PWM versus MPPT- a quick analysis

lazza · August 2013

HI Forum

We've been looking at some bigger systems and the possibility of using MPPT has cropped up. The truth is that the simplicity and price of ordinary xantrex C40 type charge controllers has meant we havent ever worked with MPPT, but want to make sure that my cognitive selection process is correct. Hence I've devised a table summary. Any suggestions and additions or corrections would be most welcome... thanks

Larry

Attachment not found.

Cariboocoot · August 2013

Re: PWM versus MPPT- a quick analysis

Well, pretty much so.

Some of the points may be exaggerated though.
For instance the higher Voltage array on an MPPT controller needn't be too much higher than system Voltage and it is better not to be. On the other hand 30% greater charging is unlikely to occur in most installations as it depends on conditions which usually are not present. MPPT controllers have proven to be as reliable as PWM; there's more electronic components inside, true, but they are all as dependable as any.

MPPT's biggest advantage is its flexibility in array design: you can use just about any panel out there, including the "GT" panels which are usually cheaper per Watt. You need to consider array & controller as a whole when judging costs. With larger systems the MPPT will usually always work out to be cheaper.

Less current from the array means less V-drop and power loss too; important when distances between PV and controller start to lengthen. Not much more danger in being shocked from a 90 Volt array than a 70 Volt one (48 Volt system).

And when you want the system to do a bit more than just charge batteries, well there are no PWM controllers that offer the auxiliary functions found on an Outback FM or MidNite Classic for example.

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

Quite biased towards PWM. PWMs are only good for smaller system (no more than 2kW solar) where higher cost of MPPT controller cannot be justifyed by savings.

I don't think MPPT controller are much less durable than PWM ones. Durability depends more on specific controller.

Breakers/fuses are rated for 150VDC are cheap and commonly used for any voltage (at least here).

90-100V operating voltage doesn't pose significant threat to safety (compare to 230V AC).

There's no such thing as efficiency during absorption/float. At these stages you refuse energy that could've been produced.

If the system is in absorption/float on sunny days, it is probably designed for cloudy days. When a cloudy day comes, MPPT's efficiency will still help.

Higher voltage also means thinner wires. In bigger systems wire savings will pay for MPPT.

zoneblue · August 2013

Re: PWM versus MPPT- a quick analysis

Well, i thought like you in the start. I remember telling our local installer guy that i wanted to use a TS60 for a 1.8kW system, and he actually laughed.

niel · August 2013

Re: PWM versus MPPT- a quick analysis

i can somewhat understand why because when you have that much in pv that producing up to 30% extra becomes significant. most times it is more around 10%, but can go much higher. the 10% extra would mean you needed 1800w x .9 = 1620w in pv to produce the same power or 180w less in pvs. pv cost is about $200+ and is close to equilibrium in costs in theory by applying that $200+ toward the mppt cc, but it can pump out another 20% at times blowing the equality thing out of the water. if you have it in absorb and float most of the time you may have an argument there as low bulk times would throw the advantage more to the pwm. most of us will use the batteries at a more significant amount unless it is primarily a backups arrangement. add to that smaller wires and less breakers or combining space in a box significantly reduces costs there too where applicable.

inetdog · August 2013

Re: PWM versus MPPT- a quick analysis

NorthGuy wrote: »

There's no such thing as efficiency during absorption/float. At these stages you refuse energy that could've been produced.

One exception to that which can extend the stage(s) over which MPPT gives an advantage:
If you have daytime opportunity loads which you can schedule to happen when the batteries are in Absorb or Float, you can potentially make use of the full CC output to run your inverter, even though the batteries themselves are staying at low current.
Just a thought.

