diy solar

diy solar

Charging 48v with small arrays/100 VoC MPPT

jameshowison

New Member
Joined
Jul 30, 2021
Messages
184
I'm planning a 48v system for a travel trailer. I'm trying to figure out the arrangement so that I can purchase the MPPT solar charge controllers.

My understanding is that I need to design the voltage on the solar to two numbers. The minimum is sufficient to charge a 48v battery. The maximum is to stay under the VoC limit of the SCC, including headroom for over-production in cold temperatures. This seems pretty difficult to do with SCC with a max input of 100 VoC (e.g., Victron 100/20 as shown here: https://backwoodssolar.com/product/victron-energy-smartsolar-mppt-100-20-12-24-48vdc/

I'd like to use the smaller MPPT SCC for two reasons. The first is that I see us having three separate arrays, one on the travel trailer, one on the truck, and one portable one. I'm thinking separate given that these will probably see different conditions during the day, the truck might drive off for a bit during the day, and the portable ones might not be set out. The second is cost, the smaller SCC are half the price of the 150 VoC.

Some sources place the minimum for charging 48v batteries as high as 72v. Presumably that's the number under load, so the VoC would be even higher? With a multiplier of 1.25 to account for potential cold weather that seems to leave a very small window for charging 48v batteries from smaller MPPT SCC?

For example I'm looking at the Rich Solar 160W CIGS modules. https://richsolar.com/products/160-watt-cigs-flexible-solar-panel Those show a VoC of 23.3v and a Vmp of 18.9. To get sufficient Vmp to charge 48v I think I need four panels (4 x 18.9 = 75.6) but that puts me at unadjusted VoC of 93.2 or multiplying by 1.25 I get 116.5 VoC. So I can't do 3 panels because the Vmp would only be 56.7 but I can't do 4 panels because the VoC would be too high. So I have to bump up to the Victron 150/35 controller? That's almost double the price and for three small arrays it just seems a lot (like $500 extra).

Am I thinking about this right? I tried to use the more advanced calculators but I can't find the temperature coefficient of the panels.

Any experiences in designing to charge 48v systems from small solar arrays? (btw, I'm committed to 48v due to the 48v DC HVAC system from Hotspot, already bought along with 48v batteries.)
 
The victron needs 5 V over the battery voltage. For LiFePO4 that is ~57.6 + 5 = 62.6V.
 
And what about the Epever XTRA 3415N ?

30A, 138V PV nominal (150V max)

With a Voc of 23.3V you can have 4 (or even 5) in series


Any why did you multiply the Voc x 1.25?
I can't find the temperature coefficient of those panels, but assuming 0.30%/C (which is about average for regular panels - don't know about cisg panels)

If I calculate -22F (-30C) that would be 55 C off the specs, thus 16.5%
I highly doubt if that's a realisitic figure, and still doesn't match your 1.25 estimate. So either you're in siberia, or the panels coefficient is much higher, I think you're safe to use a lower number
 
The victron needs 5 V over the battery voltage. For LiFePO4 that is ~57.6 + 5 = 62.6V.
Thanks, that does give more headroom. And that's calculated with the Vmp not the VoC, right? So for the Rich Solar CIGS 3 panels is only 56.7 so not enough, right? I can look elsewhere but trying to know I'm thinking about this right :)
 
And what about the Epever XTRA 3415N ?

30A, 138V PV nominal (150V max)

With a Voc of 23.3V you can have 4 (or even 5) in series


Any why did you multiply the Voc x 1.25?
I can't find the temperature coefficient of those panels, but assuming 0.30%/C (which is about average for regular panels - don't know about cisg panels)

If I calculate -22F (-30C) that would be 55 C off the specs, thus 16.5%
I highly doubt if that's a realisitic figure, and still doesn't match your 1.25 estimate. So either you're in siberia, or the panels coefficient is much higher, I think you're safe to use a lower number
No special reason for the 1.25 other than I had thought I had seen it quoted somewhere, but reading more I see 1.15 times VoC quoted too. Honestly, vast majority of the time we'll be above freezing, but we would like to be able to winter boondock so being functional down to 0°F would be useful, I reckon we'd have to bail on the trailer below that (unplug the solar I guess!).

Yeah, I searched a bit for temperature coefficients on CIGS panels but didn't find anything.

