MPPT Controller Explanation For Dummies

[Maine Sail]

Originally Posted by goboatingnow

Yes but both types connect the panel to the battery directly. , that moves Vo to the Vbat , which could be away from Vmp , with a big enough mismatch , you could loose 50% or more of the panels power with these controllers.

I mean I've seen them advertised with 30-40 V operating input voltage ranges. , unsuspecting citizen joe , connect up a 25 V panel , ( with a Vmp of say 23v) , every time the PWM switch connects the panel in , its Vo is pulled miles away from Vmp, resulting in huge looses in power generation.


Unless the panel is so small that you don't care about loosing 30-50% of say 20W , or your lucky in sourcing a panel with a Vmp close to your charging point , not doing DC DC conversion on solar , is throwing away power generation.

Dave

Agree 100%!

PWM though is significantly faster at bringing a bank back to 98-100% SOC than a shunter is but arguably less "efficient" than a good well engineered MPPT. In my own testing I've not yet found a Genasun MPPT (the OP has a Genasun) beat by a PWM (some argue this can happen) and I was specifically looking at early am charging and late afternoon.

In my experience a lot of the low cost MPPT's are pretty much worth the box they ship in (usually my customers find them on eBay). I have had some "MPPT" controllers labeled as such that were nothing more than a cheap shunter with an MPPT sticker on it. Many eBay solar buyers are first timers and would not know the difference, so the scammers get away with it.

I've also seen some inexpensive MPPT controllers that spend more time "off" searching for the max Vmp than they do charging especially with the ever changing shade characteristics on a sailboat which I can only assume "confuses" them when searching for the Vmp.

The better MPPT controllers I've used use a P&O algorithm or a modified version of P&O and they don't disconnect from the array to check Vmp..

I have seen some MPPT controllers that test Vmp and take 2-3 seconds and some that can take 30 seconds before firing back up again. I find the controllers using a modified P&O method tend to produce current both earlier and later in the day as well.

The best "low cost" MPPT's I have found are the Genasun's and I have played with piles of them both on boats in the real world and in my shop or via a solar array set up in my yard and fed into my shop.. The Genasun controllers run very cool and do what they are supposed to do. They are also very quiet noise wise..They are also built & designed to survive in the marine environment something controllers costing a lot more don't often do well with. At first I was skeptical of Genasun's concept of one controller per panel but after doing my own testing, both ways, I have seen what they see and now understand why they advise one controller per panel...

I have converted many boats from cheap Sunforce or FlexCharge shunters to MPPT and even some Morningstar PWM's and seen the average time to "full" for the bank be reduced by days not just hours.

[T1 Terry]

As I said in my post Dave, MPPT works if there is a big mismatch between panel Vmpp and battery terminal voltage, but most nom. 12v panels have a Vmpp @25degC between 16.5v and 18v. Unless you live in the snow or have ice cold winds, the panel will rapidly heat up well above 25degC, Outback Aust 75degC is not at all unusual by 10 am in the morning. Maybe on a power boat you would get good air movement across the panels, but not really on a sail boat, the panels will heat up quite a bit. Around lunch time on a good sunny day, slip your hand in behind a solar panel, don't actually touch it, no point in burning yourself to prove a point, you will feel the radiate heat well enough just putting your hand close. For every deg over 25 degC you have to drop the Vmpp by 0.462%, there isn't much leeway between the now reduced Vmpp and the steady 13.3v to 13.6v of a charging Li battery. Now factor in the circuitry that needs to be powered in a quality MPPT controller, the heat generated is lost watts and the fan also eats power to dissipate that heat, the PWM controller feeds it through a mosfet, that's it, very minimal power loss, there is a valid reason why MPPT controllers are big and have big heat sinks yet PWM controllers only have small heat sinks.
If you put a big load on a lead acid battery and drag it's terminal voltage down into the 11v or even high 10v area, the MPPT will be way ahead, but that doesn't happen with Li batteries, a big load doesn't pull the voltage down significantly, it stays in the 13v area.

