In my limited experience even a 400a peak pack wouldn’t need anything this burly BUT this is eskate specific. Might be able to make some monster motorcycle or car packs and repairs. So much is limited by the total vehicle and passenger weight - only so many watts you can put down in eskate before you take flight
Im super happy with my kweld but getting some serious tool envy here
Well the reason I want to go with 0.3 copper is that I have two very unconventional battery builds with new tabless cells that have high promises.
The first one is a 20S1P pack for my jumper / offroad mountainboard. The plan was to shave off as much weight as possible using BAK 45D tabless cells, which advertise 60A temperature limited constant and 126A pulse ratings. What makes these ratings actually somewhat believable is their roughly 3.5 mOhm internal resistance. I do plan to push this 1P pack super heavily. I’ll be monitoring temperatures during testing, but I’m hoping for at least 90A pulses. At that point, 0.2 nickel would contribute significantly to the total internal resistance and heating of the pack. The 0.3 copper should be equivalent to about 1.2mm thick nickel.
Then next up I’d need a new battery for my raceboard. The current one works fine but I can max out the throttle which means it’s too weak There’s two things to consider for this one. Firstly I want to push more than the current 13kW (21S 180A with sag calculated in). I can go up to about 270A keeping the same power distribution front to back as I currently do, so that’s the target power. The second consideration is that I want to fly to esk8con (assuming I’ll be actually be able to earn enough money to afford it). What this means is that I need a flight safe battery consisting of less than 99Wh modules, or figure out the battery situation there which I don’t want to do unless it proves necessary.
This is the unfinished version of the model, I still need to figure out the top with not a ton of vertical space to play with. I should be able to hopefully barely squeze in 12 of these modules into my pelican storm im2050. That would mean I max out at 20S3P.
And this means two things. Either I can get that 90A per cell to work, for which I definitely need copper, or I need to build two batteries, one for flying one for racing in Europe. Or perhaps build some modules to be used under the deck, while having some modules over the deck. But this is more complexity than I want if there’s a way to avoid this.
Given the complexity and uncertainty of getting such a modular pack to work well first try, I decided that the jumper’s 20S1P is going to be built in this same way to test out smaller scale if this is a viable idea to pursue further.
So there you go, two esk8 packs desperately in need of copper
Ok, looking for a second opinion and some additional insight incase i have forgotten something important here.
The story is an older Lacroix loanstar pack using the sanyo 20700 cells had some issues. Took the bad p group apart and it was broken nickel and early enough it hadn’t done anything but disconnect. As soon as it was showing something odd it got put on the workbench so im confident the cells didn’t get abused. Also the breaks were series connections and not parallel connections so everything was still nice and balanced still and no lopsided drama.
because of the style and availability of replacements it doesn’t make sense to make a new pack at this time. Probably has a good 3k miles of use left in the cells easy. They stabilize after a charge at around 4.16 and stay balanced well enough all the way to 3.2ish. Sag is ok but pretty typical for this cell and not bad enough yet to throw a red flag on test rides.
The question: so I decided it needed a thorough once over to make sure everything is in good shape before i let my friend back on it - this means pulling the shrink and inspecting the nickel to make sure all is well and i find 4 cracked spots sprinkled throughout the groups. Some partly 2 fully broken. Im going to repair them but im also lazy and have some old dead packs in the same style and manufacture methods to test some things
I avoid sandwiches like the plague but I don’t think these cells will take to getting Dremel treatment based on a few experiments on dead cells from my recycling pile. Adding welds to the existing nickel and then welding a new 10mm .2 strip on top makes a very solid connection and tearing tests are good.
What have i not thought about here as far as side effects or potential issues? All the cells will get silicone to prevent any future shifting to keep from stressing the nickel for future bumps and vibrations. Seems like it will work well and be the best solution that lets it get to the cell end of life instead of diying an early death due to poor initial construction. Whats the group think
Is there a place to buy a Malectric welder from within the US? I’m being asked for a Christmas list, but it’s going to people who I’m not sure will be comfortable buying from somewhere they don’t know and overseas. I feel like I recall seeing someone selling them in the US once, but I don’t recall where.
What temperatures are your batteries reaching when you push them hard for much of their capacity?
I ran my 10s2p battery pretty hard yesterday, and charged at about 0.7c during 2 stops via portable cbarger and mostly full throttle towing when rolling.
Noon, sunny day Florida. Ambient was about 27C.
When I got home I quickly benched my board, dropped enclosure, and with IR temp gun, hunted for the highest temperature I could find on the exposed portions of battery heatshrink, and found ~55C.
I pulled up the data sheet on my 15 amp cells and it says 60C is max operating temperature.
Being able to read 55c through fishpaper, 0.5mm G10, kapton tape and heatshrink, pretty much means my cells had to be over 60c.
I am kinda bummed I can’t safely push this battery as hard again, but am pondering methods and products to better wick heat away from cells in future builds, such as thermally conductive foam, instead of insulating foam padding between battery and enclosure.
Also wondering how much less heat a lower resistance cell actually generates.
