diy solar

diy solar

Need help! 48 volt caught fire

Here are my BMS settings
The over current trip time delay seems long at 10 seconds, however even 100 amps can cause issues with non fused under rated cables and poor joints. As pointed out above, the metal frame holding the cells may have played a part in the failure, current may have bypassed the BMS.
The pictures dont show where the other end or parts of the melted negative ended up, perhaps on the battery buss bars?
I think there is more than one fault that occurred to destroy the cells.

This and other recorded meltdowns suggest the following features should be followed in building a battery.
Use a containment method that isolated the cells from any metal, and fully protect buss bars and connections against accidental shorts.
Use a suitable fuse at the battery positive and further fuses/breakers in the systen as necessary with regard to the cable rating.
Test the battery under expected loads to identify possible poor connections before putting into service.

Mike
 
Last edited:
The burned red cable going to the battery is unsupported. Cables should be supported such that the terminal isn't carrying the weight of the cable.

Is that a contributing factor here? Possibly, but I would focus 90% of the blame on the converter wiring.
 
IEC 62368-1:2014 defines "safe" as 60VDC or 2mA DC, whichever is less. For AC the limits are 30VAC and 0.5mA. These are considered safe for an ordinary person but you still cannot leave exposed connectors/wires lying around as there's the whole chapter of "electrical fires" to consider.

It only takes about 30mA of 60Hz AC current through the heart to cause fibrillation, compared to 300-500mA of DC current. So the limits above are well below this.

The impedance of the human body at 50V is around 45kΩ. You will need to pass through the skin twice unless you really do something creative.

50V/90kΩ = 550uA.

So while 48V might make your fingers tingle a bit it is not dangerous in handling.

It certainly will make bigger sparks and the smoke genie is bigger then with 12 or 24V systems.
I promise you.
If you place 48V across your chest, you are going to have a very bad day. All it would take is slightly sweaty hands and you are completely unable to let go of 48V.
 
careful on those converters - they dont do anywhere near the rated power & go pop easily.
I used one on my solar battery charger & got no where near the rated current.
 
Hello, I need help I put together a 48 volt 3.2V 120Ah Lithium Iron Phosphate LiFePO4 battery pack. I had bms all hooked up everything was reading good I got 52 volts, i connected it to my growatt 3k all in one based off the blueprint on the mobile solar sight the only thing I had different was a 48 to 12 conversion box for my tong jack. I used a pre charge resistor no sparks when I connected everything then 30 seconds later sparks flying and one batter catches fire. The ground wire to the converter melted off, I’m not sure what I did wrong or where I should go from here are all the cells bad or just the ones that leaked? Where do you think I went wrong on my wiring. Please help!


First thing I see is that is a 24v box not a 48v.
 
Interesting photo here. https://diysolarforum.com/attachments/2f51bb4b-d0ea-42f7-8870-d9758e41c76b-jpeg.49255/

Ground wire is crispy while red wire + from busbar is not.

I take it the black ground wire burned all the way back to negative busbar?

Are you certain you have polarity at the battery correct? We had one here recently where the red lead of multimeter was in the COM port which is backwards. According to the burnt cell photo, it does appear correct where red wire is hooked to + terminal of last cell.

Very strange failure really. With the yellow converter wire to load and red from + busbar, the only real answer I can see at this point is load to high for the wires/circuit protection. You really need to get an amp clamp where you can see full load amps at motor startup. Either one that records max amps or a graphing meter. And you need circuit protection before the converter. Source a fuse panel and run a heavy gauge wire from busbar to fuse panel. This will not be cheap, a fuse panel like this one is rated for 32V max, not 48V. https://www.amazon.com/gp/product/B001P6FTHC/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

I'd prefer fuses myself for this situation but if you can't find a fuse panel, you might find a breaker like this one rated for 48V. https://www.amazon.com/Blue-Sea-713...l+48V&qid=1621294205&s=sporting-goods&sr=1-29

The issue with switches and breakers on 48V is the arc on disconnect can weld things together defeating the purpose.
 
I promise you.
If you place 48V across your chest, you are going to have a very bad day. All it would take is slightly sweaty hands and you are completely unable to let go of 48V.

The 'let go' current (above which you cannot let go of a grasped conductor) is 4 times higher for DC than AC. So even if you are freakishly low impedance and covered in sweat with the negative cable in your mouth it is far fetched that you are unable to let go.

I work around 12/24/48V systems all the time and in a marine battery room on a 100F/99% humidity day sweat is just lube to get you in and out easier. None of us treat 12/24/48 any differently and never a very bad day for anyone. There are no special safety requirements to do so.

