Power Source to Test ESC and Motor - Mimic a 22Sx36P Battery

'
I'm working on a 22Sx36P battery pack but I don't want to delay testing my ESC and motor

I need 90V DC and 288A peak and 180A continuous. If I were to power it from 220V, I would need 30kW. Quite confident that's a major upgrade and big $$$. Even if I back down to 200A peak that's still 90*200=18kW and with a 220V source that is 90A not including losses. First I need to review my shop electrical service.

I would prefer a 110 or 220V AC solution but willing to discuss series wiring of car batteries or some other hack.

Assume a budget of $750.

Thanks in advance for the help!

some references I have found on the subject:
http://www.instructables.com/answers/How-do-I-create-a-90-VDC-Power-Supply-with-a-toroi/
http://www.cnczone.com/forums/gener...13662-0-90-volt-variable-dc-power-supply.html

That's some serious juice you're talking about. Even at 220V that's going to be 118A, at 100% efficiency, 130A at 90%. Can your mains power even supply that? If it can that sounds like it would be an expensive power supply. Maybe several forklift chargers in parallel?

Car batteries could be an option, 7 in series will give you about the right voltage, but for anything more than short bursts at that amperage you might need to parallel them. You're probably going to see significant sag at high loads.

I can't think of anything else right now, but I'm interested in what others have to say.

After some review, I'm coming to the conclusion that I need a bank of batteries. Trying to modify my shop electrical system to handle 30kW of power is silly. Especially for 10-30min tests.

So to go the battery route, is there batteries on the market that will be cheaper to use for short term testing or do I just build part of my 18650 cell pack and power it off of that?

I have found this 12V, 35AH AGP deep cycle battery for $64 with free shipping:
https://www.amazon.com/ML35-12-Battery-Replaces-Mighty-product/dp/B00K8E0WAG

Let's assume I could draw 50A for 30min ( ). To get to 90V, 150A for 30 minutes of testing, I would need roughly 7Sx3P or 21 batteries. This would be about $1,260 and 483lbs.

Now, if I were to just build part of my 18650 cell pack, what would I need and how much would that cost?
The cell I plan on using is the Panasonic/Sanyo NCR18650-GA.
https://cdn.shopify.com/s/files/1/0674/3651/files/panasonic-ncr18650-ga-spec-sheet.pdf

They are rated for 10A with about 3AH capacity. So one cell could put out 10A for about 20min. The test pack could be 22Sx16P. 352 cells at $5/ea is $1,760 and about 45lbs.

So, even though I'll have to delay my tests of the ESC and motor, it's probably best to make this test pack to replicate my full pack as best as possible. This way I can work out the other issues of the pack design.

I doubt those AGP batteries would give even 10 mins at 50 amps !
So you would have to multiply your numbers if you really needed 30 mins....but do you for testing ?
And at 10a/cell that GA pack will be giving closer to 70v than 90v and dramatically lower at 288a peak pack current.
Have you considered renting a large genset ??

Hillhater said:

And at 10a/cell that GA pack will be giving closer to 70v than 90v and dramatically lower at 288a peak pack current.
Have you considered renting a large genset ??

Click to expand...

The ESC I'm testing is the newer Fliermodel 22S 500A airplane version. It is not clear to me what is limiting the voltage? For fully charged cells, the pack voltage is 22*4.2V=92.4V. Is this the limit of the Mosfets? If they see 110V at low current, is that bad? Because as Hillhater mentioned, when you start to pull 10A/cell, the voltage drops to 3.5V quickly for 77V pack voltage. So it would be desirable to have a pack at 24S with some headroom for higher loading conditions.

Note, I'm not really expecting this ESC to do much better than 250A peak and 150A continuous. Testing it should be interesting. But it will take some time to get the battery pack put together.

