So it begins...... Battery spot welder

[rvpopeye]

Next gen cells sound great ! Tabless cells , but with tabs , makes sense to somebody I guess.
As long as they meet your needs and I'm sure anything that moves will make Fiona happy ! I look at bigger packs as longer range znd more torque for the hills etc. too, leave the insane top end to the young with rubber bones and elastic muscles.....

[sternwake]

I can't share the data, but the preproduction EVE 50PL was tested against the very impressive and
currently available EVE 40PL, at 40 amps continuous till empty.
Neither cell hits the 80C temperature cut, which is impressive innitz own right, but the 50pl maintains a significantly higher voltage, and runs for about 20% longer.

The preproduction cells usually dont perform as well as the production either. The preprodudtion Molicel P50b was the most impressive cell this time.last year, but can't compete at high loads with the tabless. The production versions at high loads stayed cooler, held more.voktage and delivered more mah.
They want 10$ for the p50b.

Molicel is gonna blow up the game when they release a tabless cell, hopefully soon.


No hills in florida to worry about, but having the torque desired when needed is nice.

But speed and torque are addictive. I got.to keep outta the throttle with my limited power now.

[sternwake]

One other consideration.of these cells are cycle life. Coming from lead acid this is always in the back of.ones.mind. Gotta get em.full ASAP, but its better with liion to not get to 100% whennone xoes not require full capacity.

The claimed cycle.life of Liion cells is usually something like 800 cycles, and they will have equal to or greater than 60 or 80% of their original capacity. Charged at X amps, discharged at x amps to 2.5v.

Usually these cycle life tests are using 100% of the capacity each cycle, and charging at a high rate, with only a short rest before next cycle begins.

Anyway, there is no standard to compare cells from different manufacturers and often even from same Manufacturer, nd all claims need to be taken with a grain of salt, but unlike Lead acid claims, liion often far outperforms the claimed cycle life, especially when charging slower and avoiding 100% and the bottom 10%.

The Tabless cells max charge rate are not as impressive as their max discharge rates. There are still some bottlenecks, which can cause the separator piercing Dendrites.

Like lead acid, above 80% ish state of charge the charge acceptance rate does slow down, but not nearly to the same degree. No 5 hour absorption stages like aged lead acid.

The Molicel P50b is not a huge step above the venerable p45b, in terms of discharge performance, but while the p45b says 9 amps is the max charge rate, the p50b can charge at 25 amps.

Max charge rate is usuallly about 2C, and this is always at the cost of cycle life.

My current battery, I usually charge at below 0.5c, 2.6amps. 1c charging is 5.2 amps, max 10.4.
My dumb Daly BMS has a 10 amp charge limit, 20 amp discharge. I bypass bms on discharge as it is too dangerous on esk8 to lose power and brakes.
My 5.5x2.5 barrel plug charge port says 10 amps max, but I am leery of this.

It has not been an issue, as I don't charge very fast, so far, but the BAK45d has a max charge rate of 12 amps. So in theory i could charge them at 24 amps. 800 to 1000 watts for 10s2p.

Rather insane.

I am limited to 10 amps with my BMS, but achieving 10amps would mean parallelling charging sources, which might not enjoy that.

When the new battery is complete, my old one will not have a job. It still can deliver most of its original capacity, at lower rates.

I am thinking of ways to safely parallel the old with the new if I want to extend my range.

Can't just hook a 42v battery in parallel with a 35v one. They need to be very close in voltage, and.of. course charged separately
I have some plans. Might not ever implement them, but the possibilities must be explored.

Time to go cormorant hunting.

[rvpopeye]

arf arf arf arf arf arf arf arf arf arf arf arf arf !
Have fun and take it easy on the shoulder.

[sternwake]

   

Little dogs go 'arf arf arf'
Fiona has the 'rrrrWoof!!' of a dog twice her weight.

[rvpopeye]

WoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoofWoof
I sit corrected. Very happy pup !
Ya know , she's singing a doggie version of On The Road Again.

[sternwake]

The  cheap red spot welder, I tried using my 2017 GC-2 AGMs as the welding source, and at a 13ms pulse it fried a transistor.  Too much battery I guess.  Outside the goldilocks zone.

So I ordered another purple, just the PCB, no wires or casing..

The cheap purple spot welder has a dual sided circuit board.
The bottom  could fit 5 more transistors, in theory doubling the amps.it could send to the welding pens, or halving the load on each transistor, but they should be matched, and that would be difficult and expensive.

I have 22 BAK45D Tabless  cells, the AO4 version, which I thought might be a lesser grade than the A02. Mooch tested them both and performance wise, that is not the case.  Cycle life, not sure.

Supposedly, one can discharge  one BAK45D cell at 30 amps, to 2.5v, charge at 8 amps till full, 600 times and the cell will retain greater than 60% of its original capacity.

My ESC cuts power at 3.2volts per cell or parallel group. 10 in series. 32v.

My BMS can only handle 10amps max charge rate, and I'd have to parallel charging sources to achieve that.
 My esc can only draw 24 amps max, so 12max per  cell.

So cycle.life should be far greater than 600, in my usage.

I determined a 83 ms pulse was achieving  Solid welds on 0.1copper, 0.1nickel.plated steel sandwic on a BAK45D.
3s 5.2ah lithium polymer battery had to be warmed up, and over 12.4v.

