Mppt vs pwm real world testing

mountainman · July 2018

I've been wondering about the benefits of Mppt over pwm. So I did a side by side test. I put 2 100 watt panels in series 38 volts 50' 1.40% volt drop on 10 awg on a 30 amp mppt cc. Then on a pwm 30 amp cc. 2 100 watt panels in parallel 65' 19 volts 4.57% volt drop on 8 awg. On a cloudy 88f day starting at 10 am both systems charging the same roughly 65% soc 12 volt bank. Connected to rc watt meters. After 2.5 hrs pwm 210 whrs 15 amp hrs. mppt 203 whrs 14.52 amp hrs. Conclusion on a small system and in summer pwm wins.

BB. · July 2018

Yea... The are about the same "efficiency" for a properly designed MPPT and PWM systems. That is because the Vmp~35-38 volt array on a warm day is just about the "perfect voltage" for charging a 24 volt lead acid battery bank in nominal/warm conditions.

Also, many MPPT controllers really don't do MPPT "tracking" if the input voltage is too low... More or less, a minimum of 1.3x Vbatt-charging voltage is needed:

29 volt battery charging * 1.3x = ~38 volts

And remember that warm/hot solar panels, Vmp-std of 38 volts does drop and can be much close to 35-31 volts (depending on actual temperature of cells).

With very low Varray voltage, a MPPT controller does not do much different than a PWM controller. A nice working Vmp-array voltage is ~2x battery charging voltage. If your MPPT array was Vmp~72 volts (roughly), then you know the MPPT controller is functioning as designed and "Tracking" Vmp/Imp of the array.

Also note that for off grid DC power systems... The loads+battery state of charge control the actual harvested energy--Once the battery bank is full and no loads, there is no more harvest.

And when FLA batteries are over ~80-90% state of charge, the controllers go into "absorb" mode and are supplying only the charging current need by the battery bank and MPPT/PWM does not make any difference (i.e., both controllers are outputting less than maximum current that Bulk Mode charging is capable of).

So unless you controlled for loads and battery state of charge--Your test may actually not reflect the actual panel+controller harvest capabilities. Day to day weather/temperature/etc. condition variations can easily ad 10% or more error/change in harvest between otherwise "identical" weather days.

There are just a bunch of variable test conditions that have to be controlled for to get useful results here.

-Bill

mountainman · July 2018

Bill this was on a 12 volt battery with no loads. So the mppt was 2x battery charging voltage. The mppt was consistently showing higher watts amps etc. Then it would drop to 100 watts for 20 seconds or so and then back to 140. At the same time pwm was showing constant 138. Even during clouds mppt would show slightly more. But over all pwm even with more voltage drop from cc to panels had the advantage. I'm gonna try it again on a clear cool day just to see the out come.

NANOcontrol · July 2018

I think this is a very hard thing for the general public to measure and your results do not mean a lot. A battery is an indeterminate load and the pulsed nature of the power is hard to measure. In this situation, I wouldn't see much gain because the difference in voltage is so small. I think of MPPT as solar for dummies. You hook things up with no understanding and the controller figures it out. Someday, they might be called panel killers. But, going to something like grid tie panels, smaller wire and long runs has distinct advantages. I would never run MPPT in my systen, I run panels at power point voltages (MPPC) which gives you better than 90% of the advantages of MPPT. This allows me to set priorities as to where the power goes so nearly 100% of the panels capabilities at the moment are used. My charging system is based on cheap $5 buck converters and because I choose to store energy elsewhere I only need a very small battery. It's solar upside down.

mcgivor · July 2018

mountainman said:

Bill this was on a 12 volt battery with no loads. So the mppt was 2x battery charging voltage. The mppt was consistently showing higher watts amps etc. Then it would drop to 100 watts for 20 seconds or so and then back to 140. At the same time pwm was showing constant 138. Even during clouds mppt would show slightly more. But over all pwm even with more voltage drop from cc to panels had the advantage. I'm gonna try it again on a clear cool day just to see the out come.

In order to achieve more accurate results, if both controllers were connected to the same battery with a resistive load greater than both arrays combined, applied to a common output this would divert all production of both controllers to the load, the only purpose of the battery would be to power the controllers, this may be more accurate to compare outputs. Truly I appreciate you experimenting, just trying to provide information to obtain a more accurate result.

myocardia · July 2018

NANOcontrol said:

I think this is a very hard thing for the general public to measure and your results do not mean a lot. A battery is an indeterminate load and the pulsed nature of the power is hard to measure. In this situation, I wouldn't see much gain because the difference in voltage is so small. I think of MPPT as solar for dummies. You hook things up with no understanding and the controller figures it out. Someday, they might be called panel killers. But, going to something like grid tie panels, smaller wire and long runs has distinct advantages. I would never run MPPT in my systen, I run panels at power point voltages (MPPC) which gives you better than 90% of the advantages of MPPT. This allows me to set priorities as to where the power goes so nearly 100% of the panels capabilities at the moment are used. My charging system is based on cheap $5 buck converters and because I choose to store energy elsewhere I only need a very small battery. It's solar upside down.

