Mismatched Panels?

[Riverwind]

I want to hook up a system with 4 x Evergreen 170's and an old Outback MX with 4 new Cdn Solar 230 watt panels. I know I can't make this work on the Outback MX. Is there any way to make them all work on one Midnite? Here are the specs:

Evergreen VOC 32.4, VMP 25.3, IMP 6.72, ISC 7.55
Cdn Solar  VOC 37.1, VMP 29.9, IMP 7.7, ISC 8.22

Riverwind

[Westbranch]

Tell us how you have them arranged and what the battery bank voltage is.  Thse are key to the final answer.

Without doing the math for exact numbers, it could work (but not efficiently) on the MX 60 at near its capacity BUT only to the combined Amps of the lower array (6.72A) assuming you have them in 2 series array configurations, (1 Evergreen + 1 Cdn Solar).

Might even be over Voltage for the MX60... can't remember right now.
 
Series/Parallel may work, but only with the right number of panels and max voltage. 

[Riverwind]

Battery voltage is 24v and the arrangement of the panels is what I'm hoping to get some advice on here. It would be great if only one controller was needed saving the high cost of another length of wire (about 250').

RW

[Halfcrazy]

You are going to take a hit do to the difference in VMP or IMP but if it where me I would probably wire them 1 new and 1 old in series and then do 4 strings. This would put you at a theoretical wattage of 1600 or a max of say 70 amps. Now I am not thinking you will truly get the 1600 do to the mismatch I would expect the IMP to average sort of. I would try it on the MX and see if it is acceptable. You may find (And I think you will) it may be worth getting the Classic and putting the 4 new panels on it.

Ryan

[Kent0]

Putting these two brands of modules in series will work very poorly. Notice that Imp for the Cdn modules is greater than Isc for the Evergreen modules. By the time that the Cdn modules are near peak power, the bypass diodes in the Evergreen modules will be conducting!

With the modules in parallel the voltage is different by 15% so the mismatch losses won't be more than 10% or so. Would you consider that much wire loss as acceptable? I wouldn't; it would be better to use two charge controllers.

[Riverwind]

Some great advice and thank you. Kent are you saying that the max losses would be around 10% if I paralleled all the modules or if I series connected each module type then paralleled the arrays? If I did that I think I would end up with close to 16 amps output and I think around 148 v VOC but I'm not sure how to calculate that - should it be averaged? Just trying to figure out the wire size that would keep the line loss low. The reason I'm considering this is that the alternative of adding another controller and wire will cost over $1k.

RW

[RossW]

I'm not sure how to calculate that - should it be averaged?

The current in a series circuit is the same and is basically limited by the lowest individual device.
The  voltage in a parallel circuit is the same, but individual panels output will vary depending on where you are on their production curve.

If you have one 170W Evergreen in series with one 230W other panel, the current is likely limited to around 6.7A at around 55.2V for around 371W out (vs the 400W you would expect if they were ideally matched).

In parallel the volts will likely be close to 26 at around 13.5A for 351W output. It will depend on where the MPPT settles on as the "best" output, but my guess is that the voltage will be too high for the evergreens optimum and too low for the Cdn Solar.

Cable losses are on top of this of course, and I^R losses on 13.5A will be far higher than on 6.7A
Eg, if you have 100m (50m out, 50m back) of 6AWG copper (13.3 sq mm) you will lose another 5.7 watts just in the cable at 6.7A vs 23 watts at 13.5A, or 100m of 10AWG copper (5.2 sq mm) that is 14W at 6.7A or 58W at 13.5A

[zoneblue]

It would be great if only one controller was needed saving the high cost of another length of wire (about 250').

One  or two controllers doesnt change per se how much wire you need. Unless  you are ramping up the pv voltage, then either way youll be carrying more amps, and hence need more cable. 250 feet (each way???) is a lot of cable, and under the rule of thumb puts you into a high voltage controller.

[Kent0]

Put the Evergreen modules in a series group of four, the Cdn modules in a series group of four, then put the two strings in parallel and connect them to a Classic 200. The peak power voltage, at STC, will be between 101 and 120. As you can guess it will probably be near the average. The mismatch losses will be less than 10%, probably closer to 5%.