Quick and dirty test pack solution ..
https://hobbyking.com/en_us/multistar-high-capacity-4s-16000mah-multi-rotor-lipo-pack.html
6 of those will give you 24s, 16 Ahr with 160/300 amp potential for $300 !
Double up the quantity for more run time and less voltage sag ..still only $600
Quick and easy to connect up too.
Sell them on here after your testing and get most of your money back.

could just use some server power supplys.

xadmx said:

could just use some server power supplys.

Click to expand...

.?? Powered from what ?

Hillhater said:

Quick and dirty test pack solution .. https://hobbyking.com/en_us/multistar-high-capacity-4s-16000mah-multi-rotor-lipo-pack.html 6 of those will give you 24s, 16 Ahr with 160/300 amp potential for $300 ! Double up the quantity for more run time and less voltage sag ..still only $600 Quick and easy to connect up too. Sell them on here after your testing and get most of your money back.

Click to expand...

Hillhater, that pack is pretty darn impressive. $0.21/Wh is budget friendly. 184Wh/kg and 411Wh/L is pretty good at the pack level.

The cells are strange. 160g/ea and 8Ah/ea. What kind of cell is that?

I'm curious how hot those little wires get at 160A continuous. Any performance curves such as current vs capacity? Do you think these have less of a voltage drop than the GA cells at 10A (at equivalent capacity)?

Edit: The cells are flat and rectangular. Similar to cell phone batteries. http://www.electric-skateboard.builders/t/how-i-made-my-lipo-packs-slimmer/1415

They can't. They're 12AWG with a fairly high temp coating, but they're also connected to an XT-90 - rated for 90 Amps.

But they are also only 10cm long, and at 10c, only running for 6 minutes tops, designed for drones in the open air, getting full cooling.

So far from ideal, but will probably survive their designed purpose. I'm uprating my set to 2AWG on Anderson SB175s.

If you're in Sydney and you're ready to test soon, I may be able to lend you 6 packs for testing purposes. Too much later though, and they'll be installed into my motorcycle.

mkp007 said:

.
The cells are strange. 160g/ea and 8Ah/ea. What kind of cell is that?

I'm curious how hot those little wires get at 160A continuous. Any performance curves such as current vs capacity? Do you think these have less of a voltage drop than the GA cells at 10A (at equivalent capacity)?.....]

Click to expand...

Those are classic "RC Lithium Polymer". Foil pouch cells, as Sunder said , commonly used for RC hobby models and Drone flyers.
.. but they are useful for many other things as you can see on this forum.
I have never tested or seen test results to confirm their performance at maximum discharge levels but they do have very low IR so they should not sag any more than the GAs at max discharge .. IE less than 1.0 v per cell.
And yes, the connection wireing will be questionable at that level also.
..but as i said, you can halve all those issues, by using dual packs (32Ahr) to reduce the individual loads.
..Alternatively, you can find some similar but much more powerful RC packs on that site ...but also more expensive !

Not to be pessimistic, but I think the ESC will give out way before the wires or batteries on a 30 minute test...

'
I'm considering ordering 18 of those packs and configure them in 6Sx3P (packs) for 89V and 48Ah. This would get about 15min at 150A for $1,000 including shipping. Each pack will be drawing 50A which seems more reasonable. I may need to remove 2 cell off 3 packs to make is a 22S due to the limitation of the ESC. Not sure I want to risk blowing the ESC before I even test it. Is there a voltage regulator that I can plug a 100V+ battery pack into that will output 90V?

Sunder said:

If you're in Sydney and you're ready to test soon, I may be able to lend you 6 packs for testing purposes. Too much later though, and they'll be installed into my motorcycle.

Click to expand...

Thanks for the offer Sunder but I'm in San Diego, CA.

Punx0r said:

Not to be pessimistic, but I think the ESC will give out way before the wires or batteries on a 30 minute test...

Click to expand...

That is quite pessimistic, but rightly so. The ESC is beefy and I know that's not an indication of quality, performance or endurance...but it does provide some hope. I'll post some photos of it shortly.

I was about to point you towards the thread where liveforphysics bench tested the 420A Alien ESC but I see you've recently posted in it.