diy solar

diy solar

New Supplier for 280Ah cells: Shenzen Basen

LFP batteries produce very little heat when charging/discharging. For example my 8 cell pack with 1/2" foam insulation all around only gained about 10F over ambient after a complete charge/discharge cycle at 0.2-0.3C.
 
I see compressing the batteries between large boards. I wonder if it would help to put some small holes in the boards and perhaps very thin separators between the batteries with slits as well and/or with copper to draw out the heat to keep them cool during use/charging. Maybe CPU type vertical heat-sinks with dual sides for pressure.
Would trying to reduce temps of batteries with cooling per cell extend their life?

From that table, it looks like a fan would greatly increase AMP capacity of at least the bus-bars by 15a. Keeping batteries cooler could only help that also.

I went down this thought-path a while back. But when I looked into it, my takeaway was similar to what @Luthj said above.

Passive and Active cooling measures are not necessary or very useful at low C-rate. Certainly not under 0.5C and maybe not under 1C continuous. At < 0.5C cell temperature should be within a few degrees of ambient. With these small temperature differences heatsinks or even active air cooling can't achieve much.

Where a cooling system of some sort does become more important is higher intensity applications like EV's. Where charging/discharging at or above 1C is commonplace

See here (and click the blue spoiler tag)
And Here

I think anything you can do to keep your batteries at happy temperatures will extend there life. Keeping cells within the happy temperature range (ideally ~10*C-30*C) and keeping the terminals cool would be of benefit. But I think passive or active air cooling won't be the most effective path, beyond maybe exhausting hot air from the compartment if there is heat generating equipment located nearby (like an inverter for instance). I think locating the battery bank near thermal mass could be of benefit as well (in a vehicle maybe near the water tank, in a home or garage, in contact with a cement slab floor for instance. Thermal mass has a moderating effect on temperature, and helps even out the high and the low. If your busbars are heating up significantly there would be some benefit to pushing air over them, but the proper solution would probably to use larger busbars that stay cooler. Those are my thoughts at least, mostly just speculation.
 
I see compressing the batteries between large boards. I wonder if it would help to put some small holes in the boards and perhaps very thin separators between the batteries with slits as well and/or with copper to draw out the heat to keep them cool during use/charging. Maybe CPU type vertical heat-sinks with dual sides for pressure.
Would trying to reduce temps of batteries with cooling per cell extend their life?
I think that you should find a cool spot for the batteries (mine will be in an underground concrete cellar : 15-20°C), I've seen a guy from Australia that was having some hard time with temperatures in his shed. In this kind of case there are solutions, for example what we call in my country "puit provençal ou puit canadien", a ground coupled heat exchanger.
I've seen Will Prowse using an air conditioner in his shed, which I found pretty.... irrelevant, funny but not smart way of cooling.
Keep batteries over 10°C and under 30°C and you will get the best of them, do put high charge rate when they are cold.
 
This needs to be mounted in my garage which doesn't have A/C and we get outside temps up to 105 in the summer. I'll never be charging at that time however. Is discharging ok during higher temps?
 
This needs to be mounted in my garage which doesn't have A/C and we get outside temps up to 105 in the summer. I'll never be charging at that time however. Is discharging ok during higher temps?
I have a similar situation and I will use a temperature sensor so the battery won't charge or discharge if the temperature is too high.
But I also thought I could get a portable air conditioner and point it right at the batteries for the occasions when I need to charge when it's really hot.
Sorry I don't know the answer about discharging at high temps.
 
Here's a study about LifePo4 charge/discharge effects from temp.. but the study was from -40c to 20c !! :oops: warmer is better.


It was found that the temperature combination for charging at +30 °C and discharging at -5 °C led to the highest rate of degradation. On the other hand, the cycling in a temperature range from -20 °C to 15 °C (with various combinations of temperatures of charge and discharge), led to a much lower degradation. Additionally, when the temperature of charge is 15 °C, it was found that the degradation rate is nondependent on the temperature of discharge.
 
