Rednecktek
Solar Wizard
Well, AIO's usually come with 60-80a charge controllers and a Victron gets really expensive really fast to get half that so...
Standalone MPPT more efficient on small arrays?
- Thread starter ilprepper40
- Start date
timselectric
If I can do it, you can do it.
- Joined
- Feb 5, 2022
- Messages
- 18,610
Victron is more efficient at charging the battery.
But when you factor in the double conversion, it's almost a wash. An AIO can take it from the panels to load, in one conversion.
But when you factor in the double conversion, it's almost a wash. An AIO can take it from the panels to load, in one conversion.
I'm not sure what you mean by double conversion, either way it has to go through an inverter. Separate SCC also divert to load depending on how much the inverter needs depending on where the voltage is lower. Anything left over goes to the batteries. An AIO does not change this, it just puts it in a convenient package.
Last edited:
timselectric
If I can do it, you can do it.
- Joined
- Feb 5, 2022
- Messages
- 18,610
Well, I oversimplified it as "double conversion".
A Victron charge controller converts the higher PV DC voltage into lower battery voltage. Then it must be converted back to the higher DC voltage. And then inverted to AC voltage.
The MPPT in the AIO provides the higher DC voltage directly to the inverter.
Last edited:
Thats interesting. It would be curious to see how much power is saved skipping the up convert on the DC side. Thanks for clarifying that. Im guessing that is the reason for the low voltage input specs on allot of the AIOs?
timselectric
If I can do it, you can do it.
- Joined
- Feb 5, 2022
- Messages
- 18,610
Probably
In an AIO everything is connected to a high voltage DC bus. Inverter, SCC, and the bidirectional DC to DC converter/battery charger.
I help out at an install that has two arrays connected to a single 48V FLA battery bank.
One array is connected to a MidNite Classic 200, and the other to a MidNite MN5048DIY which also provides the AC power there.
Earlier this year, I was watching what both were doing on a particularly cloudy day. The Classic had found a point at which it was still charging some. The MN was going through sweep after sweep and not charging at all, nor sending any array power to the inverter. Charge params and operational mode of the AIO were carefully set and triple-checked to ensure that they did not conflict between the two devices. On sunny days both charge fine. Have had other reports of similar behavior on other AIO's when weather conditions are "less than optimal".
My unscientific opinion is that the MPPT algos on dedicated CC's are much better tweaked than the ones in the AIO's. Probably due to years of experience gleaned before AIO's even came on the market.
That, combined with the multiple conversions that timselectric did an excellent job of explaining could explain the higher performance of stand-alone SCC's.
The Chinese AIO mgfrs made a marketing decision - "good enough" will do for new users who have never had a system composed of discrete components and therefore nothing to compare it to. Cheap and easy is where it's at these days.
Somebody said some time back that you need apx. 400W of panel capacity just to offset the +/- 100W "self-consumption" of most AIO's. I will never accept that unless I have absolutely no choice in the matter. It will be separate CC's and big honkin' LF inverters for any system I have.
One array is connected to a MidNite Classic 200, and the other to a MidNite MN5048DIY which also provides the AC power there.
Earlier this year, I was watching what both were doing on a particularly cloudy day. The Classic had found a point at which it was still charging some. The MN was going through sweep after sweep and not charging at all, nor sending any array power to the inverter. Charge params and operational mode of the AIO were carefully set and triple-checked to ensure that they did not conflict between the two devices. On sunny days both charge fine. Have had other reports of similar behavior on other AIO's when weather conditions are "less than optimal".
My unscientific opinion is that the MPPT algos on dedicated CC's are much better tweaked than the ones in the AIO's. Probably due to years of experience gleaned before AIO's even came on the market.
That, combined with the multiple conversions that timselectric did an excellent job of explaining could explain the higher performance of stand-alone SCC's.
