LiFePO4 setup with 16 x 272 Ah Cells

snoobler

Solar Honey Badger
Moderator
From what I'm reading this would only be true with 72 cell panels not with 60 cell panels.

True, but 60 cell panels are notorious for being marginal for charging 24V systems especially when they get warm and their voltage drops even further.




Electrodacus is a nifty little piece of hardware, but it is 30 year old solar charging technology.
 

djkarau

New Member
True, but 60 cell panels are notorious for being marginal for charging 24V systems especially when they get warm and their voltage drops even further.




Electrodacus is a nifty little piece of hardware, but it is 30 year old solar charging technology.
Actually MPPT was designed decades ago for AGM and FLA chemistries not LiFePO4! As LiFePO4 will take every bit of energy you offer it there is no need for MPPT to "optimize" the charge for this chemistry. I live in Canada and will have these mounted at least 4" above my roof, so heat is not much of an issue...when in the US, it would be over the winter months so once again, extreme heat conditions will simply not exist.
 

snoobler

Solar Honey Badger
Moderator
Actually MPPT was designed decades ago for AGM and FLA chemistries

True, but shorting panels directly the battery is even older technology.

As LiFePO4 will take every bit of energy you offer it there is no need for MPPT to "optimize" the charge for this chemistry.

Half of the above is not objectively true without qualification, and the other half is false. LiFePO4 has ratings and limitations and can't "take every bit of energy you offer." It can accept notably higher charge currents, which means it actually has MORE to benefit from MPPT. Which charges a battery faster? A 200W panel providing 200W of power to a battery or a 200W panel providing 150W to a battery?

Your asserted relationship is actually inverted. Charge power with LFP is lower than AGM/FLA/GEL due to the average voltage during charge being lower.

MPPT doesn't directly optimize charge, it optimizes PV input power to maximize charge power. If you want to get as close as possible to rated panel power output, you use MPPT. If you're cool with 80% max, then you go with not-MPPT.

I live in Canada and will have these mounted at least 4" above my roof, so heat is not much of an issue...when in the US, it would be over the winter months so once again, extreme heat conditions will simply not exist.

Might make sense in your case.
 

GXMnow

Photon Sorcerer
For a small system where efficiency is not that important, the basic switch charge controller does work, but it is never going to be the best solution. All the talk of it being reliable is one thing, but quality MPPT controllers are very reliable, and far more efficient. Lets take a sample system with 6 typical 60 cell 300 watt panels.

A typical 300 watt panel is 33 volts at 9 amps per panel. With the switch type charge control, you parallel all 6 panels, you get 54 amps of charge current, to your 28 volt bank. This actually looks pretty good. Hitting 1,512 watts of charge power under a perfect full sun solar noon. But how far do you have to run that 54 amps? What gauge wire do you need to make that work? Clicking 54 amps on and off is pretty touch on a relay contact as well.

Now wire those panels as 2 parallel banks of 3 in series. Now we have 99 volts at 18 amps. Just 18 amps can run a decent distance over just #10 wires. Any volt lost is also a much smaller percentage. The charge controller uses DC-DC conversion to regulate that into the battery. Now we have 1,800 watts, let's say 97% efficiency on the MPPT conversion, that is typical. 1,746 watts / 28 volts = 62.36 amps. Does not seem like a huge difference in the full sun conditions, but that is still 1,512/1,746=87% with the relay vs the MPPT

And for the rest of the day, any drop in panel voltage and current directly drops the current very quickly on the switch charge controller. If the panel falls below 28 volts, you get no charge current. You may only be able to charge for 4 hours on a good day. With MPPT the voltage can fall a fair bit and the controller can still track the power curve and find the optimum power point and convert it all to the battery voltage. Even at very low light levels, it can pull good power from the panels. The area under the curve will be much higher.

Also when the battery comes close to fully charged, the switch type can only shut off, and kick back on to full current again if the voltage falls. The MPPT can smoothly roll off the current as the battery reaches full, and even keep powering the load while the battery is just kept in the CV area. This can be a huge advantage, and also much easier on the cells. You can completely run the loads off of solar while the battery waits until the sun goes down. Rather than charge, run on battery, charge, run on battery, repeat. More modern PWM controllers are between the old relay switch and an MPPT. Most of the cheap ones are no more efficient, and still just switch on and off, but there is not contact arcing in a mosfet, and they can react faster. Some of the better ones may even have a coil and give some current boost while it is dropping voltage from the solar panel, but this seems to have gone away as the cost of MPPT has come down, the coil in a PWM would wipe out the cost savings of a cheap MPPT.
 

djkarau

New Member
Are you comfortable with getting no more then 80% of the panels rated performance under any circumstances?