So, where exactly are you storing your power?

jonr · July 2018

> Then it would drop to 100 watts for 20 seconds

Sounds like a poorly designed/configured MPPT controller. The only conclusion I would draw from the test is that PWM sometimes slightly beats MPPT.

mountainman · July 2018

I don't doubt that turning on loads of more amps than the panels can produce would make different results. But who does that. My test was about which mppt or pwm would charge my battery the best. I unhooked my array on tue afternoon. So Tuesday nite. wednesday and wednesday nite. My usage was 900 watt hrs or roughly 65% soc on my battery. I ran the test for 2.5 hrs during bulk with , both cc on 2 hrs absorb both cc were still showing 11 amps when i stopped the test. My thoughts are my mppt cc gets lost tracking ( 100 watts for 20 secs) a better one would give better results for mppt.

NANOcontrol · July 2018

At the heart of every MPPT system is a buck step down converter. There are various designs that have different conversion efficiencies from 70 to 95%. Much depends on the nominal voltage of the battery system and what the MPPT cost. Indeed there can be little difference between PWM with a 12V system with low array voltage. Your results are worth noting, but not a surprise. You would be surprised at how much power the average system wastes beyond what it can store. As I said before, I store cold and hot water.

mountainman · December 2019

For the last year or so I've ran 2 arrays
Both are mounted on the same stationary
Ground mount.
50 feet from the batteries and cc.
I'm using 10 awg on the
2 GT 190 watt panels IN SERIES on mppt.
53 vmp ~1.33℅
And on the 4 100 watt panels in parelell on pwm. Im using 2 10awg runs 18.9 vmp 5.83℅ loss.
10 days of Dec, ( 3 of which it rained) 40-50 f day temps
Mppt has logged 5.78 kwh
My pwm controller only logs ahs 561.
561 x~13.5 7.57 kwh?
With my small system
I'm not impressed with mppt.
Thoughts any one?

Estragon · December 2019

What's the spec vmp for the 190w panels)? 53v seems low for 2 in series.

BB. · December 2019

What voltage battery bank? 12 volt?

A simple difference could be different Vabs settings (either real like 14.4 vs 14.75 volts, or both set to 14.7 volts, but the internal volt meters are not in exact agreement).

Also have seen noise battery bus (loads, charger, desulfstor) can affect charging behavior too.

Bill

mountainman · December 2019

Yes 12 volts. 190 panels specs are 7.12 imp 26.7 vmp.
100 watts are 18.9 vmp 5.29 imp.
Checking with multimeter at the battery with just the pwm connected reading is 14.8 absorb.
but with only the mppt controller connected to see 14.8 at the battery
I have to set it to 15 volts.
Pwm default is 2 hours absorb.
Mppt set to 3 hours. I have tested all panels voc and isc
( Also have seen noise battery bus (loads, charger, desulfstor) can affect charging behavior too.)
@bb not sure what this means?

Estragon · December 2019

So with 26.7v spec vmp, 53v is probably about right.

Looking back earlier in the thread, there seemed to be a problem with mppt sweeps appearing to sort of get stuck. Is this still happening?

BB. · December 2019

Sounds like 14.8 volts for both controllers? 15.0 volts is a bit high for absorb (unless cold weather/thermal compensation is kicking in).

The "noise issue" was seen a few years ago with (as I recall) different generations of Outback MPPT controllers, and a desulfator... Two arrays+two controllers to one battery bank... One controller was constantly producing less AH/WH per day vs the other with the desulfator running. With the desulfator disconnected, then both controllers output the expected amount of charging AH/WH.

Desulfators, make very short current spikes on the battery bus... These short current "impulses" are created to "ring" the hardened black lead sulphate crystals, and get the sulfate crystals to go back into solution (sulphur into solution, lead back onto plate)--At least that is the theory (roughly) as I understand.

So, if you have an "electrically noisy" battery bus (from desulfator, from other loads, or even, possibly a PWM charge controller and MPPT controller connected to the same battery bus).

Is this what is happening to your system--I really do not have a clue...

One way to help reduce issues with noise... Each controller should be "home run" from the controller Vbatt output to the Battery Bus...

If you share wiring (i.e., from battery bus to Controller A to Controller

--Aka Daisy Chaining the connections--This can also create "cross talk issues" between controllers...

It is possible that the cross talk/bus noise causes one controller to throttle back sooner than the other controller (trips the absorb set point sooner). It is also possible that each controller "sees" different battery bus voltage (calibration, voltage drop vs current on charger to battery connections, different thermal compensation, etc.)...

-Bill

mountainman · December 2019

As long as there is full sun the mppt works ok. Over cast days not so good it seems to get lost. Ive tried 2 different pwm cc. Before I rule out mppt I'll try a better controller.