 
Ground wire is crispy while red wire + from busbar is not.
His hand written diagram shows a connection to the vehicles 12v battery to the converter box. Maybe this was the main source of current that could explain the disproportionate damage between main red and main black to the 48v battery?
 
His hand written diagram shows a connection to the vehicles 12v battery to the converter box. Maybe this was the main source of current that could explain the disproportionate damage between main red and main black to the 48v battery?
You are correct, it appears there is a 12V battery on the load side of the buck converter.
 
what about this... he is the battery frame as ground... what if one the cells rubbed and added a positive to the metal frame. ... looks like they aren't protected
 
Sorry about the mishap. The drawing for the voltage converter doesn't have a fuse in it. It really needs one. In your picture it looks like you tried to use a circuit breaker, and (if I am seeing it right) it is all wrong. For some odd reason, the converter seems to use black as positive and yellow as negative. You connected black to the chassis/frame, which would be a direct short. You also connected the negative input wire to the same terminal as the output positive wire.
If you redo this, instead of a circuit breaker, use 2 fuses. Put one fuse in the red wire, and the second in the black wire that is the positive output. Don't cross-connect the input and output in any way.
That's right black is only ever a chassis ground in vehicles once you get to ac black is a hot lead. Yellow is commonly ground in marine applications I believe but if it's not green read the instructions. I've also heard those breakers are slow to trip but have never used them.
 
His hand written diagram shows a connection to the vehicles 12v battery to the converter box. Maybe this was the main source of current that could explain the disproportionate damage between main red and main black to the 48v battery?

That is what it looks like from the drawing.
 
I am brand new to the off grid and LiFePo battery environment, but I was in the trucking industry as a mechanic all my life (almost 40 years) and have seen a lot. I can almost guarantee and bet a steak dinner that Steve S. and others are exactly correct with their assessment. With the thin plastic cover around the cells, and direct contact with the metal frame there was a short. It would not take long at all for the road vibration to rub through the plastic cell cover, no matter how tight everything is. Metal flexes going down the road, especially true with something thin like Uni-strut. Build the metal battery box with something more substantial such as 1/8" plate and weld the joints. Insulate the bottom and sides with some thick plastic or rubber. Flat mud flaps make great material for this, (not the kind with astroturf to hold down rain spray). Use a non conductive battery hold-down clamp or at least put plastic under a metal hold down clamp.
Also use flexible cell connectors since the threads are so fragile on these cells.
If the batteries are installed inside the vehicle, I believe you would have more wiggle room with your design since the rubber body mount cushions would absorb some of the minor vibration, but bolting your battery box directly to the frame calls for extreme measures.
Wire insulation is also critical, especially when crossing the edge of a frame rail, going through metal body panels, etc. I always secured battery cables as much as possible against anything I could and double or triple insulated against edges with heater hose or similar. I have even seen plastic wire loom rub through the insulation on smaller wire harnesses, and then the wire would corrode when the salt spray in the winter has a chance to work on it.
I know this is a late post, but like I said, I am new here and this is my first post, but I wanted to add my 2 cents for what it is worth.

Also, I have learned a LOT from reading through various threads and obviously there are some very knowledgeable and smart folks here, and I hope to learn more! I am building my first set-up in the garage where vibration will not be an issue, but fire certainly can be.
 
Hello, I need help I put together a 48 volt 3.2V 120Ah Lithium Iron Phosphate LiFePO4 battery pack. I had bms all hooked up everything was reading good I got 52 volts, i connected it to my growatt 3k all in one based off the blueprint on the mobile solar sight the only thing I had different was a 48 to 12 conversion box for my tong jack. I used a pre charge resistor no sparks when I connected everything then 30 seconds later sparks flying and one batter catches fire. The ground wire to the converter melted off, I’m not sure what I did wrong or where I should go from here are all the cells bad or just the ones that leaked? Where do you think I went wrong on my wiring. Please help!
You must add a breaker. You connect all things with the breaker off, then you will turn it on. I hope it helps
 
You must add a breaker. You connect all things with the breaker off, then you will turn it on. I hope it helps
Something shorted. Are all the cells isolated from each other? The BMS should have shut off, or a fuse blown first. Always fuse and use breakers. Disconnect everything, and start over.
 
It might sound too obvious but. Use a multimeter on EVERYTHING to double check EVERY connection to make sure it is what it is supposed to be.
Also, like everyone here is saying. Use fuses for everything. I have a mobility scooter and I have added 5 fuses that were not there before. I am not taking ANY chances.

I hope this helps. :)
 
Back
Top