Agreed…probably wayyyyy over 60°C since it was a while since you were discharging them and you can’t read the hottest cells in the pack (in the middle).
IMO, that will make very little difference.
It only has a slightly lower thermal resistance, you can only pull heat out of the pack from the outside cells through the insulating fishpaper/heatshrink, and the heat is left inside the enclosure.
You would need to open up the pack and air cool it to make a decent difference…which isn!t practical here.
Your idea of using low IR cells is probably the best approach. EVE 40PL or Ampace JP40’s would be the best choice.
I am able to measure 5 to 7.5 f increase on thicker than average fiberglass enclosure directly over battery, compared to just ahead of the battery. But this is 35f below what I measure on the battery itself.
There is about 10mm of soft squishy neoprene foam between battery and enclosure. Not sure how thick it is when compressed , when battery is strapped to enclosure. It is a single stack, in firm contact.
I went for another roll earlier, a little slower, no charging whilst walking through park, cruised not quite as far, but full throttle the last mile when I knew i had enough juice left to make it back home no worries.
Immediately Slapped board on workbench, dropped enclosure, Aimed IR gun all over battery hunting for highest reading, found 55C. Again.
Motors and esc heatsink temps well below what I have gotten them to in the past
I have a K type thermocouple I am going to tape within a blob of thermal grease on IR indicated hottest spot, try to get some real time readings.
Data to perhaps justify 20 JP40’s
This 10s2p battery has basically been treated like this or worse, since January.
It has at least 2500 miles on it, perhaps 3k.
Still seems to have most of its capacity, and grunt when above 38v, and is never more than 33mv out of balance when I check.
Wish I had some Battery temp data when new, to compare.
My K type thermocouple largely agrees with IR temp gun over two cycles dropping enclosure immediately and measuring as quickly as possible in fan free zone.
The bottom of center of battery is consistently 1.5c cooler than the highest temp I can find on top of battery, with either tool.
The top of my battery has narrow strips of same squishy foam under many velcro cinch straps, and in current set up there is two layers of uncompressed harder foam totalling 12.76mm thickness before the bottom of wooden deck. There is some open air space shared with externally heatsunk esc, and not all that much contact between squishy battery foam strips on top, and full size double gaskets.
So it appears my squishy neoprene foam under battery is conducting some battery heat to enclosure to ambient but hard to quantify precisely. lots of variables.
The foam cradle is probably thicker than it needs to be.
Wicking heat from top of battery, and or protecting it from ESC heated air, might also help, but likely nowhere near enough.
But it seems my battery is now , and has been getting way too hot when I run it from 4.2 to 3.4 volts per cell, and ignorance was blissful, but increasingly dangerous, and I need to build a better one.
Just spitballing here: Seems like this is likely a good indication for why the cheap esc’s are going bad. Sounds like they are drawing more current from the pack than they can sustain and getting overheated. Thermal performance is inadequate even with significant heat sinking to get the heat out of them. Those cells in parallel are probably doing north of 30a each to get those temperatures. Probably best to plan on 3p minimum for future builds or plan on lower performance. Unless you get your voltage higher your required wattage at the motors is going to demand it imho
I did swap in my uncycled 10s1p P42a, and I can’t believe how much more torque there is.
I went a few miles solo, then picked up my dog in her chariot for several more miles at speed, and both solo and with copilot, the whole roll, there was more torque, and re engaging the throttle from coasting there was just more lively grunt and response.
Left at 41.98v,
battery temp 30.3c
Returned to 35.29v
Battery temp 48.6C
Ambient is 27.5 f, overcast.
I only had one esc partially fail on the DMEGC battery.
But I could not get the vesc working proper on it.
The 62$ nuclear cockroach Ebay Lingyi lives on, and battery radiates way more heat than it, when i first lift off the full size foam gaskets.
I dont wanna return the dmegc 10s2p pack, but this 10s1p p42a pack is for my other in progress enclosure.
Need to finish that, and then cinch strap it to this board I guess, and make a new 10s2p battery with low IR cells…
editadd…
Went for another roll with new 10s1p P42a
41.86v and 35.1c battery temp start.
33.56v and 52.2c finish at our highest average speed, likely ever.
I wonder how many cycles you have on that bad boy? Is it just old and an aged p group? Everything else looks so tight i also wonder what the top of the charge balance looks like. Sucks when a pack goes down like this - 11p fix?
I have a cheap welder and Zee 3s 5.2ah weld Lipo, and can weld 0.1mm Copper with 0.1mm Nickel plated steel atop, without issue, using 75 to 85 ms pulses.
Not including the Nickel plated steel cap, it should be equivalent to 0.4mm Pure nickel.
I could not reliably weld 0.15mm copper, or 0.1 copper with 0.15 pure nickel atop, but wonder if adding a second Lipo in parallel, whether it could.
0.15mm pure nickel requires 35 to45 ms pulses on same welder, but copper makes me eggplant, so copper I will use.
I say go for the fattest conductor your welder can handle with a little headroom.