For a small system where efficiency is not that important, the basic switch charge controller does work, but it is never going to be the best solution. All the talk of it being reliable is one thing, but quality MPPT controllers are very reliable, and far more efficient. Lets take a sample system with 6 typical 60 cell 300 watt panels.

A typical 300 watt panel is 33 volts at 9 amps per panel. With the switch type charge control, you parallel all 6 panels, you get 54 amps of charge current, to your 28 volt bank. This actually looks pretty good. Hitting 1,512 watts of charge power under a perfect full sun solar noon. But how far do you have to run that 54 amps? What gauge wire do you need to make that work? Clicking 54 amps on and off is pretty touch on a relay contact as well.

Now wire those panels as 2 parallel banks of 3 in series. Now we have 99 volts at 18 amps. Just 18 amps can run a decent distance over just #10 wires. Any volt lost is also a much smaller percentage. The charge controller uses DC-DC conversion to regulate that into the battery. Now we have 1,800 watts, let's say 97% efficiency on the MPPT conversion, that is typical. 1,746 watts / 28 volts = 62.36 amps. Does not seem like a huge difference in the full sun conditions, but that is still 1,512/1,746=87% with the relay vs the MPPT

And for the rest of the day, any drop in panel voltage and current directly drops the current very quickly on the switch charge controller. If the panel falls below 28 volts, you get no charge current. You may only be able to charge for 4 hours on a good day. With MPPT the voltage can fall a fair bit and the controller can still track the power curve and find the optimum power point and convert it all to the battery voltage. Even at very low light levels, it can pull good power from the panels. The area under the curve will be much higher.

Also when the battery comes close to fully charged, the switch type can only shut off, and kick back on to full current again if the voltage falls. The MPPT can smoothly roll off the current as the battery reaches full, and even keep powering the load while the battery is just kept in the CV area. This can be a huge advantage, and also much easier on the cells. You can completely run the loads off of solar while the battery waits until the sun goes down. Rather than charge, run on battery, charge, run on battery, repeat. More modern PWM controllers are between the old relay switch and an MPPT. Most of the cheap ones are no more efficient, and still just switch on and off, but there is not contact arcing in a mosfet, and they can react faster. Some of the better ones may even have a coil and give some current boost while it is dropping voltage from the solar panel, but this seems to have gone away as the cost of MPPT has come down, the coil in a PWM would wipe out the cost savings of a cheap MPPT.
@snoobler Here is a video of a side by side test that seems to go against your theory of NEVER getting more than 80% with the DSSR20...in fact it is actually out performing the MPPT in some cases.

@GXMnow This video shows how the SBMS0 combined with the DSSR20 can also feed loads directly from solar and not necessitate draining the battery and then needing to re-charge once the loads are gone.

Just to be clear...I'm simply doing my research and want to decide on the best "Value" for my situation, if I need to spend more to see significant gains...I will...but if spending less with marginal looses then the "value" is obvious. The cost difference between an expensive charge controller with cheaper wiring and expensive cabling with cheaper Controllers in the end will most likely be mute. I'm a guy that will question...in a way to learn not dispute...information I'm given with no real life documented examples. The two video's I linked here simply do not, at least in the test conditions documented, support parts of what I have read.
 

djkarau

New Member
@snoobler Here is a video of a side by side test that seems to go against your theory of NEVER getting more than 80% with the DSSR20...in fact it is actually out performing the MPPT in some cases.

@GXMnow This video shows how the SBMS0 combined with the DSSR20 can also feed loads directly from solar and not necessitate draining the battery and then needing to re-charge once the loads are gone.

Just to be clear...I'm simply doing my research and want to decide on the best "Value" for my situation, if I need to spend more to see significant gains...I will...but if spending less with marginal looses then the "value" is obvious. The cost difference between an expensive charge controller with cheaper wiring and expensive cabling with cheaper Controllers in the end will most likely be mute. I'm a guy that will question...in a way to learn not dispute...information I'm given with no real life documented examples. The two video's I linked here simply do not, at least in the test conditions documented, support parts of what I have read.
 

snoobler

Solar Honey Badger
Moderator

"MPPT vs DSSR20, side by side challenge.
With my new batteries still a couple weeks away I want to expand on the brief solar charge testing I did earlie
Flat mounted panels, low winter sun, clouds and rain for 14 days while I test the relative
performance of my Epever MPPT Solar Charge Controller and my DSSR20s."

Not side-by-side except in the most literal sense - the panels were side by side. So many uncontrolled variables in horrible conditions.