SolarPowered · August 2013

Re: PWM versus MPPT- a quick analysis

I prefer PWM 10amp/12 amp mosfet digital controllers for smaller high efficiency (16.5% conversion) panels under 100watts. Personally my cut off per parallel string of panels is at 12amps with 10AWG USE-2 (PER) individual controller(s). I am not all for PWM 15 amp & higher controllers due to wire sizing to connecting buss issues, it just seems like losses after that point, when in parallel and limited to a connection that cannot exceed 10AWG USE-2.
That means either (6) 20 watt panels in parallel per 12 amp PWM controller (5) 40 watt panels in parallel per 12amp PWM controller (4) 60 watt panels in parallel per 12 amp PWM controller, or (3) 80 watt panels per 12 amp PWM controller, (2) 100 watt panels per 12 amp controller. Sure lets throw a (1) 120 watt panel in there just in case it follows a VOC range of (20~22)
My thought based off of experience is most panels on the market after 120watts exceed 22voc (18~18.5vmp), or the amperage is just to high that there would be losses exceeding what 10AWG USE-2 is meant for, and then we are getting to the point of fuses.
PWM comes much cheaper than MPPT by 35%~50% in cost.
Overall design with panels under 120watts ( with a voc of 20~22) actually has a cost savings and a much better return on investment VS. MPPT. Parrallel amperage is not affected in the same manner that series voltage is and then converting voltage to amperage in maximum power point.

lazza · August 2013

Re: PWM versus MPPT- a quick analysis

Ok thanks everyone, great help.

I guess I am quite PWM biased because we generally work under 2kW of installed power where only 1 charge controller is necessary, and working in southern spain with 40% unemployment, price is everything.

However, it seems that once you get over a certain threshold the advantages of MPPT start to have more weight and i think we must seriously consider this option.

For example, we are looking at installing an 8kW PV system. Now this would require 4 Xantrex C60 (one for each set of 10 190W 24Vn Luxor panels in parallel).. which would come close to the costs of MPPT equivalent. Then there are the advantages of less wiring, increased efficiency at periods of high consumption/low light,

I'm also not quite sure how these 4 Xantrex PWM would work in unison ??

What MPPT system would the forum recommend for such a sized system?

Photowhit · August 2013

Re: PWM versus MPPT- a quick analysis

Well I own 2 in use MidNite Classics and have 1 that, I hope will be installed this fall, so I'm not against MPPT charge controllers. If I had the chance to purchase large 24v nominal panels for my new system in the price range of my current array (.80 a watt) I might have gone with PWM charge controllers, One of the things that hasn't been mentioned is that MPPT charge controllers can now communicate with each other and will be able to charge measuring the current flowing across the shunt for better charging control!

On others replies;

I'll consider MPPT charge controller 'proven' when they have lasted 20 years, I have a 30 amp Specialty Concepts Mark IV that was in service over 20 years and is still in my 'spares' closet. MPPT CC run hotter, time will tell if they will survive.

It's hard to find panels in proper voltages to make use of PWM CC's any more, but strings of 3 'grid tied' panels can often be found that work for 48 volt systems, and might well be a cost effective way of doing 48 volt system paired with a Morningstar TS60, but with Midnite Classic lite going for $500 it's hard to justify setting up a PWM system these days.

In a well designed system, the 30% advertised is a pipe dream and good stuff at that! And more than 30% is silly talk unless your referring to poorly designed systems with Grid tie type panels with much higher vmp ratings. But! MPPT charge controller do shine when you often need them! cloudy days when panel temps aren't high and your not likely to get out of 'bulk' they are producing perhaps 20%-25% more current than PWM CCs. They also shine in winter months when some systems are worked the hardest.

An 8kW PV system, should be setup to charge 48v battery bank! So C60's won't work! 2 Morningstar TS 60 would work, but so would 2 Midnite Classics, Likely a better choice for $600 more for Classic Lites or $800 more for regular Classics.

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

lazza wrote: »

For example, we are looking at installing an 8kW PV system. Now this would require 4 Xantrex C60 (one for each set of 10 190W 24Vn Luxor panels in parallel).. which would come close to the costs of MPPT equivalent.

It's hard to work out the exact numbers, but instead of 8kW with PWM, you could install 6.5-7kW with MPPT, and it will produce the same. It immediately saves you at least 1000 EUR and pays for MPPTs and then some, plus savings on wiring, combiner boxes etc. In this case, using MPPT will save you quite a bit of money. As you said, price is everything.

Photowhit · August 2013

Re: PWM versus MPPT- a quick analysis

NorthGuy wrote: »

..., but instead of 8kW with PWM, you could install 6.5-7kW with MPPT, and it will produce the same....

