happening, or excess leakage current in a serial port. I wonder if a db9 - USB isolator like is needed for the fishbowl systems would prevent this.

The more we find out, the more we know. It really seems odd that a serial data monitor port can disable the inverter. I wonder if there is noise that is emulating a shutdown command, or if it's a leakage path swamping out a chips internal pull up/down. I know the GT can have it's display programmed by the serial port, and that may also be the factory maintance/firmware flash port. Mike

As I said eariler, GT-View has no means to command the inverter to do anything, what you descibe is connecting your PC gnd to the CPU gound is causing the unit to power up/down ...

One thing to check is measure with a DVM the db9 pin #5 of both the GT inverter and your PC and make sure they are both at Earth Ground ( ground wire of outlet ). Its possible you have a floating ground or even 120vac on the gnd if the outlet is mis-wired and maybe thats whats causing issues. I've never run without proper grounding, so not sure what that would do to the GT.

Also, PLEASE run the system with a 10 panel string and see what that does ... you need to rule out that this isn't a minimum operational voltage issue

It wasn't easy or fun but I did connect 11 panels to a string and just ran the one string for a day. The same problem occured so I ruled out minimum string sizing.
More interesting results when I took meter readings of ground on recepticles. I found there is a .28v differential on pin 5 versus earth ground or the grounded recepticle of the computer. The inverter cable measured 0v differential. This result prompted all kinds of questions in my mind.
1st let me explain that when I borrowed the neighbors laptop and hooked it up with the rs232 cable and then switched to the adapters to run the cat5 cable, the system ran perfectly for 3 weeks so I lost interest in doing any more diagnostic testing. The problem had something to do with the main computer. I failed to mention before that the main computer is hooked up to a ups backup and a separate power control center surge protector. I had to figure out if either added to my problem. Short answer...no.
The racking and panel equipment ground is straight through the inverter and terminates on the grounded busbar of the main distribution panel where it has a separate conductor linked to the grounding electrode rod.
The house was built in 1959 with knob and tube wiring. Sometime before me the main panel box was upgraded from fuse to circuit breaker. The only grounded recepticles are in the kitchen and baths. I have put in GFCI outlets and replaced all two prong outlets with three prong recepticles throughout the house. About 5 years ago I replaced/upgraded a window in the bedroom that has been converted to office central. It involved removel of all sheetrock on the wall, added insulation and I ran a separate romex line for more dedicated circuits in office central. I also ran romex for other dedicated circuits in other areas of the house at that time.
The UPS and computer with printers was plugged into an original recepticle that has an open ground. To solve my GT 3.8 shutdown problem, all I had to do is either not monitor the system by removing the data cable or move my computer plug-in cord 3 feet to a different recepticle, which is one that I installed with the romex wire to the distribution panel.
I don't think my home is wired differently than millions of american homes with the exception that they probably do not have a grounded receptcle with which to plug in their computer. It is amazing that a .28v differential is enough to bring down a system or at least make it run far less than optimum. Now am I going to trace that one circuit and find out why the neutral and the ground have a .28v differential? Probably not. I have more important things to do. Hope this post has been informative for others to read. I have enjoyed reading other topics in this forum, but time has prevented me from reading all of them.

Thanks for the help,
Clyde

Thanks for updating us with your findings; it is always good to know what causes odd issues!

A couple of points,

First, having a voltage difference is normal when comparing a neutral return to a ground return. Neutral is current carrying; earth ground isn't and should never be current carrying, that why you will see a difference with a DVM.

What is most likely happening is when you connect your PC to the GT, you are making the earth ground bond THRU the CPU card of the GT because the ground is floating at the computers receptacle.

Computers have all sorts of emi noise and themselfs depend on proper earth grounding, sending that noise into the GT is what causes the issues and in your case, your PC without proper grounding is shown to cause shutdown issues

The GT's serial port is not opto-isolated .. It directly connects to the GT's CPU board which IS opto-isoltaed to the rest of the unit. I have pointed this out to Xantrex before and your situation shows this can be a rare but real problem source.

An important FYI:

Changing outlets from two to three prong without adding the proper wiring is illegal and can kill someone with a mis-wired appliance, that’s the whole purpose to make sure live AC can't be in-touch’s reach. If someone touch’s a water-pipe, water source and an ungrounded appliance, death can be the result, I’m not being dramatic here, this is WHY we have electrical codes

I realize this is more common in older houses, I hope you can take the time to either put back to two prongs OR pull proper wiring. By the way GFI's are useless without true 3 wire wiring.

Hopefully, you can now enjoy your new system and log many wonderful hours with GT-View! Again, thank you for share this information!

to reinforce what solar guppy has said, my home was built in 1910 and it had knob and tubing until i changed it out while my parent's home was built in 1958 and was i believe the armored bx and some romex(may have been added later, not sure), but no knob and tubing. also be very sure of the wiring to the 3 pronged outlets as putting any of them in the wrong position is very dangerous also. also note that many homes were wired with only 2 pronged outlets, but have the 3 wires as some just grounded the gang box or left the ground wire open. a look inside of an outlet reveals much. i assume you wired the updated ones correctly and i'm not implying you didn't or don't know how to as this is for everyones general knowledge and safety. it helps make things run correctly too as you have found out.