Be cautious ordering generic blue wrap LFP cells

hour

hour

Observer
john61ct said:
github.com

smagicld/xiaoxiangBMS

iOS app for Xiaoxang BMS. Contribute to smagicld/xiaoxiangBMS development by creating an account on GitHub.
github.com github.com

github.com

BotoX/xiaomi-m365-compatible-bms

Xiaomi M365 compatible, open source firmware for ATMega328p+BQ769x0 BMS - BotoX/xiaomi-m365-compatible-bms
github.com github.com
Click to expand...

Yeah I posted screenshots from the xiaoxiang top link. If only he included source code on the github instead of just advertising for his app.

I'm reading the data over serial but accomplished reading it over bluetooth as well with an ESP32. I would have been done with this a while ago if I hadn't been so obsessed with getting stuff from the MPPT also, but it's invaluable for the task at hand.
 
hour

hour

Observer
Yea, shoulda known cheaper heating strips was a bad idea. Wired them in parallel and plugged in to a 12.22v brick, one curled up and died almost immediately.

The other three I tested for a long time stuck to 1/4" aluminum plate. They got extremely hot, as did the aluminum, so I figured that'd do the trick.

Not wanting to disassemble everything for better heater placement, I stuck the plate on one side of the can and let the thermostat code do its thing.

It took like 45 minutes to go from 68 to 70 degrees, drawing ~38w (12w per strip x 3) but I don't think it's because of the lack of heat, I think it's the fact that a static source can't easily warm an entire area without making the area it resides in 9000 degrees.

So for anyone reading this down the road, build a platform for your cells and slip a heat pad underneath. There are suitable options for an ammo can in the < 100w range and suitable dimensions, but they all seem to come from the orient. Literally couldn't find one with shipping quicker than December in 12v and < 10 amps (for the sake of my relay) that was the correct size.

So, what to do now? I found some chintzy 60mm heaters - literally a 60mm fan on a radiator that heats up. The 100w advertised units are typically 50w from reviewers, and the 150w advertised no one mentioned on actual draw. So I ordered one of each - if the 150w is 50% its rating it'll work fine. Will probably mount on one side of the can blowing to the other side (probes are all on opposite side of heater)

Since I'm tracking amps in through the MPPT and overall amps either in/out from the BMS, I realized I could do some calculations that a BMV won't do (despite the VE smart network stuff). This obviously only works if charging through MPPT, but I can track what's coming in, what's going out, and then what's either going in or out of the battery. I added a time til full / empty reading which is dead accurate when charging from a power supply as PV, as one would expect. Gray when discharging, blue when charging.

Long press on temp row brings up popup to pick when the heat operates, or disable completely, or run manually for 1 hour (which I'll use a lot in testing)

Long press on max pack|cell voltage box resets statistics on the microcontroller. Local toggles to remote and the other way around, so I can pull data from either source. Local refreshes every 3 seconds, remote every 20. Still a lot to do on it in the future but I'd rather just get to testing. Damn shame about the heating failure or I'd be getting out of town this weekend for an inaugural run.

IMG_5134.PNG
 
J

john61ct

Adventurer
The Chinese seller of silicon heating pads on eBay does custom orders, not expensive.

Specify watts and voltage, LxW in cm,

max or min C° for a built-in thermostat.
 
hour

hour

Observer
shade said:
Have you considered a silicone pad heater?
Click to expand...

john61ct said:
The Chinese seller of silicon heating pads on eBay does custom orders, not expensive.
Click to expand...

Yeah I forgot to put the silicone in front of heat pads in my post above and I couldn't find a thing from any seller that would arrive before practically 2020. Like a 5x5 is perfect but not when it's 175w. A 4x7 50w is perfect but not when it's 120v. A 75w 12v is perfect but not when it's 18x12. Which left the slow boat stuff.

I don't think there's a way to intelligently use either silicone heat pads or any type of heat strip(s) other than placing beneath the cells (with headroom, maybe alum floor for cells to sit on and pad an inch beneath the floor). And heat source can't touch the can unless you're hoping to turn it in to a giant heatsink.

I'd have to disassemble everything and move BMV shunt and some other stuff to accommodate a false floor so I'm hoping I luck out with one of the two incubator heaters arriving tomorrow.

