You don't need (or want) an MPPT on the output of the battery towards the inverter - it can't bring the voltage up, while you want a steady output. Batteries are not current sources, so you don't need to track a maximum power point.
excited about Sodium Ion 200 AH cells from Docan Tech
- Thread starter solarHandyman
- Start date
You don't need (or want) an MPPT on the output of the battery towards the inverter - it can't bring the voltage up, while you want a steady output. Batteries are not current sources, so you don't need to track a maximum power point.
SouthernSolar
Solar Enthusiast
Let's do a simple 12 volt system
7 sodium cells at their low point 1.9 equals 13.5 volts. So on your charge controller , you set the output to 13 volts .
When the sodium cells are fully charged it would be 27.3 volts. So you let you charge controller knock it back down to to 13.
Charge controller would be a lot more efficient. A DC to DC conversion is typically only 85 to 93 percent efficient off the top of my head. Where a charge controller is in the high 90 percent efficient.
7 sodium cells at their low point 1.9 equals 13.5 volts. So on your charge controller , you set the output to 13 volts .
When the sodium cells are fully charged it would be 27.3 volts. So you let you charge controller knock it back down to to 13.
Charge controller would be a lot more efficient. A DC to DC conversion is typically only 85 to 93 percent efficient off the top of my head. Where a charge controller is in the high 90 percent efficient.
jorby
Solar Addict
The conversion process (buck) is the same in both units with your example. The greater "efficiency" of charge controllers is typically when they are feeding the HV bus of an all in one inverter. That saves a second downconversion to battery voltage.
A decent synchronous buck converter (since we're assuming now that we only need to bring the voltage down) can achieve well above 95%, with minimal control overhead, and be cheaper, simpler and smaller than an MPPT controller.
Besides, as pointed out above, an MPPT controller also includes a buck/dc-dc converter:
Last edited:
An MPPT is a DC to DC converter, one whose input loading is adjusted to optimize the power input.
For a typical hybrid inverter, it is a boost converter to take the PV input and boost it, if needed, to the AC inverter high rail. If the PV is already above what is needed, the MPPT just passes it through unchanged. It usually does not have a buck conversion option, instead the AC inverter section simply uses lower duty cycles when making the AC waveform from a higher than necessary rail voltage. Not having a buck be part of the MPPT on a hybrid saves money and it possibly a bit more efficient.
The AC inverter high rail will be the higher of the highest MPPT input or about 360 VDC (for US 240 VAC inverter). If all you PV inputs are under 360 VDC, all MPPTs will be boosting to about 360 VDC. If one MPPT is 400 VDC and the other lower, then the higher one passes the 400 VDC unchanged and the other MPPTs have to boost to 400 VDC. When spec sheets show MPPT efficiency is 99.9%, that's the pass through mode, no switching or boosting, but that's not the efficiency over the entire MPPT voltage range. A higher than necessary rail for the AC inverter reduces its efficiency, too. If the MPPT has to boost, then you might be down to 96% or 97% efficiency to the AC inverter rail, then the inverter losses on top of that.
Ideally, you want all your PV strings to come in right at 360 VDC. This maximizes the efficiency of both the MPPT and the AC inverter section. A terrible setup is where you have one PV at near the max input, and then others near the min input. Those min input MPPTs have to boost so high, and the AC inverter loses efficiency with the high rail voltage.
Vendors are really adept at making their efficiency specs look great. They do this by operating everything at the ideal point. For real world uses where things aren't optimal, the numbers are not as good. I wish vendors would publish real charts on MPPT efficiency versus input voltage. That would be eye opening to some.
Mike C.
Last edited:
DPC
Canada, Great already
Old article from last August before we put a stupid Tariff on them..
Ford GM Chrysler dont even make cars anymore so why bother with a tariff?
Ford GM Chrysler dont even make cars anymore so why bother with a tariff?
SouthernSolar
Solar Enthusiast
Well we all agree that the wide voltage range of sodium batteries can easily be dealt with using current technology and I think we all agree it is generally more efficient to reduce a voltage, than it is to boost it.
jorby
Solar Addict
On the 18kpv, I have 3 bus voltages that I can monitor. The "P bus" is 200v, "bus1" is 400v and "bus 2" is 300v. They stay very close to these nombers all the time, even when charging at 12kw from the pv. Specified "nominal" pv voltage is 360v which has always seemed odd as there is no bus at that voltage. The mppt range is 100-500v. Mine is set up to operate at 400. They claim max efficiency of 99.9% pv to battery but if that were true, the fans would not have to run continuously, so I do not believe them. I do not see how they can claim that conversion efficiency, short of magic.
EDIT: Those bus voltages do actually vary after all! See future post below.
Last edited:
That is either a typo or lie, whichever way you want to interpret it. 99.9% is believable for an MPPT operating at the ideal bus voltage.
The FlexBOSS21, which has very similar design to the 18Kpv, claims PV to battery at 94.5%. That's a reasonable number and I suspect the 18Kpv is operating around that figure.
All they have to do is type the number 99.9 in the datasheet. See how easy it is to claim something?
Sounds like your "bus 1" is the MPPT output rail and AC inverter input rail. What does it read when there is no PV input and you are using battery to generate AC? I'd expect it to be around 360 volts.
Mike C.
wpns
Solar Joules are catch and release
jorby
Solar Addict
Yup! Same here.
The 250ah were available up until this past oct…
99.9% might be accurate when the PV voltage is the same as the high voltage bus. The 94.5% figure includes the high voltage DC to battery voltage DC conversion.
jorby
Solar Addict
I have been monitoring the 18kpv bus voltages more closely and see that the 400v has gone up as high as 470v when the batteries are full and the mppt allows the pv voltage to rise. So it works like stated below:
I am guessing your string voltage is about the same as mine from the red line on the graph.
My 5 strings to this inverter are all the same at 12 in series and two of the mppt have 2 parallel strings each. They run 400v to 410v at MPP and 470 OCV. While 11 series would initially seem to be "more optimum" as it is closer to the 360v level, I notice that the 400v buss is at 400v all night long when no charging is taking place, and the buss goes up to 425 while pv charging at maximum power.. The only time it is much higher is when lower charge power is being transferred by the mppt because the batteries are full and efficiency is not very important anymore. I don't think I would benefit from rewiring them to a lower voltage. I think if the input was 360v, it would be boosted to 400 or 425 anyhow.
I am guessing your string voltage is about the same as mine from the red line on the graph.
wpns
Solar Joules are catch and release
String voltages:
and bus voltages for the last 4 days:
Looks like when any string exceeds 400V it bumps the bus voltage up, but I do have '300v' strings, can't tell without turning off the '500v' strings if they would bump the bus, but I tend to doubt it.
jorby
Solar Addict
I think you are correct. That would go along with the idea that the mppt will boost to 400v but just let it go through if above 400v. I also have 3 generic 6.2kw hf AIO inverters that I bought just for the 450v mppt charge functions. I took one apart and traced everything out, all the way to the battery input. The first active component after the input current sensor is the inductor in the classic boost converter. If the voltage is high enough, the mosfet switches do not operate and the voltage goes right on through the diode to the buss. The mppt function is implemented downstream by loading the bus with inverter output and/or battery charging via the 48v transformer based bi-directional inverter stage. Hard to believe that they were only $300 each. (220v only version)
Similar threads
