Mixing modules with slightly different Imp and Vmp?

ahisma · January 2011

I had 24 flawlessly working 12V 100W modules wired in six series strings for 48V using a Outback inverter and MX60 MPPT charge controller. Three of the modules were recently shattered when a tall pine tree fell on them. The charge controller is currently remaining in standby mode due to insufficient voltage being produced by the array with those 3 modules down.

I'd like to replace the 3 damaged modules but unfortunately Sunwize, the manufacturer, no longer makes the modules I have. I've located somewhat comparable 12V 100W modules, however their Imp and Vmp vary slightly. My current modules have an Imp of 4.35A and Vmp of 23V while the new replacement modules I am looking at have a rating of 6.54A and 17.1V respectively (there are a few other possible ones but they all seem to range by 1-2A and 4-6V). Will this cause any issues? Thanks!

niel · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

a vmp of 23v is somewhat odd, but what you are trying to match is not the wattage so much as the vmp voltage total as seen by the 4 pvs in series. this is 4x23v=92v. any combination of like pvs in series that add up to this within 5% is a definite go. just be sure the voc of all the pvs in series do not exceed the controller's operating max calculated for the coldest temp on record for your area.
this higher power pv below is a good match on the vpm, but of course one has to watch controller capabilities too and this might mean adding another cc to make it work so it would then not need to match with separate controllers,
http://www.solar-electric.com/kyocerakc170gt.html

Cariboocoot · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

Just my opinion, but it doesn't look good.
For paralleling panels or strings of panels it is the Vmp that is most important. As Niel said, you're looking to make up a string from the new panels with a Vmp that is within 5% of the old ones. This means you want to stay in a Vmp range of 87.4 to 96.6. Five of the 17.1 Vmp panels would be just under this range at 85.5 Volts, and six would be over at 102.6.

Some say that up to a 10% range is acceptable. How this actually effects performance has been debated here quite a lot. In essence (unless I'm being confused today) you will end up with the controller picking a power point that is less than ideal, so that none of the panels will actually be producing their maximum potential.

Options. There are always options! You could hunt around for panels with specs closer to the existing ones (used, perhaps). That may be a long and difficult hunt. You could go with 5 of the new panels in series and be just a little bit off ideal and accept a small (perhaps insignificant) loss in performance. You could get an inexpensive PWM controller and appropriate panels to make up the missing Watts.

Others here will no doubt have further suggestions. It's a clever bunch.

BB. · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

Solar Guppy has measurements.

My guess is a 10% mismatch in current or voltage will cost a maximum of 10% (or less) in string output.

You have to decide if that is acceptable or not (I think it is ok-but it is your system).

-Bill

mike95490 · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

the MX60 will throttle itself back, if it gets into unsafe (too high power) range, but that may only happen for 30 min or so, a few days per year.

ahisma · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

Thanks a bunch! This helps me understand the problem much better than talking with the reseller and installer. They said nothing of the Vmp differences mattering, only that a module with 5.74A Imp was too high. It is strange these Sunwize modules have such a rare Vmp. Wish I was aware of that before I bought them!

At this point I think I'll look for another module with a higher Vmp. The Kyocera KD185GX-LPU would be perfect if it was just 100W and a bit less.

Cariboocoot · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

Just to clarify:

You would not want to put dissimilar panels in series. You'd be replacing one of your strings with all new panels. The Imp difference matters when connecting panels in series. That might be what your vendor was on about.

BB. · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

It is not that they are rare--It is just that everyone now, for the larger arrays (typically over 100 watts), simply makes the panels for any size/ratings they wish that fits their market/needs/etc... With Grid Tied Inverters and MPPT (maximum power point tracking) the exact voltage does not matter anymore when you have a complete set...

EXCEPT if you break a few or want to increase the size of the array. Then it is a huge pain in the behind.

Sometimes, as another poster figured out... It was cheaper to buy current production 235 watt panels to fit in with a 200 watt array than it was to by discontinued 200 watt panels--even assuming the 235 watt panels only output ~200 watts.

Regarding the MX 60 and current limit... Just to be clear, MPPT charge controllers (like the MX/FM and other MPPT controllers)--They are not unsafe with more wattage than the controller can output--They simply (and safely) throttle back to maximum output current (and usually have thermal shutdown if they get to temperature limits).

What you do have to watch out for--depending on how cold it gets in your area--Arrays with Vmp>~100 VDC may approach, exceed, the typical controller limit for Voc-cold of 150 VDC.

-Bill

ahisma · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

I've located a SolarWorld 80W module that has a similar Imp rating (off by only 0.13A) although it's only 17.7Vmp. So would it be viable to just replace the 3 broken modules with 3 of these and spread the new lower Vmp modules across different strings in the array to minimize the difference? This would make 3 mixed strings have a total Vmp of 86.7V while the 3 original intact strings have 92V, a difference of only 5.7%.

I'm in central Florida and our record low is only 17F so I don't believe the Voc-cold issue would be a problem here either way.

BB. · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

Sounds like a reasonable solution if I understand you correctly.

-Bill

mike95490 · January 2011

Re: Mixing modules with slightly different Imp and Vmp?

ahisma wrote: »

would make 3 mixed strings have a total Vmp of 86.7V while the 3 original intact strings have 92V, a difference of only 5.7%.

