Understanding current limits for MPPT

[Eric_Jensen]

I'm planning to build a modest solar generator in order to have a bit of battery backup during a power outage. I have no plans to power the whole house for an extended time - just enough to keep the fridge on for a while, and power a few small lights, etc.

I'm planning on modeling it on this milk-crate build from Will Prowse (probably with a larger battery), which is a slightly modified version of the one in this video. I actually prefer the design in the video (with the AC power going into the MPPT rather than directly into the battery as in the first link), so that the MPPT can manage the battery charging cycle. (As I understand it from the comments on the video, the modification was to avoid problems with having the AC power and solar connected at the same time, which I wouldn't do - though I might also add some in-line diodes as suggested in the video comments.)

My questions involve trying to understand better how the MPPT controller handles different current loads, especially for non-solar input sources:

1. If I have a controller that is rated at, say, 15 A, does the controller manage that limit itself (only drawing 15 A or less), or is it on the user to design a system that will never provide more current than that?
2. Does the answer to #1 change for controllers that allow programming of the max battery charge current, like the Victron SmartSolar? For example, if the max charge current is set to 10 A on, say, a 15 A controller, does that mean that it would not draw more than ~10 A (with possible differences based on different input vs. charging voltages, e.g. if charging at 13 V, 10 A, I understand that it would need to draw 130 W, resulting in input current of 130 W / 12 V = 10.8 A if connected to a 12 V source).

Part of the motivation for these questions is that, in the event of a longer outage (esp. overnight), I'd like to also be able to draw some charging power from the 12 V accessory port on my Tesla. The newer models have a continuous current limit of 12 A from the 12 V accessory port (16 A peak) - there's no physical fuse, but if the car senses too large a current draw from those ports, it assumes something is wrong and shuts down the port. So I need to be able to proactively limit the current draw. (I realize that this isn't a lot of power, but I've measured the fridge average current draw to be 63 W, so getting even 100 W from the car would definitely help a lot with riding out an outage.)

Thanks for helping me understand all of this better! I appreciate people taking the time to read and answer questions in this community.

 

[A.Justice]

An MPPT controller generally requires a voltage of 2-4v higher than the batteries voltage. You would need around 18v to charge a 14.4v battery with an MPPT controller.

To answer your specific question, yes, the controller would limit itself internally, but you wouldn't be able to charge anything from a 12 volt port without a different charger.

For something that low current, you can get 100w or so DC2DC converters from Amazon for under $20. I wouldn't recommend one for constant use, but it would definitely work during a blackout or emergency.

 

[DThames]

I didn't watch the video.

The statement "with the AC power going into the MPPT rather than directly into the battery as in the first link" is not correct at all. From the drawing, the output of the MPPT is connected to the battery. The output from the AC charger is connected to the battery. This is correct.

The AC charger output does not go to the MPPT. However, they are both able to charge the battery.

 

[DThames]

You can charge 12v to 12v with a boost converter. You are actually boosting the 12v to 14v or so to charge 12v with it. The boost converters that I have played with have a current limit setting. These are dumb devices where you have to adjust output voltage, current limit, by turning a tiny pot.

 

[Eric_Jensen]

I didn't watch the video.

The statement "with the AC power going into the MPPT rather than directly into the battery as in the first link" is not correct at all. From the drawing, the output of the MPPT is connected to the battery. The output from the AC charger is connected to the battery. This is correct.

The AC charger output does not go to the MPPT. However, they are both able to charge the battery.

Sorry if I worded this in a confusing way. In the video (at about the 8:50-9:15 point), the DC output of the AC/DC power source goes into the PV input of the MPPT. As you note (and as I was trying to say originally), in the other linked schematic, it goes directly to the battery.

 

[DThames]

Sorry if I worded this in a confusing way. In the video (at about the 8:50-9:15 point), the DC output of the AC/DC power source goes into the PV input of the MPPT. As you note (and as I was trying to say originally), in the other linked schematic, it goes directly to the battery.

I would not consider that a smart/wise design. The power supply output is a switching supply. The input to the MPPT is a dynamic load. This load is not constant, as the MPPT varies the load trying to find the spot that it can get max power. The power supply is trying to old a constant voltage output regardless of the load. If the MPPT charger could demand 20 amps and if the AC charger could only supply 20 amps, it would be on the edge for that reason alone. The AC supply would need a DC output that was in excess of the MPPT's max load. If you had a 10 amp PWM charger and a 20 amp AC power supply, you would be (maybe) in a better realm. There is also a lot of noise on the output of some of those AC-DC supplies that a solar charger is not designed to deal with on the inputs.

Bottom line, it would be wise to avoid such a setup.

 

[Vigo]

I might be misunderstanding something. If you're going to the trouble of building something, the added 'difficulty' of skipping the 12v accessory port and going straight to the 12v battery is negligible and gives you access to the full potential of the car's dc-dc converter, which is probably 700-1000w. Im assuming the car would have to be 'on' to get continuous 12v power from the accessory port anyway (and not simply unswitched access to the 12v battery itself which would drain down during use).