The Chinese AIO mgfrs made a marketing decision - "good enough" will do for new users who have never had a system composed of discrete components and therefore nothing to compare it to. Cheap and easy is where it's at these days.
Somebody said some time back that you need apx. 400W of panel capacity just to offset the +/- 100W "self-consumption" of most AIO's. I will never accept that unless I have absolutely no choice in the matter. It will be separate CC's and big honkin' LF inverters for any system I have.
Last edited:
Thats the route I have gone and will continue for now. I won't buy another HF inverter ever again if I don't have to. One thing that has been important to me is reliability/redundancy and low power strings. I don't have any strings over 800w. All fully independent to its own charge controller. I have currently 12 strings/SCCs going on two systems and am adding 5 more strings to one of them. Cant do that with an AIO.
With that said, I can see why many like the convenience of the AIO. Its a brilliant marketing strategy and has made it easy for many to get into the solar world.
timselectric
If I can do it, you can do it.
- Joined
- Feb 5, 2022
- Messages
- 18,610
You missed the point.
It was to the exact opposite.
The extra conversions are with separate components.
not to derail this tread, I shared my ideas here; Feel free to comment on that thread.
Why I keep my panel strings under 1KW and under 80v
link - solar panel fire / google images Edit; updated title.
diysolarforum.com
Ah, dang. My bad. I see the "separate" reference now.
Last edited:
Do not give up on PWM.
MPPT provides more versatility and allows higher voltage panels or arrays to be used with lower voltage battery. If panel operating voltage is only slightly higher above battery charging voltage, PWM controller will operate in MPP zone.
For example a 12V battery and 18V panel with OC voltage about 21V as shown in picture below. PWM controller does not do DC/DC conversion and is just connecting panel directly to battery. The charging voltage of 12V LA battery is about 15V that is very near MPP in optimal conditions 95W and pretty much dead-on MPP in suboptimal 55W for this panel.
If a 36V panel (42V OC) is used with 12V battery, PWM will perform terribly. But it will work with close to MPP efficiency with 24V battery. It may work even with higher efficiency than MPPT because there are no DC/DC conversion losses. Also less heat burden on the controller = higher reliability.
Image credit to Paul Scott from Spheralsolar.
MPPT provides more versatility and allows higher voltage panels or arrays to be used with lower voltage battery. If panel operating voltage is only slightly higher above battery charging voltage, PWM controller will operate in MPP zone.
For example a 12V battery and 18V panel with OC voltage about 21V as shown in picture below. PWM controller does not do DC/DC conversion and is just connecting panel directly to battery. The charging voltage of 12V LA battery is about 15V that is very near MPP in optimal conditions 95W and pretty much dead-on MPP in suboptimal 55W for this panel.
If a 36V panel (42V OC) is used with 12V battery, PWM will perform terribly. But it will work with close to MPP efficiency with 24V battery. It may work even with higher efficiency than MPPT because there are no DC/DC conversion losses. Also less heat burden on the controller = higher reliability.
Image credit to Paul Scott from Spheralsolar.
Attachments
Last edited:
I believe you are missing the point. The main difference between an MPPT and nonMPPT charge controller is the lack of tracking as conditions change.
When I say tracking, I mean the ability to find and keep the higher output at current conditions. Conditions could be solar irradiance, shading, temperature, etc.... The controller will very the draw down/load on panel in the sweet spot, you cant do that with out an MPPT circuit in the controller.
Most MPPT are PWM on the output side (to regulate charge rate) if memory serves correctly.
I agree with what you said except that I missed the point.
Look at graph above. MPP area spans wide voltage range and when comparing optimal 95W and sub-optimal 55W, PV voltage does not change a lot, mainly output current changes with different solar exposure. Solar panel is non-ideal constant voltage source, which is exactly what we want from battery charger. Voltage matched solar panel is already an optimal battery charger without any mumbo jumbo, all we need to do is disconnect it when battery is full.