I can not use the language needed to describe how much I disagree with this statement!!!!!!!!!!!!!!!!!!!!

It is WRONG! A well designed system should rarely be in bulk, other than cloudy days, so on those days when it is just 'ramping up' in bulk it saves you nearly nothing! That should be most days unless your in a very poor climate. I posted a more complete answer in how they will work differently which is 'more' correct.

I currently have both types of systems setup a 1700Kw array at my cabin that I lived in and a 4Kw array at my new home. Both spend little time in bulk unless we've had poor weather. I've run and air conditioner on both systems for parts of the day for the last week, ran a hot water heater for an hour 3 mornings as well. It's been pretty nice and both systems have spent minimal time in bulk.

If panels are available in proper charging voltages, which is becoming rare, it would not be a huge mistake to do a good sized system with a PWM charge controller, other than the advantages of them being able to turn on loads at different stages of battery charging and properly charge batteries while loads are present, and recovery and charging during poor weather conditions.

Cariboocoot · August 2013

Re: PWM versus MPPT- a quick analysis

The somewhat better efficiency of an MPPT controller will allow some smaller array size for a given application. For example:

If you need 20 Amps @12 using a PWM controller you multiply current by the typical 17.5 Vmp and get 350 Watts minimum panel.

Same process for an MPPT looks like this: 20 * 12 / 0.77 = 312 Watts minimum. 38 Watts, or roughly 10% difference. Hmm. Where have we seen that before?

You definitely would not want to drop an 8kW array down to 7kW or less and expect the controller to make up the difference: it won't happen.

People who try to live on the edge usually fall off it.

Photowhit · August 2013

Re: PWM versus MPPT- a quick analysis

All very true, but once you reach absorb, usually the last 15-20% of your charging, your charge controller reduces the amount of amperage 9or actually the amount of amperage the battery takes) is reduced from the max, so even the little advantage is reduced quite quickly to zero in the later stages of charging. If your system isn't balance and you regularly need to charge your batteries from 50%DOD then there is a greater advantage, but IMHO a unbalanced system requiring a generator for charging in poor conditions.

I know you run a similar system, and I'm NOT judging you, 'Coot old buddy old pal... but most recommendations I've seen are to use the upper 20-30% of your batteries capacity.

Cariboocoot · August 2013

Re: PWM versus MPPT- a quick analysis

I'm not disagreeing with you Photowhit, rather the opposite; I'm pointing out that the charging advantage to MPPT is, under most circumstances, small and only found when maximum current is needed. The claims of "30% more power" are always "up to" and yes, when it's -40 outside and the panels are super-conducting you can get a lot more charging power through MPPT - provided it is also needed.

I'm not sure people always grasp the concept of "minimum" either; they tend to think it means "all that you really need" when it fact it means "the least you can get away with provided everything else goes right". After all, the sun does not shine brightly every day.

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

Photowhit wrote: »

I can not use the language needed to describe how much I disagree with this statement!!!!!!!!!!!!!!!!!!!!.

Photowhit,

You're arguing the wrong point. It doesn't really matter whether your batteries are in bulk, in absorb, or doing something else. Panels and controllers are about producing energy. It is not about using it or wasting it for absorption.

Say, you have 4kW array with MPPT controllers now. You can do two modifications.

- You can take off 500W and make it into 3.5kW array

or

- You can replace your controllers with PWM.

The result will be the same regardless of your use pattern.

Of course it is not straitforward, and MPPT controllers have more advantage in cold wether and less in hot days, and so on ... but do help to produce more energy and do allow for smaller array to achive the same result.

Lefty Wright · August 2013

Re: PWM versus MPPT- a quick analysis

How do PWM and MPPT compare if there is periodic shading of parts of one or more panels?

This is my situation and for that reason I've chosen PWM for my next revision. Am I wrong about that?

I'm now running MPPT but am installing more panel ampacity and planning to go PWM as I'm not impressed with the performance of my present system.

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

Lefty Wright wrote: »

How do PWM and MPPT compare if there is periodic shading of parts of one or more panels?

Both can charge batteries only if the panel voltage is higher than the battery voltage.