Plenty of people have done that kind of thing, use the onboard dc dc of an EV to power a smallish inverter (or a bigger inverter, if you have more battery capacity added between the car and the inverter) during power outage. It's a solid plan.

As far as hooking a power supply to an mppt vs to battery directly, the battery serves to 'smooth' the power supply's output (if it matters) and somewhat isolates the power supply from demands beyond what it is designed for since the power supply is never having to feed into a dead short, it will only have to feed into whatever the battery voltage has dropped to under that load. Additionally if you hooked it to an mppt it would have to be a higher voltage than the battery itself as mentioned. Something like a laptop charger @ ~19v would work for that.

 

[Eric_Jensen]

I might be misunderstanding something. If you're going to the trouble of building something, the added 'difficulty' of skipping the 12v accessory port and going straight to the 12v battery is negligible and gives you access to the full potential of the car's dc-dc converter, which is probably 700-1000w. Im assuming the car would have to be 'on' to get continuous 12v power from the accessory port anyway (and not simply unswitched access to the 12v battery itself which would drain down during use).

Plenty of people have done that kind of thing, use the onboard dc dc of an EV to power a smallish inverter (or a bigger inverter, if you have more battery capacity added between the car and the inverter) during power outage. It's a solid plan.

Thanks - that is what I did with my Nissan Leaf, and with my Toyota Prius before that. The problem with the Tesla is that it’s too smart for its own good (or at least for my good!). From reading long discussions of this in other (Tesla-specific) forums, tapping the car’s 12v battery directly has the same problem and results in the same kind of overcurrent shutdown I described above. (This was less of an issue in some older Tesla models.) There is a spot underneath the rear seat where you can tap the DC-DC converter more directly (designed e.g. for high-powered audio systems) where the car will tolerate a higher load without shutting down, but it’s a bit of a pain to get in there and wire something (and honestly I’m a little leery of getting too deep into the innards of this complicated car). I was hoping to avoid that hassle and go with something more straightforward to connect. It would indeed require building something separate, but would have some other flexibility (e.g. being able to have a portable unit for running other things away from home if desired).

Regarding connecting a power supply to a LiFePO4 battery directly, my (possibly incorrect) understanding was that it was better for these batteries to have something “smarter” to manage the charging cycle, rather than just a constant source. Is that not the case?

 

[DThames]

"Regarding connecting a power supply to a LiFePO4 battery directly, my (possibly incorrect) understanding was that it was better for these batteries to have something “smarter” to manage the charging cycle, rather than just a constant source. Is that not the case?"

You need to understand the charging profile but a cc/cv supply with charge them just fine. You need to have a BMS and to stay away from the point where a fully charged cell starts running (to a higher voltage). As the cells are nearing full charge, a little more charge to the cells will make the voltage to up more quickly. There is always one cell that takes off toward 3.6v before the others. If you change to about 3.45 to 3.50 volt per cell, you don't see this happen (as much). I charge my 24v packs to about 27.2v (3.4v per cell). You can hold them at that voltage without much issue. I would rather turn off the charger than hold them at 3.4v for where you would not want to hold them like that for weeks, but for hours, not an issue. Just plug in a modest power supply and leave overnight and they cells will be charged. Works well for generator charging as well, if you have an emergency system and no sun. Maybe double up on them if you know you need more amps quickly, then remove one at the end of the charge cycle.

 

[Vigo]

Regarding connecting a power supply to a LiFePO4 battery directly, my (possibly incorrect) understanding was that it was better for these batteries to have something “smarter” to manage the charging cycle, rather than just a constant source. Is that not the case?

Well, if you're talking about a pre-built battery that has some kind of BMS managing its cells, then the bms will disconnect from the charger whenever any one cell hits the 'max voltage' the bms wants to see, and from there it will probably start discharging the full cell into the lowest cell to 'balance' the cells. That's my understanding, anyway, im pretty new to it. So any battery with a bms you COULD just let the BMS disconnect the battery from the charge source when it's 'full'. But i believe how fast you charge it up to that point will cut a bit off your cycle life and contribute to greater imbalance over time, as i believe most bms' balancing happens at a snail's pace compared to regular charging and discharging, such that the amount of balancing it accomplishes between 'cycles' might not keep up over time, letting the cells get further and further apart.

Either way, when it comes to recharging you have to replace whatever you use in a 'cycle' and that's going to require a certain number of amps X hours. When your hours are limited (daylight, for example) your amps are somewhat dictated to you as well. At that point the only thing you can do is not to slow your charge but to add more batteries such that the charge current is split between them, resulting in a lower 'c rate' of charge, and less issues over time.

Hopefully someone will confirm or deny all that because i'm not 100% confident in my explanation.