MPPT vs. PWM – Still Not Sure

[4arch]

I’m planning on installing solar panels (around 300 watts) on my boat in the spring. I’ve done a lot of lurking on solar threads on this and other sailing sites and also on solar energy forums. Opinions about controllers for small scale systems (<600 watts) seem to vary a bit between the sailing and solar forums. The general consensus of the sailing forums seems to be that, if the budget allows, MPPT is always the way to go as it will make the most out a limited amount of real estate on a boat. The consensus of the solar forums seems to be that while MPPT is nice to have, the real-world performance difference between a high quality, programmable MPPT controller and a similarly high quality, programmable PWM is so negligible for a small system as to almost never justify the added cost of MPPT, even when panel real estate is limited. The solar forum people also generally say that those who claim significant performance boosts by changing from PWM to MPPT probably aren’t comparing apples to apples – that they probably swapped a low quality, non-programmable PWM for a good quality MPPT and are simply seeing the “boost” from no longer undercharging with a “one size fits all” charging sequence. I’m certainly paraphrasing and generalizing the arguments made on both sides, but that’s what I’ve taken away.

I had initially been planning on going with PWM, then became sold on MPPT, and am now back to thinking a Tristar 45 PWM would be my best bet. It really comes down to real-world performance. If there really is a “boost” from MPPT of 5 or 10 AH or more daily average, I can see its value. But I’m just not convinced I’ll see that theoretical gain “out there” and it seems like the extra cost of MPPT might be better put toward conservation. I’m probably already over thinking this, but am I coming at this all wrong?

 

[kloudie1]

Maine Sail will want to weigh in on this one.. Meanwhile, check out posts 1,3, 7 of this thread:
http://forums.hunter.sailboatowners.com/showthread.php?p=979379&highlight=MPPT vs PWM
Seems to me that the differences would be exaggerated by small, boat sized panels that move and get shaded occasionally .. The land based panels are generally optimally oriented and not partially shaded..

 

[JohnShannon]

MPPT uses a bit of power electronics that converts the solar panel voltage to the battery voltage with about 97% efficiency. Typically the battery voltage and the optimum solar panel voltage are not the same voltage. The end effect is a panel with a MPPT controller between it and the battery will deliver higher charging current to the battery than a direct connection of the panel to the battery.

The PWM type of solar charger is switching between a direct connection of panel to battery and (when the battery is full) no connection. So PWM will avoid overcharging but won't get max out of the panel

 

[JohnShannon]

so the final analysis you would have to decide I suppose if a cheaper pwm controller and a bigger panel would be a better deal than a smaller panel and an mppt controller.

where both systems are dimensioned for equivalent Ahr's

 

[Maine Sail]

I’m planning on installing solar panels (around 300 watts) on my boat in the spring. I’ve done a lot of lurking on solar threads on this and other sailing sites and also on solar energy forums. Opinions about controllers for small scale systems (<600 watts) seem to vary a bit between the sailing and solar forums. The general consensus of the sailing forums seems to be that, if the budget allows, MPPT is always the way to go as it will make the most out a limited amount of real estate on a boat. The consensus of the solar forums seems to be that while MPPT is nice to have, the real-world performance difference between a high quality, programmable MPPT controller and a similarly high quality, programmable PWM is so negligible for a small system as to almost never justify the added cost of MPPT, even when panel real estate is limited. The solar forum people also generally say that those who claim significant performance boosts by changing from PWM to MPPT probably aren’t comparing apples to apples – that they probably swapped a low quality, non-programmable PWM for a good quality MPPT and are simply seeing the “boost” from no longer undercharging with a “one size fits all” charging sequence. I’m certainly paraphrasing and generalizing the arguments made on both sides, but that’s what I’ve taken away.

I had initially been planning on going with PWM, then became sold on MPPT, and am now back to thinking a Tristar 45 PWM would be my best bet. It really comes down to real-world performance. If there really is a “boost” from MPPT of 5 or 10 AH or more daily average, I can see its value. But I’m just not convinced I’ll see that theoretical gain “out there” and it seems like the extra cost of MPPT might be better put toward conservation. I’m probably already over thinking this, but am I coming at this all wrong?