61Yv6vOjReL._SL500_.jpg
 
shade

shade

Well-known member
Diffusing the heat will be key. If the incubator heater doesn't work out, multiple low watt silicone heaters may be worth a try.
 
hour

hour

Observer
shade said:
Diffusing the heat will be key. If the incubator heater doesn't work out, multiple low watt silicone heaters may be worth a try.
Click to expand...

Think it'd still have to go beneath the cells. Can't fit anything on the sides of them unless stuck directly to the wall of the can which would waste most of it. If I add any insulation to the sides then I won't be able to clear the sides of the cells-zero diffusion. The incubator fan still isn't ideal, it'll be blasting a wood panel that's 99.8% the width of the can, and partitioning it 3/5ths of the way up. Hot air will go upwards and ********** my router and charge controller, then drift over the top of the cells. I might try to make a slant of reflectix that sort of partitions those two things off from the direct rise of heat (heater will be in bottom of can in the void, controller and router are in the top area of void)

May be able to smooth it out with software - if heater on time exceeds 3 minutes or so then stop for a minute. When that minute is up and fresh readings come in, re-evaluate. Still not the perfect solution as it'd be toggling my relay more than necessary. Hindsight 20/20. Will update tomorrow
 
luthj

luthj

Engineer In Residence
My suggestion would be to wire several higher wattage silicone pads in series to cut the wattage down. 50W per sqaure foot is ideal. Another option is to find a 12V heating blanket, and just stuff that in the box.
 
hour

hour

Observer
shade said:
Prices are reasonable, too. Bookmarked.
Click to expand...

Are they, though? 12 watt, 12 volt 2x4" silicone heater is $70.. and I thought Wolverine battery heaters were expensive... Imagine the made-to-order units are even less bang for your buck.

The 150w incubator heater arrived a few mins ago and was promptly connected to one of my portable LFP packs with a little TK15 coulomb meter.

Thing was pulling 194w. So that sucks, I was banking on it being half (or certainly less than) the advertised power like the 100w unit allegedly is. Maybe it had a problem and would have run away, melting to a pile of goop. I didn't let it keep going to find out. Wiring to the actual heating element was 14 gauge and adequate for all of its four inches. Fan was setup to push air through the heater instead of pull through. That's actually ideal for how I intended to mount this thing. Air movement was pretty weak but may have been adequate.. I'll never know.

100w that apparently pulls 50w is getting delivered later today.

I discovered these have separate leads for the heating element and the fan which isn't the worst thing even though I have just a single relay in the can to control heat. I'm having this idea that I could wire relay out -> digital heat controller board of which I have several, and wire the fan to the IN side of the thermostat board, too.

Means if relay is on then the fan would be on, and the thermostat would receive power. It'd be up to the thermostat at that point to get me to the appropriate temperature range, but overseen by the microcontroller operating the relay (also based on temps). If the thermostat failed open then the relay would still disconnect later when BMS + probe temps got hot enough. Additionally, if the actual relay failed open, the thermostat (if it hadn't also failed) would keep things in check.

Also, if the relay was on between 40 and 50 degrees and the thermostat was set to 45*, I'd get some air movement in the can before and after the heating cycle.

I'm not nearly as optimistic about this working anymore, so I'm considering john's idea of using some rigid high density foam insulation to surround the box. That'd probably give me a better temperature reading inside and lessen the chance of charging being enabled prematurely because the top of the box got hot while the bottom 90% of the cells were still 10*F. Thinking I could use these aluminum extrusions to lift the box up an inch - two pieces either on the long or short sides of the ammo can. Would allow for a nice sized heating pad to fit underneath. I could skip that additional step if I maybe used two silicone heaters, one on either side of the box and positioned more closely to the side the cells are on.. but why screw up another chance to heat from beneath?

This would be less effective than heating the inside of the can so I have no clue what wattage I'd pick for that approach. Probably best to go overboard and then rely on software cycling it (5 mins on, 3 mins off - or whatever) just so my junk parts pile doesn't grow any larger.

Edit: alum extrusions I was talking about. Can could sit on the little shelf and I could drill through the flange in to the can for attachment.

100-110-1.jpg
 
luthj

luthj

Engineer In Residence
A 12x12" 50 watt pad is about 170$. Not cheap, but not crazy expensive. The big thing is you can pick your wattage.
 
You must log in or register to reply here.

Forum statistics

Threads
185,911
Messages
2,879,534
Members
225,497
Latest member
WonaWarrior

Members online

Total: 867 (members: 9, guests: 858)
Top