Oh, and focusing on the Victron since the other components are Victron, but having to bump up to the 150 SCC might just be the Victron price here!
 
Rich Solar got back quickly with the temperate coefficients for their 160 CIGS panels:
PMAX -0.39%C
VOC -0.29%C
ISC 0.05%C

Using https://www.solarchargecontrollercalculator.com/ and -20°C that gives 105V so too much for Victron 100/20.
The victron needs 5 V over the battery voltage. For LiFePO4 that is ~57.6 + 5 = 62.6V.
Thanks, Is that VoC or Vmp? Looking at the Victron spec sheet I think it's VoC. This thread seems relevant but doesn't seem to reach a conclusion: https://diysolarforum.com/threads/5v-mppt-enable-and-voc.28669/

If it is VoC then I should be fine with three panels, giving a VoC of ~70 (@25°C) rising to 81 at -30 °C (and falling to ~66 at +40°C).
 
No special reason for the 1.25 other than I had thought I had seen it quoted somewhere, but reading more I see 1.15 times VoC quoted too. Honestly, vast majority of the time we'll be above freezing, but we would like to be able to winter boondock so being functional down to 0°F would be useful, I reckon we'd have to bail on the trailer below that (unplug the solar I guess!).

Yeah, I searched a bit for temperature coefficients on CIGS panels but didn't find anything.

Oh, and focusing on the Victron since the other components are Victron, but having to bump up to the 150 SCC might just be the Victron price here!
I may be at fault on the "1.25". I use this as a quick check. If I have 25% headroom, I know I am good with pretty much any panel and any environment and don't need to go through the calculations. If I don't have 25% headroom, I look closer.

The 1.25 comes from the NEC table which is extremely conservative (It is calculated on a -.4%/degC)

1635283473802.png

If the absolute minimum temp and this table says it is OK.... it is almost certainly OK..... If this table indicates the voltage rise will be too high, then go to the actual Beta for the panel and do the calculations.
 
Is that VoC or Vmp?
It is Voc. When the batteries are charged, the controller stops all current and this looks like an Open Circuit to the panel.

BTW: This resource talks about the calculations:
 
It is Voc. When the batteries are charged, the controller stops all current and this looks like an Open Circuit to the panel.

BTW: This resource talks about the calculations:
Ok, thanks.

Hmmm, the Victron calculator is also pointing out the risk of power output dropping as temps rise, such that I can get the calculator to recommend the 100/20 for the CIGS panels only if I keep the max temp below 62°C (which is a bad idea here in Texas and with stick down CIGS panels, I'm sure.)

I think one can play around with it here:
 

Attachments

  • Screen Shot 2021-10-26 at 4.53.55 PM.png
    Screen Shot 2021-10-26 at 4.53.55 PM.png
    345.1 KB · Views: 6
  • Screen Shot 2021-10-26 at 4.58.33 PM.png
    Screen Shot 2021-10-26 at 4.58.33 PM.png
    304 KB · Views: 6
I have never tried to put a 100/20 on a 48V battery. I see what you mean about it getting difficult.

I found this in the manual:
● PV voltage must exceed Vbat + 5V for the controller to start. Thereafter minimum PV
voltage is Vbat + 1V.

If we assume 'Full' on the battery is 3.6V/cell, that will come to 57.6V.

That means:
To reliably start, the Voc needs to reach 57.6+5 = 62.6 However, that if the battery is 100% full, the charger does not need to turn on. You might be able to push that down to 58V and be ok. But then we have to look at the temperature because on a hot day the Voc will go down. The panels will be in the sun so lets use a worst case of 120deg (it will probably be less because turn-on is in the morning) With a beta of -.29, that works out to a 16.5% drop. So... 58v x .165 = 9.6V That means we should use a minimum STC Voc of 9.6+58= 67.6V to start the controller
and 63.6 to keep it running.

Using the panel Voc of 23.3V, 3 in series is 69.9V so it looks like the sweet spot for the 100/20 is 3 panels in series. This is low enough for almost any cold temperature but high enough to reliably start the charger.

However: If Vmp is only 18.9, then the operating voltage will only be 56.7V.... so it may not keep it running. I wonder if the charger would cycle on and off.

BTW: 4.4V between Voc and Vmp seems high. You might want to see if other panels work better.
 
Back
Top