Keep in mind that MPPT only really benefits, and has a good advantage, when your batteries are in bulk charge. With approx 15A +/- of current available you will be in bulk MOST OF THE TIME... Thus you will be benefiting from MPPT most of the time you are charging with solar or well into the mid 90% SOC range depending upon bank size..

I have and use a device that I track and compare solar performance with, a custom built PentaMetric monitoring system. MPPT wins on boats but the gains are usually less than 15%, unless conditions are perfect.. Still this is an excellent boost when you are out of real estate which most boats run out of faster. On boats I have yet to see PWM beat MPPT, not once, and my monitoring equipment is simply an impartial observer......

PLEASE DO NOT BUY AN eBAY SPECIAL MPPT.... There are many algorithms for MPPT and some really SUCK. The controllers I like best are the small Genasun's. the Rogue Powertech and Morningstar controllers. Rogue just launched a brand new 20A model for $250.00 which looks really, really nice. You really can't beat the Tristar 45 MPPT or the Rogue.. Theya re both very well executed products. IMHO you get more for your money with Rogue....

You need to be careful when trying to read solar forums and boat forums on solar. If you read only off-grid solar than the conclusions can be closer but grid tie etc. forget it not the same type of use.. Ohm's law does not change but they way we use it, the way we wire it, the panel angles, temps and the voltage we use them at do.. In off-grid panels are often wired in series vs. boats which are usually parallel to avoid shading issues. This changes things. Panels on roof tops get very hot where panels on boat davits etc. stay considerably cooler thus allowing higher voltages... Our panels are also not directly aimed at the sun which also helps them run cooler but also digs into overall performance.

In short it is rare that I install a PWM on a boat system that is over 100 watts unless there are extreme budget constraints..

This simulation below was to show the difference in bulk between a shunting controller (PWM would do exactly the same) and an MPPT. Keep in mind that your solar panels performance would never stay as static as my power supply. Good quality MPPT controllers use a P&O or perturb and observe algorithm. I like to call it a hill climb algorithm. This means it is constantly testing the maximum power point to see if it can covert more voltage to current. It does this by pushing up and up and up until it can't go anymore. The calculations are done on the fly and no measurable pauses exist in current delivery. They do this by testing the voltage very rapidly multiple times per second. You can see how fast the Genasun is doing this by watching the Fluke meters current reading.. It wastes no time in the algorithm.. Cheap Chinese controllers shut down, observe panel voltage, then re-start. On some real cheapies this can often mean worse performance than a PWM.

 

[4arch]

Thanks for these replies! I definitely don't want to buy a cheapo generic controller which may be why I'm (over)analyzing this so much. The Rogue units look good and seem to include a lot of features the competitors don't in that price/size range. I might just buy one and stop worrying about it.

 

[Maine Sail]

Thanks for these replies! I definitely don't want to buy a cheapo generic controller which may be why I'm (over)analyzing this so much. The Rogue units look good and seem to include a lot of features the competitors don't in that price/size range. I might just buy one and stop worrying about it.

The reason you get more is because there is no middleman. If he added multiple middlemen, distribution layers etc. it would be one of the most expensive controllers out there, and still worth it...

 

[walt]

You get a little benefit from MPPT in the bulk charging (like 5 to 15% increase in current at the battery), a big deal to some, not a big deal to others. This increase also gets worse when its warm and most people use boats when its warm.

However, the other benefit of MPPT which in my opinion is more significant is that because MPPT transfers the input power to the output power through magnetics, the input and output voltage can be much different.

Almost always we have 12 volts DC systems but 12 volts is really a poor choice if you have a lot of power to transfer. The reasons is that the power loss in the wire goes up with the square of current. For a given power, doubling the voltage results in half the current - and 1/4 the power loss in the wire distribution so considerably less copper is required.