Photowhit · December 2019

It's pretty much an impossible test.

To do it properly, the battery banks would have to never reach full and have the same voltage/resistance all the time. Then measure the total amps delivered for the 2 systems. If the systems are in use their loads will vary, if hooked up to the same battery bank they can influence each other. In the end the information you receive isn't viable for off grid systems, because they aren't used like that...

It's just very tricky to evaluate.

mountainman · December 2019

As @mcgivor and @Photowhit suggest
For more accurate results. As an experiment for a day I will add a purely resistive load larger than both arrays can produce.

mountainman · February 2020

@Photowhit I got real mppt controller Morningstar 30. After a couple rain days . while bulk charging ~13.5 volts. On cool 35f days my Pwm system ~26 amps and mppt 24. This is enough for me to say on a small 400~800 watt system mppt is not worth the price.
Mppt has its advantages (distant arrays)
But for less than ~50 foot away
Adding more panels even at 75 cents a watt.
with free shipping.
and a quality adjustable pwm controller.
you'll be money ahead. Compared to mppt cost and grid tie panels with their ridiculous shipping prices.

littleharbor2 · February 2020

What voltage are you putting into the MPPT controller? Are you charging the same battery bank with the same exact array, at the same time?

Photowhit · February 2020

mountainman said:

@Photowhit I got real mppt controller Morningstar 30.
you'll be money ahead. Compared to mppt cost and grid tie panels with their ridiculous shipping prices.

Sorry, your test is only for your self...
There are far too many variables to do a meaningful test in this manner.
Even 2 systems complete with dump loads and outside metering with the same meters would be meaningless as you could have a single poor unit.

FWIW- I am a fan of PWM CC's. I've used them more than MPPT type to date.

mountainman · February 2020

@littleharbor2. I've tried 27 volts and 54 array volts on mppt with 2 190 watt 60 cell gt panels. 380 watts.
18.9 volts with 4 100 watt panels in parellel on pwm 20 more watts than mppt.
Both charging the same battery.
For example it was 95℅ overcast This morning at 10.
13.5 volts Pwm showing 2.58 amps
Mppt .8 amps.
@ photowhit in my opinion with 400 watts and pwm adding 25℅ more panel wattage.
Is cheaper and will produce more wh per day than
A 400 watt system with mppt.
And for much less money.
For a year no matter what the season
Or weather. I have yet to see mppt producing more watts or amps than pwm.

bill von novak · February 2020

mountainman said:

@ photowhit in my opinion with 400 watts and pwm adding 25℅ more panel wattage.
Is cheaper and will produce more wh per day than
A 400 watt system with mppt.
And for much less money.
For a year no matter what the season

I have no doubt that you saw that.

For most installations, a good MPPT controller will harvest more energy than a good PWM controller.

littleharbor2 · February 2020

mountainman said:

@littleharbor2. I've tried 27 volts and 54 array volts on mppt with 2 190 watt 60 cell gt panels. 380 watts.
18.9 volts with 4 100 watt panels in parellel on pwm 20 more watts than mppt.
Both charging the same battery.
For example it was 95℅ overcast This morning at 10.
13.5 volts Pwm showing 2.58 amps
Mppt .8 amps.

Your test really needs to be done when the batteries are 75% SOC. or less, and on a sunny day. Are your panels oriented at the same angle? One thing I noticed is that the 190 watt panels are Evergreen string ribbon cell multicrystalline technology. I'm pretty sure the Renogy 100 watt panels are monocrystalline cells. Mono cells are supposed to work a bit better in cloudy conditions and the efficiency of the newer Renogy cells is higher than tho older Evergreen cells.

mountainman · February 2020

You are exactly right the renogy are mono and the evergreen are poly.
Last year I tested 2 100 watt mono panels on pwm and 2 on mppt. Overcast days where about equal. But in full summer sun
after 2 overcast days ~60℅ soc battery.
Pwm put out more daily wh.
on a 500 watt pwm system.
.75 cents a watt for 5 new 100 watt mono panels free shipping. $375 +175 for a pwm 45 morningstar. $550 total.
On a 500 watt mppt system ~40 cents per watt for grid tie $200 + shipping $350= $550.
Plus $475 for a m star 45 mppt . $1025.00
Personally I'd rather have 2 separate 500 watt pwm systems. Than 500 on mppt.
For almost the same money

littleharbor2 · February 2020

MPPT controllers are less effective on small systems and in warmer locations. There's no doubt that a real MPPT controller will harvest more power from a solar array given it is configured correctly and has high enough overhead voltage for it to do it's thing.

Are you, in fact using a Morningstar MPPT controller?

mike95490 · February 2020

And be sure you allow the MPPT controller to boot up off the battery before connecting the solar to the solar inputs. Connecting the high voltage solar without the battery having been connected, can/will instantly fry the controller

Mppt vs pwm real world testing

Comments

Categories