12 pairs of welds would have weld battery very warm, the 12awg downright hot and battery down to 12.05v, so Id recharge it after every 12, under a strong fan.

After 56 welds today, the welder quit.
   

The  copper traces right next to (-) input 8awg ring terminal fried and broke.

I bridged the break with some 1.4mm x 10mm   copper flatbar left over from some van project, and it works again.

My newly arrived backup PURPLE Welder PCB's 5 transistors, dont look to be counterfeit like my original.  I need to beef up the circuit from input to output before putting it into service, and see how.it performs, but hopefully don't need to.on this battery build.

I'd like to be able to use less than an 83ms pulse, and Id like to be able to use 0.15mm copper, on the next battery..

A more.powerful Lithium polymer weld battery, or a second one in parallel could likely do it.
There is also a   copper to copper welding flux which might allow it, on the next battery build.


Fiona and I have been rolling plenty, but the 10s2p DMEGC-26r pack is getting way too hot, even staying out of the throttle.

Been charging it from  a 7s2p battery through aV booster in the park, the  riding to next park, then charging some more, lather rinse repeat.    It just gets over 56C on the exterior, and is only rated to handle 60C.

The BAK45D tabless cell should stay far cooler.  Each cell has a 13.2amp max charge rate..

The EVE 50PL has been tested at 100 amps, and it smokes every other tabless cell at that extreme discharge rate.  Hopefully the production versions become available soon, and all this extra tax tarriff bullshit ends soon.

I have a new small enclosure made for a 10s1p battery,.originLky for a mini lightweight grocery getter build. I had made a molicel P42a 10s1p, which got submerged and ruined.  The P42a is a great cell, but I was able to get this pack to 56C  too, riding hard. The P42a is also 7year old technology.

A 10s1p of Eve 50pl would be quite the little workhorse.

As far as I can tell, discharging it to only 3.2 volts per cell, and recharging it from there, My Dmegc 10s2p still has most if its original al 5.2ah of capacity, but its resistance has increased and it gets too hot, on a hot day.  It has ~3k miles on it.  Seems a shame to retire it.

I could in theory parallel.it with the BAK, as a range extender.

I do have a 600 watt bucker.  I can step 36v nominal to 12v nominal and lose about 4-6% to heat, depending on the size of the  load

[rvpopeye]

The 50pl cells will step you up to a different level of rocket science , might have to always use buss bars instead of traces in the future lol. If you have to , pull a leg and solder wire directly to it , with a heat sink clip and very quickly of course .

[sternwake]

A real battery spot welder would be a good start.
Am lucky 20$ junk can even do what am asking it to.do.

There are premade buss bars that use mostly copper, and just have a thinner nickle tab for welding to the cell itself.

Soldering to the cell basically fries a portion of the electrolyte. It increases resistance, reduces performance, increases self discharge, to some degree.

Resistance Spot welding does too, but to a much lesser degree.

The Eve 50pl is a beast of a cell, and hopefully it is joined by some others as well.

[sternwake]

My Zee 3S 5200mah welding Lithium polymer battery started emitting a sickly sweet smell on the last few welds, attaching the main + and -  copper nickel sandwich, which had 10awg already soldered to them 

The smell is electrolyte seeping out from  one or more of the pouch cells.

The battery was still in near perfect balance, and behaving normally, but Lipo batteries are the most dangerous.  I put it on a 10ohm resistor for several days, then cut off the Xt90s and 12 awg, and shorted the leads,.so now the only danger is the toxic electrolyte.  Ill recycle it at some point.

The Bak45D 10s2p is basically done and installed as of tonight.  I took a quick solo test run on roads with wet spots, and there is more torque, but not as much an increase as hoped for.  The better comparative  test will be how fast Fiona and I can charge the crushed shell trail, with a 90F battery.

The DMEGC 26e 10s2p had 187 watt hours when new, and was rated for 30amps continuous, to 60'C

This BAK 45D 10s2p has 324 watt hours and is rated for 120amps continuous, to 80'C.

While the capacity is not double, the range might be, as the dmegc battery turned so much of  its capacity into heat

Each of the 10 parallel group of 2 cells  needs its own balance lead.

These leads should not cross over each other on an Eskate battery.  I went a bit further and spaced them apart, and held them there with black hot glue.

  Not shown, is dumb Daly BMS, and it is bypassed for.discharge. It is.pretty small, and can handle 10 amps charge, 20 amps.discharge.

If the BMS is wired for discharge, it has to handle more than the ESC can draw, and when it cuts out, no more brakes.  Hub motkrs have pretty good freeroll, but on belt drive or gear drive esk8, the BMS cutting out will basically be like slamming on the brakes unexpectedly. Dangerous.

The ESC cuts off at 32 volts, so there is no danger of overdischarging.  

NEVER TRY AND CHARGE A LIION CELL THAT HAS BEEN DRAWN BELOW 2.5 VOLTS




.    

So my bms is basicLly charge only, and will only start balancing cells above 4.18v...It uses tiny  resistors to bleed off the highest P groups to.let the others catch up, and if any Pgroup exceeds 4.25v, it disconnects the charginging source.

My Pgroups at 4.1, were within 0.015 of each other pre test ride

I set my charging source to 4.18 and 0.2 amps max for the initial top charge and balance.

I still need to finalize the battery install..so no final pics yet.