As mentioned, serial panels to get higher voltage has partial shading issues but two higher voltage panels in parallel wont have this issue. And the great thing about MPPT is that it can take this higher voltage from the panels and convert it to the 12 volts charging for the batteries. But with the higher voltage from the panels, the amount of copper used in the wiring is considerably less which benefits both weight and cost.

Most MPPT controllers do allow higher voltage panels even when charging a 12 volt battery. I think that Rogue 20 amp one works with panels even up to 100 volts.. There is probably a safety issue when the DC voltage gets very high but I would feel fairly comfortable with 24 volt panels. You see even higher voltage systems on boats (usually for electric motors) so maybe higher is safe also.. I dont know.

Don’t know about specific "marine" panels (is there such a thing?) but a lot of panels coming from the solar industry are already higher voltage - just because the power distribution is better at higher voltage. The MPPT controller is a great way to use these to charge a 12 volt battery..

 

[walt]

Maine Sail, your one of a handful of people who I trust what they say without it being backed up..

But, I have to question one thing you said since it has some broad implications

difference in bulk between a shunting controller (PWM would do exactly the same)

The issue iis saying a PWM contoller will do the same as the "shunting" contoller. I don’t think this is at all the case.

I am partially guessing on this (wish more people would say that when they are)... But the term "shunting controller" refers to a method of disconnecting the panel from the battery and this method is now pretty much obsolete. It has nothing to do with the charging algorithm.

The problem you have with the Flex charge controller has to do with its charging algorithm. I would bet that is a very old simple completely analog design. A simple analog design for the algorythm would be very “dumb”.

Any PWM controller that is designed today would use a cheap microprocessor for the charging algorithm. The processor is cheap and the software development is cheap.

The way the panel gets connected to the battery could be either shunting or FET pass through. Both work just fine but the FET pass through is simpler, cheaper and has some small benefits.

What we call PWM solar controllers are usually more modern designs and use a FET pass switch and a microcontroller that samples the battery voltage. The microcontroller can have a very sophisticated algorythm that can sample the battery voltage many thousand of times per second and also control the pass switch many thousands (or even millions) of times per second.

So what you are calling a shunt controller (like the Flex charge) is likely WAY different than say a modern Morningstar PWM controller. The difference is in the charging algorithm. The old design is extremely limited vs. the newer designs using microcontrollers.

Try a modern PWM controller in that experiment, it will behave a lot differently than the Flex charge.

FYI..

Shunting uses a switch to ground (probably a bipolar transistor) to short out the panel and an isolation diode to disconnect the shorted panel from the battery. It wont damage a panel to short it (you get the short circuit current but very low voltage so the power is almost zero). When the shunt in turned off, the panel is now connected directly to the battery through the isolation diode (which is now forward biased). I saw a charge controller patent filed about 25 years ago that used this method to disconnect the panel from the battery.

The FET switching that is now used in PWM controllers simply uses a FET as a switch and directly connects the panel to the battery – or disconnects it. FET’s are cheap and easy to drive so a simple circuit.

 

[Bill Roosa]

and all this time I thought MPP was an impedance matcher to get max power out of the panel and PWM was a way to control the charging voltage going to the batteries!!!
I would think that if you are restricted in panel size (aka panel power) and you want to maximize power delivered to the batteries WITHOUT concern for over charging them you would want both MPP and PWM.
MPP to keep the panels at max power voltage and PWM (down stream from the MPP) to control the charging of the battery.
another way to look at this is there is a voltage that produces the max power out (volts times amps) from the panel. MPP keeps the voltage of the panel at this optimal voltage and delivers some fixed voltage to the batteries for charging so you don't get the "three stage" charging that is most effective.
there is a desired voltage that you want to charge the batteries based on their SOC. so the PWM controller takes what the panels are delivering and bumps it up or down to charge the batteries most efficiently based on SOC.
these two voltages (output of MPP and output of PWM)are generally not the same for most situations. So the MPP controls the panel and the PWM controls the batteries. or panel-->MPP controler-->PWM-->batteries
But I don't have an electrical engineering degree so I'm probably FUBAR

 

[Maine Sail]

Maine Sail, your one of a handful of people who I trust what they say without it being backed up..