There are other issues with MPPT controllers and Li batteries as well but hopefully some manufacturers will address these problems soon, one is the point they return to boost, often for lead acid it's around the 12.6v mark, dead flat for an Li battery. Another problem is the length of the absorption cycle, often a percentage of the time it takes to reach the boost voltage point when the switch to absorption stage is made, works fine for lead acid, Li is all boost till it's full, holding that voltage for a few hrs of an absorption cycle will invite a cell runaway, 14v to end of boost, then 13.8v or even 13.7v and an Li battery is happy, most up market MPPT controllers don't give you that option, hopefully that will be addressed soon.
Maybe an Outback MX80 or the other one, midnight something would show some gains, but I doubt it would be value for money gains, not with Li batteries anyway, I'd prefer to spend those $$ on good battery monitoring at cell level and good cabling, more gains to be made there.
As for cheap MPPT controllers, anything under $600 and you are throwing good money away, and for that money you can buy a good PWM controller and battery monitoring, I just can't see the value for the possible small gain you might get from a top line MPPT controller.
Now, if you have a solar array greater than 1.5kW, then yes, the value would be there, but you would need a big Li battery bank to put all that in, remember they don’t need an absorption cycle, so all the hrs of sunlight are full rate charging hrs.

T1 Terry

[goboatingnow]

Originally Posted by Maine Sail

Agree 100%!

PWM though is significantly faster at bringing a bank back to 98-100% SOC than a shunter is but arguably less "efficient" than a good well engineered MPPT. In my own testing I've not yet found a Genasun MPPT (the OP has a Genasun) beat by a PWM (some argue this can happen) and I was specifically looking at early am charging and late afternoon.


.......lots of good stuff .............

I have converted many boats from cheap Sunforce or FlexCharge shunters to MPPT and even some Morningstar PWM's and seen the average time to "full" for the bank be reduced by days not just hours.

Why shunt solar panels , to turn them off, it just wastes power in the shunt ( unless the shunt is useful of itself )

Still doesnt get around the fact that mismatched Vmp and PWM are a very bad combination, you can loose stacks of power generation.

Dave

[noelex 77]

Dave you are taking about using PWM controllers on high voltage panels. I agree this is totally inappropriate, but talk of the 30-50% gain which could occur in these circumstances is only confusing the argument, and those looking for guidance when selecting controllers.


There has been is a range of opinions expressed about the gain available from MPPT regulators from 10-15% (me) to 25% (Jedi). There have been very few controlled experiments on boat systems so these sort of difference are not surprising, but I think it would be better if we talk about the results when the controllers are teamed with suitable panels.

PWM controllers do need to used with 12v (nominal) panels these will have a Voc around 22v and a Vmp of around 18v. Most small (less than 140w) panels fall into this category.

MPPT controllers can be used with these panels, or higher voltage panels which are more common as you get individual panels above 200w. (Be careful to leave a comfortable margin below the controllers maximum voltage, but with the better controllers this is often very high at 150v)

[goboatingnow]

Originally Posted by T1 Terry

As I said in my post Dave, MPPT works if there is a big mismatch between panel Vmpp and battery terminal voltage, but most nom. 12v panels have a Vmpp @25degC between 16.5v and 18v. Unless you live in the snow or have ice cold winds, the panel will rapidly heat up well above 25degC, Outback Aust 75degC is not at all unusual by 10 am in the morning. Maybe on a power boat on good battery monitoring at cell level and good cabling, more gains to be made there.
As for cheap MPPT controllers, anything under $600 and you are throwing good money away, and for that money you can buy a good PWM controller and battery monitoring, I just can't see the value for the possible small gain you might get from a top line MPPT controller.


.......loads of good stuff here ......


Now, if you have a solar array greater than 1.5kW, then yes, the value would be there, but you would need a big Li battery bank to put all that in, remember they don’t need an absorption cycle, so all the hrs of sunlight are full rate charging hrs.

T1 Terry

Yes , BUT , if you say have 400 W of panel capacity , at a Vmp of 18 ( , ie it is that ) then at 13.5 , your are throwing away 90 W , nearly a quarter of the panels capacity.

Now , your in the outback , but take a boat with shading , poor incident angles , while Vmp doesn't move a lot with illumination, it does move , let's say worse case it moves below the operating point , now the panel output tanks

Dave.

