diy solar

diy solar

Thinking about battery backup

Crimping on large gauge wires only works well if the wire has many fine strands. In many of the installations I worked on in my past, we would use those split bolt wire clamps like the picture Hedges provided. A normal 2/0 THHN wire only has about 19 very stiff thick strands in it. If it was a very high current line, we would sometimes use 2 or even 3 of those clamps along the wires being spliced together to increase the contact area. The lamp in a cinema projector runs at up to 160 amps (about 30-45 volts) and they run all day long. In some cities, they required the power supply to be in a separate room from the projection room due to old school film being flamable. We would have to run up to 4/0 cables to keep the resistance and losses low enough. Sometimes we would get a "brilliant" licensed electrician who would look up the wire ampacity and step it down to a 1/0 without telling anyone. Then we would find the lamps being hard to strike, and flickering once they have a few hours on them, and the electrical contractor was packed up and gone. He saved a ton of money putting in undersized wire, that legally handles the current, but the resistance screwed us. In one job, they also went down a conduit size, so we could not even pull in the right wire later. Thanks to digital projectors, the power supplies are usually in the projector, or the pedestal with less than 6 feet of wire supplied with the projector, but they still run a 6,000 watt Xenon bulb at 160 amps.

Working at high current on a lower voltage makes the wire sizing critical. At 120 volts, a 10 amp load is 1200 watts, and a 0.1 ohm resistance in the circuit will cause a loss of 1 volt or 10 watts, or less than 1%. Even at the same 10 amps at 12 volts, we would still just lose 1 volt, 10 watts, but this is only a 120 watt system, so the loss is 8.3%. And if we actually were trying to get 1200 watts at 12 volts, or 100 amps, that 0.1 ohm loss kills the whole system. 100 amps at 0.1 ohm is 10 volts, so just 2 volts makes it through to the load. Anything less than 120 volt, and the wire sizing really needs to take the voltage drop over distance into consideration.

In this case of a series battery bank, each block of cells is just a 3.6 volt source. At 100 amps, even a tiny bit of resistance can cause a huge loss of power. All of the connections need to be solid.

When I was getting my wire at Home Depot, they had a 25 foot pack of #8 bare copper grounding wire for like $17 You could probably run 4 or 5 of those down your strip between the cells, crimp them into a connector for a 1/0 wire, and then heat it up and fill the dead space with solder. If will be dead stiff, and trying to bend it will probably break out the solder. Even with 4 x #8, I would still connect at both ends of the cell group. At these currents, your 17 cell long row of 18650 cells is going to be about 340 mm long (13.4 inches) and the cells at the far end will end up pulling less current than the ones at the terminal. Connecting at both ends as has been suggested not only cuts the current at each connector in half, it also cuts the length from the furthest cell to the connector in half. The current and voltage balance across the 34 cells will then be much closer.


Thanks for the info GX.

I was talking with a buddy of mine who has me seriously considering switching over to Lifepo4 batteries instead. About the same price, easier to setup, and safer. Ill also end up with around 14kwh vs the 5kwh of this pack.
 
I looked at a bunch of options. The idea of stringing together a ton of 18650's was considered, but when you factor in the time you spend putting it all together, it really didn't look as good to me. I then looked into used Nissan Leaf cells, but they were already testing at just 70% capacity, which in my mind, says they are on their way out. Then I was very close to buying a bank of the aluminum shell LiFePo4 cells off of AliExpress. I looked at different combinations of 90 100 150 280 AH cells to make a 16S 48 volt bank of about 12 to 25 KwH's. The price point looked good, but I saw several reports of either old or over rated cells, so they had me a little concerned about getting the capacity I was expecting. But then I lucked out and saw the deal at Battery Hookup and snagged the three Chevy Bolt packs at an even better price per KwH than any of those. With shipping and taxes, I have almost 18 KwH for under $2,000 which is way lower than even the cheapest 18650's before all the labor. It did take some work, as I had to cut a buss bar in the 8S pack and make custom terminals to tap the middle cells. Then made mounts and wired up a BMS etc. but you need that for any battery bank. Battery Hookup did run out of these packs, but they told me they do get more in from time to time. They also have other packs that can work as well.