But, I have to question one thing you said since it has some broad implications



The issue iis saying a PWM contoller will do the same as the "shunting" contoller. I don’t think this is at all the case.

I am partially guessing on this (wish more people would say that when they are)... But the term "shunting controller" refers to a method of disconnecting the panel from the battery and this method is now pretty much obsolete. It has nothing to do with the charging algorithm.



The problem you have with the Flex charge controller has to do with its charging algorithm. I would bet that is a very old simple completely analog design. A simple analog design for the algorythm would be very “dumb”.

Any PWM controller that is designed today would use a cheap microprocessor for the charging algorithm. The processor is cheap and the software development is cheap.

The way the panel gets connected to the battery could be either shunting or FET pass through. Both work just fine but the FET pass through is simpler, cheaper and has some small benefits.

What we call PWM solar controllers are usually more modern designs and use a FET pass switch and a microcontroller that samples the battery voltage. The microcontroller can have a very sophisticated algorythm that can sample the battery voltage many thousand of times per second and also control the pass switch many thousands (or even millions) of times per second.

So what you are calling a shunt controller (like the Flex charge) is likely WAY different than say a modern Morningstar PWM controller. The difference is in the charging algorithm. The old design is extremely limited vs. the newer designs using microcontrollers.

Try a modern PWM controller in that experiment, it will behave a lot differently than the Flex charge.

FYI..

Shunting uses a switch to ground (probably a bipolar transistor) to short out the panel and an isolation diode to disconnect the shorted panel from the battery. It wont damage a panel to short it (you get the short circuit current but very low voltage so the power is almost zero). When the shunt in turned off, the panel is now connected directly to the battery through the isolation diode (which is now forward biased). I saw a charge controller patent filed about 25 years ago that used this method to disconnect the panel from the battery.

The FET switching that is now used in PWM controllers simply uses a FET as a switch and directly connects the panel to the battery – or disconnects it. FET’s are cheap and easy to drive so a simple circuit.

Walt,

That was in the context of bulk charging. Both PWM and shunting controllers essentially direct connect o the battery in bulk. Once at absorption the behavior changes and a shunter begins on/off using voltage decay to a preset level before turning back on. A PWM maintains a voltage and does this by pulsing to maintain and not overshoot an sbsorption or float voltage set point. PWM controllers charge significantly faster than a shutter for absorption but in bulk they behave the same and essentially direct connect..

If you have seen a PWM that does not direct connect in bulk I love to know which one...

Keep in mind that test was in full bulk showing the differences in bulk. I have done the same with PWM and the bulk current to the battery in no different than a shutter...

 

[walt]

difference in bulk between a shunting controller (PWM would do exactly the same)

You are right.. I see in your original message you said "bulk".

I think the thing you have the issue about the "shunting" controller is the part where it reaches 14.2 volts, shuts off and waits until it hits 13.2 volts and then turns back on. That is a vey dumb algorytm, likely a very simple analog circuit designed about 20 years ago. But it does the basic function of charging the battery but keep the panel from overcharging. As you said in your first paragraph, the modern PWM controller works much better in the absorption phase.

 

[Maine Sail]

You are right.. I see in your original message you said "bulk".

I think the thing you have the issue about the "shunting" controller is the part where it reaches 14.2 volts, shuts off and waits until it hits 13.2 volts and then turns back on. That is a vey dumb algorytm, likely a very simple analog circuit designed about 20 years ago. But it does the basic function of charging the battery but keep the panel from overcharging. As you said in your first paragraph, the modern PWM controller works much better in the absorption phase.

Yes VERY DUMB!!! However there are still PILES of these controlelrs sold every year especially by places such as eBay etc... A general rule is that if it does not specifically say PWM than it is likely not PWM...