[goboatingnow]

Originally Posted by noelex 77

Dave you are taking about using PWM controllers on high voltage panels. I agree this is totally inappropriate, but talk of the 30-50% gain which could occur in these circumstances is only confusing the argument, and those looking for guidance when selecting controllers.

There has been is a range of opinions expressed about the gain available from MPPT regulators from 10-15% (me) to 25% (Jedi). There have been very few controlled experiments on boat systems so these sort of difference are not surprising, but I think it would be better if we talk about the results when the controllers are teamed with suitable panels.

PWM controllers do need to used with 12v (nominal) panels these will have a Voc around 22v and a Vmp of around 18v. Most small (less than 140w) panels fall into this category.

MPPT controllers can be used with these panels, or higher voltage panels which are more common as you get individual panels above 200w. (Be careful to leave a comfortable margin below the controllers maximum voltage, but with the better controllers this is often very high at 150v)

Sorry Noelex , the tests that have been done are often very faulty

Using your own example a Vmp of 18 in a 12 operating point is loosing over 20 % of available energy. Even simple voltage conversion with a manually set Vo ( operating) improves things , never mind tracking the power point.

20% is 20% ( you'd be happy to get it in gas mileage !!) or boat speed on a best !

When I mean high voltage I'm only talking about pulling Vo a little ( 3-5v) away from the Vmp. I mean if you have hot panels , you could potentially end up with Vo> Vmp , and that's where the output of the panel collapses.

The chip sets are out there and power tracking is common place in industrial applications , from pay machines to lighting , only the marine industry continues its merry process of overcharging you for its products.

I dont see how PWM can help absorption , charge any faster. Its just the same as connecting the panel and waiting around with a multimeter .

Dave

[noelex 77]

A lot of atypical numbers are being used for Vmp.
13.3-13.6v from one poster and 18v from another. These sort of Vmp,s are possible and will be seen from most panels briefly, but for the majority of the time a much more realistic Vmp will be about 16.5v. (From a nominal 12v panel.)
FWIW just checked mine and the Vmp is 16.6v

[noelex 77]

Originally Posted by goboatingnow

Using your own example a Vmp of 18 in a 12 operating point is loosing over 20 % of available energy. Even simple voltage conversion with a manually set Vo ( operating) improves things , never mind tracking the power point.

Vmp for most 12v panels is listed around 18v.
However this is measured at STC. This includes a cell temperature of 25C.
In fact the only way they can get this is with very brief flashes of light. Any normal light exposure will heat the black panels up considerably.
We are fortunate on boats to have a good airflow typically and the cell temperature near the peak production of about 45C is typical.
This lowers the Vmp to about 16.5v.

FWIW just checked my cell temperature with a IR thermometer 41C with an air temperature of 24C, but it is blowing 30knots at anchor which is helping the cooling.

[goboatingnow]

fine Vmp 16.6 say on a 100W panel , heavily discharged battery , ( just when you want all the energy) , say 12v, thats 27 watts lost in the PWM, ( ie a quarter of the panel capacity).!

Dave

[endoftheroad]

Originally Posted by FlyingCloud1937

Endoftheroad,

I don't want to burst your bubble,

but food for thought.

Solar panels can easily fool a volt meter, especially if you never plug-in to shore power.

It's called surface charge.


If you have a 440 amp-hr bank and draw out 100 amp-hr over-night.

Now add 2 solar panels that generate 12 amp's per hour. Total by 12 noon in your latitude is going to be not more then 48 amp hrs added back...leaving a deficit of about 52 amp hrs.

If your bats are showing 14+ voltage you have a surface charge. And the real state of charge is....something less than 100%

Amps out overnight X 125%, plus amps in from the solar = net re-charge.

So now if this is a day in day out occurrence and you have a negative, then, at some point the deficit is going to come home to roost.

Lloyd

I'm unsure what surface charge means.
During the day for about an 8 hour period my LED indicator on my controllers are reporting a "full charge" for much of the day while running a fan, vhf and refrig. Naturally the controllers fall back into charging mode briefly from time to time as I'm drawing a load from my batteries.

Are you saying that my batteries aren't truely full?

When my voltmeter showed my batts at 14+ could this have been while the controllers were charging the batts?