At no time in his testing did he demonstrate the DSSR20 captured more than 80% the rated power of the panels, in fact, he demonstrated that in full sun, they performed notably worse than the MPPT.

He demonstrated that when in the rare full sun conditions, the MPPT substantially outperformed the DSSR20:

1617900102766.png

The magenta bars, which represents 2X panels in series on the MPPT, are substantially greater than 2X higher than the cyan bar of the single panel on the DSSR20 indicating that 2X panels produced substantially more than 2X the power of the 1 panel. Dacian even tried to explain that away because the panels got hot and reduced the performance on the DSSR20.

When conditions are bad for solar, solar sucks, regardless of charging technology. When conditions are favorable for solar, MPPT trounces everything else.

If you want the potential to get rated power out of your array, you want MPPT. If you're okay with only getting 80% of your array's rated power, then Electrodacus or PWM is fine. It's basically a $ decision. Do you want to buy more panels, or do you want to buy an MPPT?

The other day with mostly clear skies and intermittent clouds:

1617901377691.png

2930W out of 2970W rated (98.7%).

Had these been installed in a 24V configuration, my 9 24V panels in parallel would have an Imp of 8.76A * 9 = 78.84A working at a battery voltage of 26.69V.

26.69V * 78.84A = 2104W

2104/2970 = 71% rated.

Even using Isc, the best would be:

26.69V * 83.43A = 2227W

2227/2970 = 75% rated.

It's worth $500 (price difference between my MPPT and 10X DSSR20) to me to get 25-29% more power from my panels.

MPPT: 98.7% rated.
DSSR20: 71-75% rated.

I choose MPPT.

If you've determined that you're cool with only getting 800W for every 1000W installed, then that's awesome. It's about being aware of the limitations and making an informed decision.
 

GXMnow

Photon Sorcerer
I watched both videos. Even he agrees the MPPT does pull more power, but under poor conditions, the difference was pretty small. On the second video, that small shadow is actually a big deal. The reduced current from just a single cell will impact all of the current in the series chain. When he runs in parallel, the impact is much less, but the voltage going in to the MPPT is not high enough to get it in it's most efficient range. So it really is not a fair test. Putting the panels in parallel, and basically shorting them straight into the batteries does work pretty well in his case, because those panels are very well matched to his battery bank. The amount the voltage is being pulled down to match the battery is actually pretty close to the MPPT point so it makes good power. This is a specific case that does work. If you can deal with the high current on the wires, sure, do it that way.

If he would series all 3 panels, and place them where there are no shadows of leaves on several cells, the MPPT would really start to shine. For my proposed 6 DC charging panels, MPPT is the only real way to go. Sure, I could parallel 3 sets of two series connected 60 cell panels, and just use the DSSR-20 to stuff the power into my battery bank, but that would not be ideal charging. The best I could hope for would be 27 amps of charge current at 55 volts totaling 1,485 watts at solar noon. Going to my proposed 355 watt 72 cell panels, I would actually get a little less power as they are just 8.5 amps instead of 9 amps. So then it falls to 1,402 watts with larger and more powerful panels. Using the 300 watt panels, I could run 2 parallel of 3 in series on the MPPT controller. That would make more like 1,746 watts about 250 more than the DSSR-20 style. And moving up to the 355 watt panels, the MPPT can actually make over 2,000 watts in ideal sun. That is nearly 600 watts more than the same panels with the DSSR type control.

As the conditions are typically not going to be ideal, the MPPT with enough voltage overhead can maximize the output much better. In those videos, that did not show for several reasons. With a single panel, there was not enough voltage difference for the MPPT to optimize the current, and when he did run 2 panels in series, he seemed to have some shading issues. Any shadow on either panel will hurt in a series string. The same is true for a passing cloud. To be a truly fair test, both systems should be the same size array, so passing clouds are effecting them both for closer to the same time, any fixed shadows from trees should be totally avoided, and the input voltage to the MPPT has to be in the proper range to get good tracking.

Whether it was planned or by accident, his setup is pretty well optimized for the DSSR-20 setup, and not even close to optimum for the MPPT controller. Any the results still showed the MPPT doing better most of the time. The only time the DSSR setup looked better was with series panels in poor conditions with random shadows on the MPPT. If you want a more accurate test, try it with 6 panels, 2 rows of 3, out in a field with no shadows, interleave the setup, so any passing cloud will be pretty even for both systems. Parallel 3 for the DSSR, and series the other 3 for the MPPT and let that run for a week. I will bet the MPPT ends up wining by a larger margin.