Keep in mind that most any MPPT controller essentially becomes a PWM once at absorption voltage. It is in bulk where the gains are made with an MPPT controller...

On boats, due to real estate constraints, we are in bulk with solar most of the time so your current gains from MPPT can go from 50% SOC all the way into the high 90's SOC wise, depending on your panel current to bank size ratio....

PWM will finish charge a battery far better and far faster than a shunter will as the decay time from 14.2V to 13.2V or what ever the particular shunter does can burn most of your sun up hours away. It will also gas your battery far better, which is a NEEDED part of the charge cycle, to keep sulfation at bay.

I have had AGM banks that will hold over 13.1V for a few hours and a Sun Force shunting controller that did not turn back on until 13V. It would hit 14.2V than take an hour or more for the voltage to decay in order to kick back on and bring the voltage back to 14.2V before turning off for another hour.

*A PWM brings the bank to target voltage and then HOLDS it there....

*An MPPT controller brings the bank to target voltage and HOLDS it there...

*A shunter brings the bank to target voltage then SHUTS OFF and only turns back on when the voltage has decayed to the ON set point...

 

[Bill Roosa]

Mainsail wrote
"*A PWM brings the bank to target voltage and then HOLDS it there....

*An MPPT controller brings the bank to target voltage and HOLDS it there..."

Are you saying that the MPP controller is also a three stage charger? That would imply that it has a PWM section inside would it not?

 

[Maine Sail]

Mainsail wrote
"*A PWM brings the bank to target voltage and then HOLDS it there....

*An MPPT controller brings the bank to target voltage and HOLDS it there..."

Are you saying that the MPP controller is also a three stage charger? That would imply that it has a PWM section inside would it not?

Yes absolutely. Never seen an MPPT that did not incorporate three stage charging and voltage control. It is most often in the form of PWM regulation which is widely used today for voltage regulation and not just in solar controllers..

The gains in MPPT come when the bank is in bulk mode and the controller can turn excess voltage into more current when the excess voltage is available..

Once at absorption voltage the bank will only take what it will take in current to maintain that voltage and this accepted current will continue to decline at that voltage until the bank is full. When an MPPT controller hits absorption voltage they essentially go into PWM mode...

This is but one description, from a manufacturer, about MPPT and its PWM voltage regulation stage:

"Blue Sky Solar Boost 2000E:

Patented Maximum Power Point Tracking (MPPT) technology
allows Solar Boost 2000E to increase charge current up to 30%
or more compared to conventional charge controllers.

The Solar Boost 2000E provides a precision Multi-stage Pulse Width
Modulation (PWM) charge control system to ensure the battery is
properly and fully charged, resulting in enhanced
battery performance with less battery maintenance. An equalize
function is also included to periodically condition liquid
electrolyte lead-acid batteries."

 

[Bill Roosa]

Sooooo if you get a MPP controller you would also get (assuming it ain't a cheap one) the PWM features.
Sounds like a no brainier for the panel size challenged. Is cost the only venerable here?

 

[Bill Roosa]

I hate auto complete. Should be variable not venerable

 

[dawa996]

well, I am just curious, and find out that the MPPT controller will increase 30% efficiency. not so sure of the PWM. I think MPPT is more expensive and looks better than PWM controller, they should not be in the same lever to compare.

 

[mitiempo]

...and find out that the MPPT controller will increase 30% efficiency. not so sure of the PWM.

More like 10% increase than 30%

 

[zeehag]

I THINK ONE WILL FIND THE PWM BY SELF IS NOT ADEQUATE FRO BRINGING BACK ALMOST DEAD OR DEPLETED BATTERIES. THE CONTROLLER MUST BE BYPASSED TO RETURN LOOOWWWW BATTERIES TO FUNCTION.
ask me how i know..i will be trying an mppt controller this year to see if there is actually a difference.
i have been using pwm for 10 years, and is time for a change.