If you have panel voltage at 3x of battery voltage, then even if you lose 2/3 of your voltage because of shading, you still can get the rest. You can do that with MPPT, but not with PWM.

If you have panel voltage at 120% of battery voltage, even a small voltage loss will be devastating. This can happen to both MPPT and PWM.

Cariboocoot · August 2013

Re: PWM versus MPPT- a quick analysis

PV's are a current source. Shaded panels quite often have enough light on them to produce enough Voltage to charge; it is the current that drops off drastically. Putting four panels in series for higher Voltage on an MPPT controller has no discernible advantage over having all four panels in parallel on a PWM. And yes, I've actually tried this variation in the real world so don't bother trying to argue theory with me.

All other factors being equal there is no difference in shaded array performance between the two controllers.

vtmaps · August 2013

Re: PWM versus MPPT- a quick analysis

Lefty Wright wrote: »

How do PWM and MPPT compare if there is periodic shading of parts of one or more panels?

This is my situation and for that reason I've chosen PWM for my next revision. Am I wrong about that?

Possibly. A lot depends on the shading AND the configuration of the shaded panels. Any shading on a panel will be devastating to the output of the entire string that the panel is part of. It will not affect the other unshaded strings that are in parallel with the shaded string. Try to make sure that all of the shaded panels are in the same string.

As far as PWM vs MPPT is concerned, I don't think it makes much difference. Some folks have said that an MPPT controller might be fooled into finding the wrong power point when the strings are unevenly shaded, but others have said that it makes little or no difference. I wouldn't make the MPPT vs PWM decision based on shading of some panels.

--vtMaps

EDIT: Cariboocoot beat me to it. He wrote:

Cariboocoot wrote:

Putting four panels in series for higher Voltage on an MPPT controller has no discernible advantage over having all four panels in parallel on a PWM.

That is generally true. But Lefty Wright has shading on one or more panels. If his choice were 4 in series with MPPT or 4 in parallel with PWM, he would be better off with the PWM.

Cariboocoot · August 2013

Re: PWM versus MPPT- a quick analysis

Excellent point vtMaps: "shading" is a terribly vague term which can describe a wide variety of situations from a single shadow line crossing one cell segment to heavy cloud cover blocking the whole array!

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

Cariboocoot wrote: »

PV's are a current source. Shaded panels quite often have enough light on them to produce enough Voltage to charge; it is the current that drops off drastically.

You're talking about single cell.

Usually, the panel is a collection of cells connected in series. They also have bypass diodes, which bypass groups of the cells. For example, each of mine panels consists of 60 panels connected in series. There are three strings of cells. Each string consists of 20 cells and a bypass diode. These three strings are connected in series.

Each cell produces 8A and 0.5V (roughly).

A group of 20 cells with a bypass diode produces 8A and 10V.

Whole panel produces 8A and 30V.

When a single cell get shaded, it looses ability to conduct current, so the current through that cell falls to, say 1A and still 0.5V.

The string of 20 cells would procuce 1A and 10V. However, other strings try to push 8A through this string. Reverse voltage rises and bypass diode breaks. Once this happens, you have the string totally disconnected, 8A going through the bypass diode, and about 1V lost in the diode. Net result is 8A and -1V for the string.

When this is combined with other two strings in the panel, you still get 8A and (10+10-1) = 19V. Producton is 19x8 = 152W - roughly 1/3 of the full.

It would be different without bypass diodes. Current would be 1A and voltage perhaps even slightly more - 35V. You would get 35x1 = 35W.

As you can see, to get some production from this shaded panel, you desperately need to drop voltage to 19V, so that the bypass diode would open and let high current for healthy cells through. If you need 24V for the battery, you simply cannot do that. If you only need 12V, you can. That's the difference between MPPT with higher-than-battery voltage and PWM with almost-battery voltage.

I got tired a bit, will continue this post later.

Cariboocoot · August 2013

Re: PWM versus MPPT- a quick analysis

No, I'm talking about the way panels behave: they try to produce current @ any Voltage. When shaded they produce little to no current, which means the load will pull the Voltage down.

I don't want people getting the Voc ability confused with the actual power output. It happens all the time.

