Load testing/discharging a custom battery pack

Not sure whats the situation where u live but u could try to find decent power tool li ion packs.

These should have high drain batteries inside of them, usually already spotwelded.

Those high power resistors are meant to be mounted to a larger metal heatsink for cooling. Your carbon cellulose slab is not an ideal heatsinking material :smiley: (You probably just wanted to attach them to something, fair enough)

But for reference, something like this. Might not even need anything huge size wise, if you add some fans for forced convection cooling.

1 Like

That’s basically a HiPOT, which is something i need. They’re used for load testing and discharging. Most people use lightbulbs now because they’re cheaper than these huge resistors that also require some kind of framing and have to be kept away from everything.

maybe there’s a fancy way to make a hipot out of lightbulbs. Maybe some kind of hilarious sign full of bulbs.

2 Likes

Just hang a bunch of em up all around your work shop. When it goes dark plug em into your skateboard and poof your shed is lit. Oh and make sure to get multi coloured ones :stuck_out_tongue_closed_eyes:

2 Likes

LHB -Best bet is to get an 8 ohm 100W wirewound resistor (Got mine from Aliexpress) and do a full discharge test from 42V to 30V (for 10S) with regular voltage readings of each cell pack. This will show up any dud cells.See my blog at https://evolveforums.com/threads/bamboo-gt-at-50km-range-project.1169/page-7 for how I did it .Check voltage and AH against cell specs -use exel for graphing and ah calculation -I can send file to you if you want. Note fan for cooling of resistor . 8 ohm at 40V will give 5 amp and 200W -but with a fan 100W resistor will be fine

3 Likes

Liked your idea so much @SkaterBoy58 I, errrr, stole it :grimacing:

Wrote some python code to periodically poll the DieBieMS and write the data out to a CSV file I could create a chart from fairly easy. I’ll give a few more details on this over in the other thread.

a123 LiFePO3 based pack, so if the voltages seem a little odd…

2018-01-17-21-09-09_cells

Discharge was no more than 1.6a, so I really wasn’t expecting to see much drift across the cells. Having a less well balanced pack probably would have made for a more interesting chart though.

2018-01-17-21-09-09_status

One of the nice things about the DieBieMS is how it tracks state of charge, and doesn’t just base percentage remaining off the current pack voltage. It’d be tough to get a reliable charge % based on the voltage change shown above. Red line seems right as the pack was drained using a constant resistance.

Will be interesting to run the same tests after the pack has been used for a while.

1 Like

Maybe use an engine starter? But it probably draws way more amps than batteries for eksate

I just noticed the wood under the resistors is singed :open_mouth: .

Do you have any spare fans hanging around? Just some forced airflow over them would greatly increase their heat dissipation

no worries -ideas are for sharing here My DIY pack is 10S8P with MJ1 cells so about 1000wh in total-explains the long time for my discharge test!

.I used a 200W load to simulate average riding conditions . That is at 20km/hr and 10wh/km =average of 200W .

Cheers

Just letting you know but the 10wh/km is for distance referencing not speed. Although it may work for speed too but I don’t think it does

Nah 10wh/km is typical board energy consumption going at 20km/hr Cheers

it takes about 440 watt hours to take ~150 pound man ~25 miles at ~20 mph on a calm day on terrain that is on average mostly flat.

That metric is based on the 12S4P 25R pack, which is what i used to build.

i base all of my gut calculations on that and have been getting adequately close results to what i suspect will happen.

Actually it is the typical consumption of power from your battery pack. That can vary greatly depending on hills, rider weight and quality of the cells

Other people have come up with results showing that around 4000 watts is required to reach 40mph (64km/h). According to what you have said it would only take about 650 watts.

Load test setup v2. Added some bullet connectors so I can switch between 2s2p and 1s4p. 4p drops the resistance and ups the amps to about 6, power works out to be approx 250w.

It was fairly obvious something was going to catch fire if I didn’t change anything, so here’s the new setup. As suggested I dug the fan out, but also g-clamped a meter long strip of aluminium I had lying around across the top of the resistors.

Kept an eye on temps throughout the test, and it was cooler than when I was testing at 1.6a. Nothing too surprising about the discharge curves, which is a good thing. Kept recording data after the BMS shut down discharge to see how the cell voltage bounced back. chart

And here’s the charge cycle after this test. The DieBieMS cell mismatch is set to 0.01v, which makes sense when you look at the chart. chart (2)

1 Like

LHB Ater doing US/Aust conversins this equates to about 10.9 Wh/km which is about what I get on streets on flat ground on a good day.

Good to know that the laws of physics work world wide?

Cheers

1 Like

Pat Think you getting mixed up with instantaneous power required for a particular speed and friction overcomming requirement with energy (with a time element) over a range of riding conditions

Cheers