Last edited:
Here's a study about LifePo4 charge/discharge effects from temp.. but the study was from -40c to 20c !! :oops: warmer is better.


This last paragraph is interesting
 
What awg are you using for your 16cells pack...?
Awg 4....2....? let's say for 3m of cable.
 
I have 1 cell I may have a problem with but I'm not sure yet. I had my batteries built into 24v packs just getting the bms and wiringnall sorted out while waiting on a new power supply. They were not top balanced altho all read 3.295v. when the power supply came in I put it on 28.8v and started charging and the bms was reading one cell running 3.8v
I disassembled the pack put the all in parallel and started to charge the fro top balancing. I'm only home on weekends this is going to take a while. But all week when I'm away they are balancing themselves also. I never did see 3.8 on my multimeter so I'm not sure what to think
 
What awg are you using for your 16cells pack...?
Awg 4....2....? let's say for 3m of cable.
It depends on how much current you plan to draw.
If you are going to run 1C charge/discharge on a pretty constant basis, its a different answer than if we're talking maybe 0.5C maximum.
 
It depends on how much current you plan to draw.
If you are going to run 1C charge/discharge on a pretty constant basis, its a different answer than if we're talking maybe 0.5C maximum.
Yea I know, I was curious of what people use and which rate will they use their pack (16s).
I plan on 0.5C, my inverter is 5kw only and perhaps 1kw on my dc circuit ( connected to the batteries directly) that's a total of 120A max. I was planning on AWG 2.

Hooo what cheap charger can you guys use ...? I will only use it once.... A Riden..? This one could perhaps be off use later...
 
Yea I know, I was curious of what people use and which rate will they use their pack (16s).
I plan on 0.5C, my inverter is 5kw only and perhaps 1kw on my dc circuit ( connected to the batteries directly) that's a total of 120A max. I was planning on AWG 2.
I think your math is slightly off. Mine may be a bit on the conservative side but the formula I use is:
[Total Watts] / [Inverter Efficiency] / [Low Voltage Disconnect Voltage]
So for example..
[5000W+1000W] / [0.85] / [48V] = 147A maximum

Hooo what cheap charger can you guys use ...? I will only use it once.... A Riden..? This one could perhaps be off use later...
I think the Riden makes some sense for people who will only use it a few times. If you buy the DC-DC portion of the Riden and a separate meanwell AC-DC PSU one or both components can be repurposed in some way.
 
I think your math is slightly off. Mine may be a bit on the conservative side but the formula I use is:
[Total Watts] / [Inverter Efficiency] / [Low Voltage Disconnect Voltage]
So for example..
[5000W+1000W] / [0.85] / [48V] = 147A maximum


I think the Riden makes some sense for people who will only use it a few times. If you buy the DC-DC portion of the Riden and a separate meanwell AC-DC PSU one or both components can be repurposed in some way.
It's not off, you misread, the 1000w are on the dc circuit ( no inverter) only 5000w are on the ac 220v circuit. (I got 2 circuits in my project, one ac 220v, one dc 48v ( lights, water pump, phone/ pc chargers)
And the 0.85 efficiency is pretty low... Inverters got low efficiency at low power, typically under 30% of their rated power, we are looking for max draw, max power > efficiency over 90%.

Typical-per-unit-efficiency-curves-for-grid-connected-solar-inverters.png


Yea you're right, perhaps I could do with the Riden...
 
And the 0.85 efficiency is pretty low...
Its not low as a generic, conservative ballpark estimate. Many/most people on this forum are using cheap off-brand inverters with efficiencies in the 80's. My approach is to always default to conservative assumptions if I don't know the details.

But yes, if you are using a quality inverter, and they publish an efficiency curve, use that value instead (I would still subtract maybe 5% to be conservative).

Calculate how you will, my broad point is you should account for inefficiency

It's not off, you misread, the 1000w are on the dc circuit ( no inverter) only 5000w are on the ac 220v circuit. (I got 2 circuits in my project, one ac 220v, one dc 48v ( lights, water pump, phone/ pc chargers)
I did consider that but I assumed you would be using a DC-DC converter (48v to 12v). A wrong assumption it seems.