Except, that 8 v difference will be noticed
4.3A * 8 V = 34W per string, or 103.2W lower for the entire array.

Just saying, so you won't be too suprised.

ahisma · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

Thanks for all the helpful feedback. I'm grateful to have found such a resource, especially when my vendor has been unsupportive.

Just to ensure I don't have any other issues I re-arranged my array to have 5 complete strings (just to re-cap the damaged modules were spread out across 6 different strings). From my understanding this should still work: 5 strings of 4 100W panels (23V 4A) with a 48V battery bank. However the Outback MX60 charge controller says the input voltage is only 40V, 10V less than the battery bank so it stays in sleep mode even in full sun. Strangely, the battery bank read 58V two days ago and the input voltage displayed 46V but yesterday and today the battery bank read 51V and the input then displayed 40V. I even tried taking out 2 of the 8 (6V) batteries and changing the system voltage in the MX60 to 36V. The input voltage on the MX60 then displayed 29V and the batteries at 38V.

Tried the Outback forums and someone said I probably have more damaged panels or wiring, so used a digital DC multimeter to read the outgoing voltage from the array. Both in the PV combiner box and in the Epanel PV inputs it read 100V, with everything on and connected (so I assume under a load). I then checked the voltage and amperage of all 20 modules yesterday when we had full sun. I checked the voltage both disconnected and connected but only have a 10A multimeter so I disconnected the modules to check the amperage. All were around their rated values except one module that was consistently under by 0.2 - 0.3A. I also tested each string one by one in the PV combiner box, disconnecting the others, and got usual readings: around 100V and 4A. I also tried resetting to factory default settings in the MX60 and double checked proper float and absorb set points for my batteries.

So there seems to be something else up. Is there anything else I could try or look for? Thanks again!

Cariboocoot · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

You need a better meter. When diagnosis potential panel problems you must be able to check the Isc - short-circuit current - and that requires being able to read each panel's Isc or the string in total. Checking the panels individually is best. Panels are a current source, and even defective one will spring up to "full Voltage" in sunlight without necessarily producing any actual power. So you might have 100 Volts going in to the MX60, but it's like 6 penlight cells; put the load of discharged batteries on the output and it gets dragged down to the battery level.

Assuming something hasn't gone wrong with the MX60 or someone accidentally reset its system Voltage to less than 48. :roll:

All diagnosis of this type follows the same pattern: disconnect and check each component individually. If the panels work properly and you combine them and get something other than the expected Voltage/current then there's likely a wiring fault. Et cetera. It's tedious and time-consuming, but it's the only way.

P.S.: If your batteries are down to 40 Volts they need charging desperately and may already be ruined.

ahisma · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

I disconnected each panel and tested amperage in full sun, all read around 4A. I also disconnected all strings then connected 1 at a time to test amperage and voltage, read ~100V and 4A.

Are you saying I need to discharge the batteries and try this again so it's under load?

MX60 is currently at 48V system voltage.

Cariboocoot · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

ahisma wrote: »

I disconnected each panel and tested amperage in full sun, all read around 4A. I also disconnected all strings then connected 1 at a time to test amperage and voltage, read ~100V and 4A.

Right there is a problem. One panel reads 4 Amps, next panel reads 4 Amps, et cetera. Put two, three, four panels in series and the Isc should read 2,3,4 times 4 Amps.

Are you saying I need to discharge the batteries and try this again so it's under load?

MX60 is currently at 48V system voltage.

As far as the MX60 is concerned, batteries always need a little something unless the battery Voltage is above charge set point. Then it will read "BAT FULL". If you don't have that reading then it should be trying to at least Float.

But a string of four 4 Amp panels should read 16 Amps short circuit, not 4.

ahisma · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

Cariboocoot wrote: »

Right there is a problem. One panel reads 4 Amps, next panel reads 4 Amps, et cetera. Put two, three, four panels in series and the Isc should read 2,3,4 times 4 Amps.

Ugh sorry I'm confused. I thought series adds voltage and keeps amperage constant, while parallel does the opposite. But you are saying 4 panels connected in series should add voltage and amperage?

Cariboocoot · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

ahisma wrote: »

Ugh sorry I'm confused. I thought series adds voltage and keeps amperage constant, while parallel does the opposite. But you are saying 4 panels connected in series should add voltage and amperage?

Nope: I am the one who is confused.

You are correct: the series string should read same current. higher Voltage. For some reason I was looking at it in terms of parallel and/or MX output. Probably not getting enough sleep.

I think I'll shut up now before I tell you plus is minus and vice-versa.

Cariboocoot · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

Been reviewing the thread. Please check my understanding of your array now:

Six strings in parallel.
Each string consists of three 100W panels Vmp 23 and Imp 4.35 and one 80W panel Vmp 17.7 and Imp 4.48?

ahisma · February 2011

Re: Mixing modules with slightly different Imp and Vmp?

Ah np I was worried there was some exception I wasn't aware of.

I currently have 5 strings with 4 panels each (100W Vmp 23V Imp 4.35A). I just broke up one string to fill in the gaps and make 5 complete strings to see if there were any other issues going on before I purchased additional panels.

Mixing modules with slightly different Imp and Vmp?

Comments

Categories