24V solar system questions

Z

Zagy

Member
Joined
Apr 5, 2017
Messages
36
Hello all, I am new to the whole Powerwall thing, and I am in the process of gathering info andmaterials for a mini DIY Powerwall and I have some questions. I would very much appreciate if someone can point me in the right direction.

The idea I have is for a 3S12P battery pack (currently 25.5 Ah, eventually will double it - laptop batteries are hard to find around here and it takes time) and I am planning on running the lights in myapartment from it(all lights are either LED or CCFL). Due to limited budget and space I have settled for a12V system, with a 80W solar panel (now I have a 20W one, but will purchase the 80W), and the biggest dilemma is the charge controller. I have a Steca Solsum 6.6F, but this is best suited forlead-acid batteries, and the ones I will use are Li-Ions. Should I use a buck converter with the Steca(like XM4015 set to 12.4 or 12.5 volts) for a poor man's MPPT charging, and if I use this option - should I tie it before or after the Steca? Or get thisBQ24650 MPPT solar charger? Which will give me the best possible results?
 
3S li-ion makes the whole electronics choice tricky at the moment, IMO of course.

I got an MPT7210 after hearing about them here and have it charging 24p6s nicely from a single 12V 100W panel..

I'd consider going 6s with the MPT7210 and then you can use them as two independent 3s packs or even better just parallel both of them for discharge.

Charge at 24V (the 7210 won't really do 12V but does 24V happily) and discharge at 12V.

It's a convoluted little beast but about the lowest cost for a starter DIY with lithium with the caveat you need to be charging 5 or more serial cells with it on a 12V panel.
 
Also, you need to be aware that 3s cell arrangement isnt really good on the cells as you would be needing to charge all the way up to 4.2v to get any length of time from them. You should instead go with at least 4s because it gives a larger buffer and you can safely opperate in the 3.4v -4.1v range; which is a lot happier for the cells to run at.

However, I agree with elmo with 24v set, instead.
 
Thanks... I am starting to ponder the 24V option now.

If I use 2x80W panels and 2 x BQ24650 MPPT controllers for 24V system - can I (and how exactly) join them in series? Or if this cannot work - can I use a buck converter to lower the voltage on the 2 panels in series (they put out more than 40V, and the BQ24650 can handle 28V at most).

In this case I can go 7s on my battery pack.
 
As i understand the BQ24650 only do 6s so it wont work properly with 7s system. For 7S you can use the 7210 as sugested. You said that 2 panels of what you have would in series do 40V. That would give you 20v in paralell and that would work with the 7210 on a 7s pack.

Since the 7210 only boost it would need the panels to have its peak point below the battery. Preferable 2-3V.

/D
 
Thanks... I am starting to think about 7s 24V system now - this means I can use my Steca 6.6F 6A controller in 7s config directly with either 2x80W 20V panels, or one 160W 40V panel (I need to check this because of the limited space I have for the panel)... and I have another question (correct me if I am wrong): the Steca 6.6F is PWM type - lets say I go with 2x80W 20V panelsin series, and each panel's Vmpp is 18V - if I use a buck converter and set it at 36V total and place it after the panels but before the solar controller, or use 2 buck controllers with the two panels, and tie them in series - would it make some kind of MPPT-ish system?
 
MPPT is a bit like getting the correct "gear" when driving a stick, if you are in too high a gear the engine lugs down and has poor performance, too low a gear and the engine races giving poor performance again. MPPT selects the correct "gear" to be in for the conditions at the moment, just like driver in a car with a stick does.

MPPT is a real time event processing sequence and requires a little bit of intelligence in the controller so what you are talking about with a bucking converter doesn't do that, the 7210 solution does although maybe not optimally so.

Use the 7210 to charge your 7s string and then the bucking converter to run a bunch of 12V lights...
 
The 7210 will work from using 18V panels to charge a 24v bank. I would go with that solution all days if you ask me. I got 3 of those controllers so far for just playing around and they work just fine for such a setup.
 
I have a BQ24650 board too it's good for low voltage banks but that's it. You can get 10A from it with heatsink, but solar voltage input limited, 33V abs top. Ohh and it's basically a buck so input voltage should be bigger than the output.

I think no one argue about the 7210's price. It's cheap as hell, 30 bucks or something and for a starter it's a best bet imho.
 
Its enough if u have decent balanced packs from start in terms of capacity and self discharge
 
If you start off with all your cells charged to the same voltage and 66mA can't keep them balanced then you have a problem in the pack and should be troubleshooting that.