But even assuming 100% efficiency for DC and 91% for AC
[5000W / 0.91] + 1000W] / 48V = 135A


But regardless of whether we take your calculation or mine, 2AWG seems like it would be reasonable (considering both Ampacity and Voltage drop). 1 or 0 seem reasonable as well.

105*C rated 2 AWG wire (ampacity: 210A at 30*C, 178A at 50*C).
 
Its not low as a generic, conservative ballpark estimate. Many/most people on this forum are using cheap off-brand inverters with efficiencies in the 80's. My approach is to always default to conservative assumptions if I don't know the details.

But yes, if you are using a quality inverter, and they publish an efficiency curve, use that value instead (I would still subtract maybe 5% to be conservative).

Calculate how you will, my broad point is you should account for inefficiency


I did consider that but I assumed you would be using a DC-DC converter (48v to 12v). A wrong assumption it seems.

But even assuming 100% efficiency for DC and 91% for AC
[5000W / 0.91] + 1000W] / 48V = 135A


But regardless of whether we take your calculation or mine, 2AWG seems like it would be reasonable (considering both Ampacity and Voltage drop). 1 or 0 seem reasonable as well.

105*C rated 2 AWG wire (ampacity: 210A at 30*C, 178A at 50*C).
The 1000w are a large estimation on what I will use on this circuit.
2awg it will be, I do not plan on using massive power anyway, î will certainly never reach 5 kw. And the bms will be 150A.

Did you heard on someone having done a "switcher" based on arduino..? I would like to connect directly some of my PV to my water heater ( I plan on using 3 resistive elements in there, with 3 different power and resistance that match my panels)... Arduino will calculate which resistive elements to switch on/off and it will also reconnect pv to the inverter in case power is needed ( energy going out of the batteries).
 
  • Like
Reactions: Dzl
Did you heard on someone having done a "switcher" based on arduino..? I would like to connect directly some of my PV to my water heater ( I plan on using 3 resistive elements in there, with 3 different power and resistance that match my panels)... Arduino will calculate which resistive elements to switch on/off and it will also reconnect pv to the inverter in case power is needed ( energy going out of the batteries).
I highly suggest that you look at the electrodacus SBMS0 + DSSR20 w/ Diversion (either to purchase for your system or for ideas on how to implement direct connecting PV to resistive heaters. This pair of products is designed to do what you want to do. I am not sure if it is compatible with your inverter or system though.
 
I highly suggest that you look at the electrodacus SBMS0 + DSSR20 w/ Diversion (either to purchase for your system or for ideas on how to implement direct connecting PV to resistive heaters. This pair of products is designed to do what you want to do. I am not sure if it is compatible with your inverter or system though.
I know Dacian, we both got a lot of discussion about is dmppt and how it was kind of unusable for other people then him, system was not flexible to satisfy different situations.
The dssr is better but sbms only take 8 cells.

In a way my system, on paper, look like the dssr but just not as pro cause I plan on using dssr and Arduino to pilot them, I'm in computer science and not electronic.
 
  • Like
Reactions: Dzl
I know Dacian, we both got a lot of discussion about is dmppt and how it was kind of unusable for other people then him, system was not flexible to satisfy different situations.
The dssr is better but sbms only take 8 cells.

In a way my system, on paper, look like the dssr but just not as pro cause I plan on using dssr and Arduino to pilot them, I'm in computer science and not electronic.
Well I am excited to see how your arduino based system develops
 
Just got notified my 8 cells were received by FedEx. The following is the sequence:
22 September - completed order
25 September - Fedex tracking received.
12 November - received by FedEx at Los Angeles
17 November - scheduled delivery at my home in Sonoma.
The unknowns in this process are; how long after Fedex tracking number was received before they were in a container on a vessel and; when the vessel sailed and arrived in Los Angeles. Usually cargo is loaded and unloaded within a few days. The other unknown is how long it took customs to clear these two boxes and release them to Fedex.
 
Back
Top