Check out this video
This is a PWM vs MPPT. While in bulk charge, the PWM is basically the same as the DSSR as it is just tying the solar panel straight to the battery. SO the result is about the same. 8 amps from PWM and 11 amps from MPPT, not even close.
 

djkarau

New Member
"MPPT vs DSSR20, side by side challenge.
With my new batteries still a couple weeks away I want to expand on the brief solar charge testing I did earlie
Flat mounted panels, low winter sun, clouds and rain for 14 days while I test the relative
performance of my Epever MPPT Solar Charge Controller and my DSSR20s."

Not side-by-side except in the most literal sense - the panels were side by side. So many uncontrolled variables in horrible conditions.

At no time in his testing did he demonstrate the DSSR20 captured more than 80% the rated power of the panels, in fact, he demonstrated that in full sun, they performed notably worse than the MPPT.

He demonstrated that when in the rare full sun conditions, the MPPT substantially outperformed the DSSR20:

View attachment 44255

The magenta bars, which represents 2X panels in series on the MPPT, are substantially greater than 2X higher than the cyan bar of the single panel on the DSSR20 indicating that 2X panels produced substantially more than 2X the power of the 1 panel. Dacian even tried to explain that away because the panels got hot and reduced the performance on the DSSR20.

When conditions are bad for solar, solar sucks, regardless of charging technology. When conditions are favorable for solar, MPPT trounces everything else.

If you want the potential to get rated power out of your array, you want MPPT. If you're okay with only getting 80% of your array's rated power, then Electrodacus or PWM is fine. It's basically a $ decision. Do you want to buy more panels, or do you want to buy an MPPT?

The other day with mostly clear skies and intermittent clouds:

View attachment 44261

2930W out of 2970W rated (98.7%).

Had these been installed in a 24V configuration, my 9 24V panels in parallel would have an Imp of 8.76A * 9 = 78.84A working at a battery voltage of 26.69V.

26.69V * 78.84A = 2104W

2104/2970 = 71% rated.

Even using Isc, the best would be:

26.69V * 83.43A = 2227W

2227/2970 = 75% rated.

It's worth $500 (price difference between my MPPT and 10X DSSR20) to me to get 25-29% more power from my panels.

MPPT: 98.7% rated.
DSSR20: 71-75% rated.

I choose MPPT.

If you've determined that you're cool with only getting 800W for every 1000W installed, then that's awesome. It's about being aware of the limitations and making an informed decision.
Thanks for this info...it helps a lot...like I said...just doing my research! Looks like I will stick with my original Victron Charge Controller but will use the SBMS0 for my battery management!
 

snoobler

Solar Honey Badger
Moderator
Thanks for this info...it helps a lot...like I said...just doing my research! Looks like I will stick with my original Victron Charge Controller but will use the SBMS0 for my battery management!

The Electrodacus works great with Victron hardware in signaling low and high voltage cut-offs.

I don't mean to come across as negative on the Electrodacus. It's a pretty awesome piece of hardware. It just has limitations like everything else, and folks that are sweet on Electrodacus can be irrationally fanatical about them in spite of its shortcomings. I'm just trying to temper that perspective. :)
 

djkarau

New Member
The Electrodacus works great with Victron hardware in signaling low and high voltage cut-offs.

I don't mean to come across as negative on the Electrodacus. It's a pretty awesome piece of hardware. It just has limitations like everything else, and folks that are sweet on Electrodacus can be irrationally fanatical about them in spite of its shortcomings. I'm just trying to temper that perspective. :)
Luckily 95% or my portfolio is in Tesla not Electrodocus so I don't have a horse in the race. Like I said previously, so much of what you read on line is complete BS, but folks come across as knowledgeable. I like to see facts...no opinions. I do agree that their SBMS0 is a really cool device...basically a mini-Batrium for a fraction of the cost and complexity.
 

snoobler

Solar Honey Badger
Moderator
Don't get me started on Batrium.... my wallet is still quaking in fear of what is to come. I'm working on a parallel 14-15S NMC Li-Fire bank with a planned Batrium BMS. I'm looking at around $1200 IIRC.
 

rcrracer

New Member
For now, just remove the RV AC system, fill the hole and get a normal Mini-split. If you really want to use a roof mount, there are several options out there that use ceiling mounted "indoor units" but wold need some work to cover them on the roof. New Horizons RV have made a few 5th Wheels with ceiling mounted head units!
https://www.energystar.gov/productfinder/product/certified-room-air-conditioners/results. List sorted by efficiency. Inverter technology window ACs at top down to 13.8 ceer, top 18 on list.
 
Top