OK, I've tried to figure out the Xantrex calculator program, and I'm probably more confused than before I started. I think it calculates for one controller, so if you have two controllers on one XW inverter you can just add the amount up.

But, can you have a different number of solar panels on each XW-60-150 charge controller?

Also, it seems to indicate that you can only have at most three Sharp 224 panels hooked up together, and the charge controller can handle either three or four strings of three panels each.

So am I now limited to either 18 Sharp 224 panels, with 3 strings of three panels on each controller, or can I have 21 Sharp 224 panels, with 3 strings of three panels on one and 4 strings of three panels on the second one? Do the number of panels have to be a multiple of 3?

If you go back to strings of two panels on each controller, you could do 4 strings of two panels each, but if you wanted more than 16 panels would you have to hook a third XW-60-150 controller to the XW-6048 inverter?

Just about all charge controllers are limited to a maximum of 150 volts, and cold temperatures is when this condition is most likely

3 panels in series ( using 24V ) is the limit. You then have a total wattage limit of about 3kW for a charge controller. I looked up the specs on the Sharp 224 watt panel and its a real oddball, somewhere between a 18V and 24V panel, the only one with those voltages

If you used a panel like an 195 watt evergreen you could have 4 panels to a string as those are 18V panels. Then you would have 3 parallel strings for a combined wattage of 2340 for each XW-60-150 ( total system wattage of 4680 )

I would recommend the Evergreen 195's ( or 190's ) over any other panel. Evergreens power tolerances are the best in the industry and its a standard voltage. Sharp is an odd-ball one off voltage and they make there solar cell wafers 1/2 to 1/3 the thickness of other manufactures, I would not be selecting those panels for my system.

Also if you ever have to replace one, it could be impossible if the manufacture stop production which they do almost every year, if you have a common voltage of 12/18/24V you have many more options. Evergreen has made the 190's for close to 6 years now I believe

Whatever you decide, yes each parallel string must have the same number and type of panels ( voltage in each string need to match the other parrallel strings ), so that is what your incremental wattage will be as in your example, ( 3 ) 224 watt panels means each increment is 672 watts. Having different amount of strings on the two charge controllers is technically fine, I like to have them the same as it makes monitoring the system for problems easy as the two units should be performing within the tolerance of the solar panels for side by side comparisons

How critical is it to have all strings for a MPPT60 to be the same 'vintage' panels? I have a situation where I have two MPPT60 feeding a XW-4548 inverter charger battery bank of 396Ah AGM's. One MPPT60 has 2 new strings of 5 Sharp ND-N2ECUF 142W panels and 1 old string of the previous year 5 Sharp ND panels and the other MPPT60 has 3 old strings of the previous year 5 Sharp ND panels. I "feel" that the mismatched string MPPT60 is not harvesting consistently and that the matched string MPPT60 is harvesting consistently (I don't have any hard data to back up this unfortunately, just observations). Would it be better to create one MPPT60 to have 4 old strings of the previous year 5 Sharp ND panels and the other MPPT60 to have 2 new 5 Sharp ND-N2ECUF panels? Will the battery charging, harvesting, etc be affected negatively by having "unbalanced" MPPT60's or does it matter?

assuming that the panels are both http://www.beyondoilsolar.com/Sharp142.pdf , it should not matter.

You solar array's are actually small enough that you could run all the panels on a single controller so any combination should be fine to try out. As it is now, during the day the reported input and output values should be close between the to chargers and would provide an excellent way to see if there is issues with a string or panel. You can also used the history information from the mppt units to see how close the daily generation is.

This assumes that your in bulk charging, once the unit goes into absorb or float, its very likely one unit will out perform the other as not all the available energy can be used