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

Cariboocoot wrote: »

No, I'm talking about the way panels behave: they try to produce current @ any Voltage. When shaded they produce little to no current, which means the load will pull the Voltage down.

Exactly. They prodce current at any voltage. But this current is very small even when only one cell is shaded. When voltage goes down, at some point, it gets low enough that the bypass diode across the shaded section opens and boom! you get your full current back (but at lower voltage). However, to get this full current, you need to drop the voltage low enough. And if this happens to be below battery voltage, you're out of luck.

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

NorthGuy wrote: »

I got tired a bit, will continue this post later.

Ok. Here's the continuation.

I left where a 240W panel (8A 30V) with one shaded cell could produce 8A @ 19V = 152W (because of bypass diodes clipping out bad cell group), but only 1A @ 30V = 30W.

What happens when panels are combined. Let's look at 4 panels. Assume 3 "normal" panels and one panel with shorted cell. Without shading it would be 960W.

1. All 4 panels in parallel. At 30V, you get 8A from normal panels and 1A from shaded panel (8x3+1)x30 = 750W. At 19V, you get 8A from all panels 19x8x4 = 608W. Apparently, it's better to maintain higher voltage. Higher voltage practically elimibates the shaded panel, but others perform well. This is the best configuration for PWM. You can get 750/960 = 78% of unshaded production.

2. 2 x 2. 2 parallel series of 2 panels each. Now normal voltage for the string is 60V. At 60V normal string production is uninterrupted, but the shaded string produces only 1A. (8+1)x60 = 540W. If we drop voltage to 49V (because in the shaded series, the normal panel will be at 30V and the shaded at 19V), we'll get full 8A current in both strings (8+8)x49 = 784W. Here it is beneficial to drop voltage a little. If this doesn't drop us below battery voltage, we can get 784/960 = 82% of unshaded production.

3. All 4 panels in series. Full voltage is now 120V. At 120V, we only get 1A and 1x120 = 120W. But we need to drop voltage only a little - to 109V, and then we get full current through our string. 8x109 = 872, and we achive 872/960 = 91% of unshaded production. For this configuration, we'll need MPPT even with 48V system.

Shortly, an ability of the system to use less paralleling (one long string) and an ability to work at lesser voltage both help minimize losses from shading. Only MPPT provides such capabilities.

Caveat emptor: this sure a very primitive explanation that cuts some corners to achive certain simplicity, but it does give the correct idea.

inetdog · August 2013

Re: PWM versus MPPT- a quick analysis

vtmaps wrote: »

If his choice were 4 in series with MPPT or 4 in parallel with PWM, he would be better off with the PWM.

Except for eliminating the voltage drop across the bypass diode(s), I do not see the benefit.
And with the panels in series the power gained from the difference between Vmp and Vbat of the remaining unshaded panels should still be enough to make it more efficient than PWM at full current.
PS: Basically what NorthGuy said with more numbers.

niel · August 2013

Re: PWM versus MPPT- a quick analysis

lazza,
the mppt will be the best choice in a high power array imo. if you aren't sure then purchase a known good mppt cc and see for yourself what it is you like. if i were to miss my guess you will like it and won't go back to pwm excepting in some circumstances.

mppt does improve the overall efficiency of a system by producing more current within the same pv power present. it gets sticky with what ifs and such, but for most times mppt works wonders. even a simple setup like a 12v pv to a 12v pwm cc an mppt will show an improvement in the current unless the pv is hot, but that same hot pv will be reduced in output with a pwm too.

normally you will get the rated pv amps as in imp only with your pwm. when using the mppt it will utilize the power wasted when the voltage is higher than that needed. so you could see a pv with for example 18v as vmp and say 11a (198w) now produce upward of about 12a or so and possibly more at the rated charged battery voltage. 14.4v absorb v point? now here's something interesting is that when the battery is drained further and it reduces the battery voltage that the mppt cc will produce even more current than possibly seen at the absorb voltage point. 198w/11.5v=17.21a. now there will be some loss from inverter operations and such so figure 2a or 3a lost there and you get 14a-15a to the battery. a gain over the pwm of 3a-4a and you can do the %. this is the area where they make the 30% claims from and this is not all of the time this will occur, but occur it can. bottom line is with mppt that if the batteries are taken low then the mppt will provide more current for recovery.