Example:
PV 95W 18V/5.3A and 0.1ohm wiring, 2.79W lost on wiring, 92.2W on charger input.
PWM uses MOSFET with 10 miliohms ON resistance 0.3W lost on MOSFET, battery receiving 91.9W. There are likely better MOSFETs used, I am just being conservative. Losses on controller are negligible.
MPPT sits at same point of the solar panel's voltage/current curve and uses DC/DC buck converter. 0.5V*5.3A=2.65W lost on schottky. 6%*92W=5.5W lost in switching on transistors and inductor. I am being quite generous with 6% switching losses, real world DC/DC converters are usually worse. Battery receiving 84W about 9% less than PWM.
It is going to have to do lot of tracking during the twilight to catch up to underdog PWM who is getting 9% more during bulk exposure.
To be objective, this is not fair comparison for MPPTs, who thrive at higher PV voltages. Just wanted to demonstrate that there are common scenarios where MPPT would not improve overall performance, and may even hinder it.
Last edited:
Yea, that is interesting. I just check todays output and I was at 50% volts when at 25% amps. So there is some power that would be zero with PWM since voltage would be to low. Maybe not tons of power but over lots of time its good, especially if its a cloudy day. It could just be 15% output for the day but thats better than zero.
You could up volts with PWM to make up lower light but you would loss lots of panel output usage with the high voltage. A 24v panel used with a 12v battery would yield 50% output from the panel nominaly. This is a big advantage with MPPT.
Do not get me wrong, if I am designing a proper system, I use MPPT. My main system is 2.5kW 400V array, 10kWh battery, 5.5kW MPPT AIO
At this point I have lot of solar components on hand. Needed light in the shed, so used 250W panel, 480Wh battery and 10A PWM controller to power 10W light for less than 10min a day. An overkill, but whole setup cost less than $100 Canadian, most parts are second hand. Old 24V lawnmower battery for $20, it has 2 good SLAs inside, rewired to parallel. Panel is from bigger project where I bought 20 for $1200 and used only 18. Found Renogy Wanderer PWM controller in Amazon flash sale for $15, bought 5 of them. PWM, MPPT does not matter, I am just trickle charging the battery anyway. Renogy has well defined charging profiles with regular monthly equalization, if I wanted to buy a decent 10A charger, I would not be able to find it anywhere close to $15. This one has millivolt temperature compensated accuracy and understands MODBUS protocol.
I have better components on hand, Rover 30A MPPT controller, several 100W monocrystaline panels and 24DC battery, that would be better fit for small solar system, but I paid full prices for these components and they would be wasted on the shed where they get <10min of use per day.
At this point I have lot of solar components on hand. Needed light in the shed, so used 250W panel, 480Wh battery and 10A PWM controller to power 10W light for less than 10min a day. An overkill, but whole setup cost less than $100 Canadian, most parts are second hand. Old 24V lawnmower battery for $20, it has 2 good SLAs inside, rewired to parallel. Panel is from bigger project where I bought 20 for $1200 and used only 18. Found Renogy Wanderer PWM controller in Amazon flash sale for $15, bought 5 of them. PWM, MPPT does not matter, I am just trickle charging the battery anyway. Renogy has well defined charging profiles with regular monthly equalization, if I wanted to buy a decent 10A charger, I would not be able to find it anywhere close to $15. This one has millivolt temperature compensated accuracy and understands MODBUS protocol.
I have better components on hand, Rover 30A MPPT controller, several 100W monocrystaline panels and 24DC battery, that would be better fit for small solar system, but I paid full prices for these components and they would be wasted on the shed where they get <10min of use per day.
Last edited:
Benjamin Yang
New Member
Overall, standalone MPPT controllers may perform better in complex conditions, while integrated MPPT controllers may be better suited for smaller, simpler systems. It's best to consider your array size, environmental conditions, and budget when choosing a controller that fits your specific needs.
Similar threads