The complete BMW 1 series E pack is a bit of money, but it is also a lot of storage. They may have some of those left. They are listing the BYD 8S 24 volt LiFePo4 packs at a great price, but they are only quoting them as 50% capacity, so I have not idea how well they will age. I saw at least one guy complain that one of the packs he got came bulged enough to crack the plastic housing. What pushed me towards the Bolt packs was that they are basically new cells. Chevy rejected the modules because the assembly was out of spec so the robot assembly line could not install them in the car. They looked brand new when I got them. Evidently the bolt holes may have been just a fraction of an inch out of position.
 
I looked at a bunch of options. The idea of stringing together a ton of 18650's was considered, but when you factor in the time you spend putting it all together, it really didn't look as good to me. I then looked into used Nissan Leaf cells, but they were already testing at just 70% capacity, which in my mind, says they are on their way out. Then I was very close to buying a bank of the aluminum shell LiFePo4 cells off of AliExpress. I looked at different combinations of 90 100 150 280 AH cells to make a 16S 48 volt bank of about 12 to 25 KwH's. The price point looked good, but I saw several reports of either old or over rated cells, so they had me a little concerned about getting the capacity I was expecting. But then I lucked out and saw the deal at Battery Hookup and snagged the three Chevy Bolt packs at an even better price per KwH than any of those. With shipping and taxes, I have almost 18 KwH for under $2,000 which is way lower than even the cheapest 18650's before all the labor. It did take some work, as I had to cut a buss bar in the 8S pack and make custom terminals to tap the middle cells. Then made mounts and wired up a BMS etc. but you need that for any battery bank. Battery Hookup did run out of these packs, but they told me they do get more in from time to time. They also have other packs that can work as well.

The complete BMW 1 series E pack is a bit of money, but it is also a lot of storage. They may have some of those left. They are listing the BYD 8S 24 volt LiFePo4 packs at a great price, but they are only quoting them as 50% capacity, so I have not idea how well they will age. I saw at least one guy complain that one of the packs he got came bulged enough to crack the plastic housing. What pushed me towards the Bolt packs was that they are basically new cells. Chevy rejected the modules because the assembly was out of spec so the robot assembly line could not install them in the car. They looked brand new when I got them. Evidently the bolt holes may have been just a fraction of an inch out of position.

I saw those same BYD packs, but the capacity had me worried as well. Which chevy bolt packs did you get and what type of cell are they (lipo, lifep04, li ion)
 
I have two of the 10S 3P modules, and one of the 8S 3P modules. Each cell is an LG Li Ion NMC cell, so they do have some Cobalt. They are 3.7 volts nominal and 60 amp hours per cell.

In the Bolt modules, they are arranged in 2 cell holders, with aluminum cooling sheets between each pair. But then the tabs are welded in groups of 3 in parallel. So the 10S pack is 30 cells. The 8S pack is 24 Cells. The full positive and negative ends come out to a 6 mm stud for the output cables.

What I did was cut through the middle buss bar on the 8S pack, Each buss bar is held by 3 plastic rivets to the battery case. I made sure both halves still had one rivet by cutting it at an angle through the center rivet. Then I made a clamp to hold a copper tab against the buss bar that extends out and has a bolt hole for a 6 mm bolt. So it is now two 4S packs. I then wired each 4S half of the 8S in series with one of the 10S so I now have two 14S groups. I used #2 awg wire to make up each pack, and they join at a fused buss bar. I am using a single BMS to monitor the pack. Each cell monitor lead is going to two fuses and then to the cell junctions on each pack. If for any reason the 2 packs go out of balance, it will just pop the little fuse. I have 7.5 amp ones in there now. The Balance function can pull 2 amps on those leads. Here are a few picture of the battery bank.

Battery management.jpg IMG_2408.JPG IMG_2412.JPG
 
Man thats impressive stuff.

If i got 4 of those BYD packs and wired both sets in parralell and then series them together, i would be looking at a 48v ~250AH system, correct?
 
Back
Top