[RE-wrenches] trouble shooting system with Sun Tie

[William Miller]

Dana:

The failure mode of the original ST inverters were a failure in the MPPT 
algorithm.  The inverter would sweep the impedance too high until the 
voltage would collapse.  The inverter would then take several minutes until 
it allowed the PV voltage to rise again.  This would repeat throughout the 
day resulting in a significant decrease in overall output.  You could 
witness this over any given half hour period.

Regarding arrays in dissimilar orientation:  The loss due to differing 
orientations should be about 1%, IF all modules in each string are facing 
the same direction.  This is because the different IV curves are of a 
similar shape when superimposed on each other.  The ST inverter might be 
less able to average these curves and more susceptible to collapse.

Did you check the array fuses?

William Miller


At 03:23 PM 9/11/2008, you wrote:
>Hi all,
>
>I was called out this week to trouble shoot a system I didn't install.
>It's a five year old 1.44kW system with the following: Sun Tie ST 1500
>inverter and 12 Astro Power modules. It only produced 634kWh last
>year. With our sun here, the shading, orientation, equipment, etc. it
>should have produced around 1123kWh.
>
>I've heard a lot about Sun Tie failures, but don't know the details.
>What are the common failure modes? Is it likely that a failure of the
>Sun Tie would leave it operating, but only producing half what it
>should?
>
>When I showed up the system was producing 950W on a relatively sunny
>day - seemed pretty reasonable. After restarting the inverter it only
>produced around 150W. Is the MPPT very, very slow to find that sweet
>spot, or what was going on there?
>
>The system is comprised of three strings of 4 modules each. When I was
>there all three strings were producing about the same current, so that
>leads me to believe it's not a module failure.
>
>To make things more fun, the modules were installed at two different
>orientations. One string is at 38d slope, 30d east of true south,
>while the other two are at 36d slope, 75k east of true south. This
>certainly has a detrimental impact on the performance, but it doesn't
>seem to me that it could account for all of that energy loss.
>
>Thoughts?
>
>Thanks a lot!
>
>Dana