[noelex 77]

For those interested in the theoretical calculations of the gain with MPPT, this is my estimation.
If we have a Vmp of 16.5v and a average battery voltage during charging of 14v there is a theoretical gain of:
About 18%.

From this we need to deduct
1. The self consumption of the circuits to track the MPP.
2. The fact the current will increase slightly as we lower the voltage
3. The tracking is never perfect
4. Multiple panels connected to one controller will have different VMp

Note the above is for the peak of the day. Shadows or late in day will reduce the Vmp and point one also becomes more important.

My 5-15% estimation from observing my system with and without a MPPT controller seems to at least fit, ball park, with the theoretical calculations.

[goboatingnow]

hang on,

a battery voltage of 14v even with surface charge, you are likely not pulling enough power out of the panels to make use of power point tracking anyway.

rerun the computations with a low battery !

dave

[noelex 77]

Originally Posted by goboatingnow

fine Vmp 16.6 say on a 100W panel , heavily discharged battery , ( just when you want all the energy) , say 12v, thats 27 watts lost in the PWM, ( ie a quarter of the panel capacity).!

Dave

Yes it is, that is why MPPT manufacturers can get away with listing "up to 25%" improvement. If they factor in subzero arctic temperatures (where the Vmp is higher) they can even claim 30%.
However battery voltages of 12.0v under charge are not typical and will only be seen for a small fraction of the charge cycle, or with a heavy load.

To get realistic average results you to use realistic average voltages. I don't believe 12v is average, but the the calculations are easy to do so feel free to plug in whatever numbers you feel are appropriate.

[colemj]

Originally Posted by T1 Terry

[COLOR=black][FONT=Verdana]

......

12v panels have a Vmpp @25degC between 16.5v and 18v. Unless you live in the snow or have ice cold winds, the panel will rapidly heat up well above 25degC, Outback Aust 75degC is not at all unusual by 10 am in the morning. Maybe on a power boat you would get good air movement across the panels, but not really on a sail boat, the panels will heat up quite a bit. Around lunch time on a good sunny day, slip your hand in behind a solar panel, don't actually touch it, no point in burning yourself to prove a point, you will feel the radiate heat well enough just putting your hand close. For every deg over 25 degC you have to drop the Vmpp by 0.462%, there isn't much leeway between the now reduced Vmpp and the steady 13.3v to 13.6v of a charging Li battery. Now factor in the circuitry that needs to be powered in a quality MPPT controller, the heat generated is lost watts and the fan also eats power to dissipate that heat, the PWM controller feeds it through a mosfet, that's it, very minimal power loss, there is a valid reason why MPPT controllers are big and have big heat sinks yet PWM controllers only have small heat sinks.

We have a lot of experience using a Trace 40 PWM and Morninstar 60 MPPT on the same panel setup.

The Trace and Morningstar both have roughly the same heatsink area, but the Trace heatsink was always significantly hotter than the Morninstar's. Even accounting for the possibility that the MS heatsink is larger or more efficient, the Trace was very hot comparatively.

Neither controller has a fan.

Our 17.5V OC panels near the equator during full sun get so hot you can fry an egg on them. The MPPT seems to optimize at ~15V at these temps.

When it is necessary to control the current in the absorption phase, our MPPT switches into PWM mode (it notifies you which mode it is in). So it seems that MPPT is unlikely to be useful if your system spends much of a day in absorption mode, regardless of how efficiently you reached that point.

We use enough power to ensure that we rarely reach the point where the MPPT goes into PWM (this statement is to set up the one below it).

Near the equator with nearly normal incident sun and perfectly symmetrical days, we experience very little measurable difference in average daily output between the two controllers. However, when we switched the 4 panels into a series/parallel configuration to provide a ~35V OC to the MPPT, we experienced a 10-15% gain in daily output. I attribute this to taking better advantage of the lower incident light of early morning and late afternoon, where the individual panel voltage is too low for PWM.

Mark

[goboatingnow]

where the individual panel voltage is too low for PWM.

Panel voltage doesnt change much based on illumination , current capacity does. What you are seeing is the benefits of moving the Vo closer to Vmp ( any improvement helps). thats the advantage of voltage conversion, never mind the Mpp tracking part.

Dave