Bear in mind that with the sort of BMS you linked to the high cell is discharged until it's not the high cell any more and then the new high cell is discharged and so on until they are all at the level of the low cell, if you have a leaker that can cause you a lot of lost amp hours over time as all the rest of the battery is dragged down to match one leaky cell out of a great many.
 
Thanks, guys. I do not plan to run it 24x7 - I will put individual LED voltmeters on every pack so I can easily check the voltages and if I see the whole pack has to be balanced - I will switch thins thing on.
 
So this is Ver. 2 of the system I want to build:

1) 2 x 80W solar panels (preferably monocrystaline) in series - I will put them on my balcony - the space I have is long and narrow;
2) Steca Solsum 6A PWM solar controller - these are made in Germany and as per the datasheet have some advanced PWM functions (whatever this means); eventually will get a MPPT controler;
3) 7s12p pack for 24V system; currently I have about half of the cells needed, and I still continue to source new ones; all of them are in the 2000-2300 range and of several different brands, and I will spend some time matching the capacity of each pack as closely as possible; I also plan to put the highest capacity cells near the busbar terminals, the busbars will be either 6mm2 or 8mm2 solid copper, each cell will be fused to the positive side;
4) cheap 1000W 24V to 220V inverter (still need to get that), later on I may get something better;

This will be the initial setup, I will continue to gather cells and upgrade the capacity with time.

What do you think? Any suggestions?
 
Also - I have some concerns about the charging mode of the Steca Solsum controller - the datasheet specifies the max voltages for 24V battery as:

End of charge voltage - 27.8V
Boost charge voltage - 28.8V

The 7s pack will be 29.4V when fully charged. The controller is not designed for Li-Ion batteries and I am a little concerned about the float charge... Are these voltages above harmful to the pack? Is this controller OK for a 7s lithium pack or I should get a new one?
 
You're assuming you're going to charge to 4.2V (4.2 * 7 = 29.4V). This is not advisable as this shortens the life of the cells drastically; by 1000's of cycles.

What you should do is charge to 4.1V (4.1V * 7 = 28.7V). That's a lot closer to your charge requirements of you controller. And it will give loads of life out of the cells.

Also, if you can set the controller for cut off, set the cutoff for 3.8V/cell (26.6V), or no less than 3.5V/cell (24.5V). Again, this drastically increases the life span of the cells.

That's why you see lots of cells in packs on here. We're trying for longevity as well as capacity. If each cell only delivers .5amps, that increases life, and if you can drop it to even .3amps or so, even better. A lot less likely a cell with go bad; or at least catastrophically.
 
The 7s pack is max 29.4V, the controller charges max at 28.8V, so I don't think this will be a problem - that's 4.11 per cell/pack. The 29,4V will never be reached in this config - it was just for reference. As per the controller's datasheet:

Deep discharge protection (SOC / LVD) - 22.4 V (23.2 V)

The first one is listed as nominal, the second is listed as lowest... not sure how to interpret that. The controller does not offer any modifications - it is a basic one. If I take the lower of the two (22.4V) - that's 3.2V per cell/pack. Is that OK?
 
You're probably fairly safe in the order of charging/discharging, then. The 7s string would be fine with that inverter. I would try to make sure you stop the draw before the 3.2v/cell, though, if possible.

But you could put in place a circuitry that would cut power to the inverter at a set voltage drop. A few resisters, diodes, transistors and a couple mosfets (or more in parallel) and you can build a safety switch. I'm not good at circuit design, but I can come up with the ideas ;)
 
Thanks - I can breathe now :) I will use that controller in the project and see what power I get out of the panels. If I am not happy with the result I will get one of these Chinese MPT-721A controllers and see what I can suck out from the panels.
 
Korishan said:
You're assuming you're going to charge to 4.2V (4.2 * 7 = 29.4V). This is not advisable as this shortens the life of the cells drastically; by 1000's of cycles.

What you should do is charge to 4.1V (4.1V * 7 = 28.7V). That's a lot closer to your charge requirements of you controller. And it will give loads of life out of the cells.
Click to expand...

Even better charge only to 3.93V ... this table shows aprox how many discharge cycles before the battery dies ...every powerwaller should have this pinned above his bed....


image_lzymrc.jpg


At 3.93V you get four timesas many cycles as you do going to 4.1V ... although slightly less capacity if charged only to 3.93 , still over 3 times energy handling in a batteries life,compared to charging to 4.1

3.93 x 7 = 27.51 .... still on the high side for electronics designed for 24V lead acid.
 
You must log in or register to reply here.
Back
Top