rambling right now, but hopefully you get the general picture. if you cycle your batteries 10%-20% it is not as worthwhile with mppt and most should cycle to at least 80% soc anyway.

vtmaps · August 2013

Re: PWM versus MPPT- a quick analysis

NorthGuy wrote: »

Ok. Here's the continuation.
<snip>
2 x 2. 2 parallel series of 2 panels each. Now normal voltage for the string is 60V. At 60V normal string production is uninterrupted, but the shaded string produces only 1A. (8+1)x60 = 540W. If we drop voltage to 49V (because in the shaded series, the normal panel will be at 30V and the shaded at 19V), we'll get full 8A current in both strings (8+8)x49 = 784W. Here it is beneficial to drop voltage a little. If this doesn't drop us below battery voltage, we can get 784/960 = 82% of unshaded production.

NorthGuy, thanks for the numbers... I realize that some of my generalizations don't hold up under all circumstances. I realize that I must do the numbers for every array, taking into consideration the array configuration, the Vmp and Imp of the panels, and even the internal construction of the panels.

Thus, using the same panels as in your example, consider 5 parallel strings of 2 panels per string: if there is shading on one cell there will be a disadvantage to dropping the voltage to 49 volts and allowing the bypass diode to conduct.

--vtMaps

Photowhit · August 2013

Re: PWM versus MPPT- a quick analysis

NorthGuy wrote: »

...You're arguing the wrong point. It doesn't really matter whether your batteries are in bulk, in absorb, or doing something else...

If you don't understand that MPPT is advantageous in the bulk stage and very early stages of Absorption, and that normal recommendation is for systems to stay out of Bulk... you have missed my point and perhaps could stand a review of how things work...

I was trying to explain your ignorance in your estimating the huge advantage of MPPT. I could go ahead trying to explain, but fear it's a lost cause. My major use is in the summer. I would gladly use PWM with my array with little or no advantage, if I could have purchased panels for the same price in 35 vmp. Your comparison is silly and shows a lack of understanding IMHO. It is exemplified by This statement "The result will be the same regardless of your use pattern."

I feel there are other reasons for using the Midnite classics I have purchased, I'm a fan of the Midnite classics, and I believe there are 2 threads here crediting or blaming me for there use/purchase. I'm NOT against use of MPPT charge controllers! but better understanding of how things work and how batteries charge is advantageous, so I'm putting it out there.

Photowhit · August 2013

Re: PWM versus MPPT- a quick analysis

niel wrote: »

... but that same hot pv will be reduced in output with a pwm too...

Niel as usual is right on the money, though the drop is minor in the case of PWM as heat effects the current minimally while it effects the voltage greatly. So with MPPT there is little 'overhead' for it to take advantage of so the effect of heat is much more dramatic with a MPPT CC compared to PWM.

Cariboocoot · August 2013

Re: PWM versus MPPT- a quick analysis

Let's not be getting adversarial and insulting about this guys.

NorthGuy · August 2013

Re: PWM versus MPPT- a quick analysis

Photowhit wrote: »

If you don't understand that MPPT is advantageous in the bulk stage and very early stages of Absorption, and that normal recommendation is for systems to stay out of Bulk... you have missed my point and perhaps could stand a review of how things work.

I said that lesser array with MPPT is the same as bigger array with PWM. You're saying they're different.

Let's review how it works.

PWM: 4kW array. Vmp = 36 Imp = 111; MPPT: 3.5kW array. Vmp = 36 Imp = 97. Assume best conditions.

In bulk. Battery voltage 27. PWM produces 27*111 = 2997W; MPPT produces 36*97*0.95(efficiency) = 3317W (a little bit better; 3.2kW array would be the same)

In absorption. Battery voltage 29. Current 50. Both MPPT and PWM produce 29*50 = 1450W + whatever goes to loads (both arrays are the same)

In float. Battery voltage 27. Current 0. All production goes to loads (both arrays are the same)

I don't understand at what point 4kW array with PWM is different from 3.5kW array with MPPT. Perhaps you can tell this to me?

PWM versus MPPT- a quick analysis

Comments

Categories