Help me decide on a battery build

[maniac4bikes]

I've slowly been upgrading my ebike from the cheaper products it came with (mostly Chinese manufacturers like Kunteng) to higher quality components.

My next upgrade is going to be two-fold, upgrading to an ASI BAC 855 controller and a new battery to go with it.

My current battery is a 48v 10Ah in a 13S 4P configuration. I get around 20 miles round trip.

With the new controller it's tempting to go up to 72v since that's supported but the dimensions of the battery compartment mean I would have to go with a 20S 2P configuration to reach 72v and be able to fit everything. As an aside I'm thinking of using Sony: US18650 VTC6 cells.

From everything I read you always want higher voltage for greater efficiency, and if that's compounded with the greater efficiency of the programming in the ASI controller I'm wondering if could potentially get the same range from my proposed 6.24aH 72 v battery compared to the previous 10aH 48v with the obvious advantage of added performance and lower heat/wear on the entire battery system. What do you guys think, is it worth going to 72v?

 

[calab]

How fast do you want to go? Voltage.
Motor kv is?
Wheel size is?
Your desired speed is?
Dont forget setup and gearing, having a comfortable and pleasureable ride with the body position and gearing to match your style and preference along with pedaling cadence.

Remember - The faster you go, the more wh/mile you use up.

 

[amberwolf]

I doubt you will get that much greater efficiency to see a significant difference in Wh/mile between the two systems, but part of it will depend on your actual use case and the differences between the usage of the two systems.

If you ride exactly the same between the two, no change in speed, acceleration, conditions, riding habits, etc., then the more efficient system (whichever that ends up being) will use less wh/mile by however much more efficient it is. Whether that is enough to get the same range or better depends on how much lower the wh/mile is.

Riding conditions will probably not change, but your riding style might if you setup the new controller to take advantage of it's potentially higher capabilities, and that will affect power usage negatively if so.

72v x 6.24Ah is 449.28Wh, and 48v x 10Ah is 480Wh. So your newer significantly (almost 10%) smaller pack will probably give you less range. My guess without knowing how you calculated the pack Ah is that you won't actually get 6.24Ah out of the cells, because of BMS limits for LVC / HVC, and the loading on the cells in use that causes voltage drop that forces LVC to kick in sooner than with no load. If you haven't already, you may want to look at cell testing charts like the ones at lgyte-info.dk to see how that works on different cell brands, models, and versions, and see if they have a chart on the cell in your existing pack, and one on the cells you want to use in the new pack.

Being only 2P vs 4P, the higher voltage pack will be less able to provide current without pack heating; higher current per cell also means more voltage sag and less watts to your wheel. As long as your new system draws proportionally less current than the old one, that shouldn't be a problem, but it means ensuring you set a proportionally lower battery current limit in the ASI vs the KT.


The efficiency difference between the controllers will depend on how well each one matches the motor and usage, so the better you can fine tune the ASI to match your motor's characteristics, and the better you tune it's feedback loops and limits to match your actual usage, the better it will run the motor.


Something else to keep in mind is that a motor is wound with a specific kV, or RPM/volt. So it will try to spin faster at a higher voltage. This means it will require a significantly lower average output voltage from the controller to the motor to spin it at the same speed it now goes (you don't want it to spin faster, because that will increase your speed, which will use more wh/mile and decrease your range). Or if it is a middrive, not a hubmotor, change the gearing from motor to wheel proportionally to your average voltage change.

If you like, you can simulate the general behavior of the two different systems using the A/B compare mode of the ebikes.ca motor simulator, using the Phaserunner in place of the ASI, and the Grinfineon in place of the KT, then putting in your other system components using the custom fields, icnluding the motor (unless your motor is already listed), and your riding conditions. That will let you guesstimate the approximate usage for each system; it could vary significantly from the guesstimate in reality, though, since the simulator is a static setup, and your actual conditions vary all the time during a ride. (there is a trip simulator that can help with that, but it *is* still a simulator).

 

[electric_nz]

I was using vtc6a 21700 and they would give their rated capacity, but after a few hundred cycles the voltage would drop a bit and they weren’t as zippy as at the start. I’d be reluctant to put just 2 in parallel as they would probably sag lots after a while, depending on how much you were asking of them

 

[maniac4bikes]

Remember - The faster you go, the more wh/mile you use up.

For the sake of comparison my current pack and controller requires around 600watts to maintain 20mph.

I doubt you will get that much greater efficiency to see a significant difference in Wh/mile between the two systems

If you ride exactly the same between the two, no change in speed, acceleration, conditions, riding habits, etc

Being only 2P vs 4P, the higher voltage pack will be less able to provide current without pack heating

can simulate the general behavior of the two different systems using the A/B compare mode of the ebikes.ca motor simulator

So here's the part I left out, my current 13s 4p pack has really crummy cells that are rated at 5A max continuous discharge which means that my entire pack can only produce 960 watts max... Assuming an even comparison of 600 watts draw my current pack is running at 62.5% capacity.

But the new pack I'm building 20s 2p is with batteries that have 30A max discharge, which would give me 4320 max watts and assuming 600 watt draw on the new controller (which is probably high, I bet ASI is more efficient at the same wattage and absolutely more efficient at higher wattage) so to assume 600w draw on the new pack would mean it would be running at around 13.9% capacity... That's where my logic is heading with the idea of building a 72v 20s2p pack .... And last part I left out (because I wanted a direct comparison of one 48v pack vs one 72v pack) I'm actually going to be building 3 of these packs so I'm not as concerned with there being less total capacity of a single pack, but I'd like to think in a perfect world with running a better more efficient controller at a more efficient wattage and running the pack at much lower % of total power capacity I would like to think that my 72v pack with far more advanced cells and a lower mAh capacity could outperform my 48v pack with a higher mAh capacity and crappy cells and outperform in every way in terms of acceleration, raw performance, and distance...

The ebikes.ca simulator is really cool, I did not know about that.... thanks for the heads up..

depending on how much you were asking of them

Exactly, as noted above I'm planning on running the 72v pack at a far lower percentage of total available current but it'll be nice to have 4300+ watts available for getting out of bad situations on the road and for kicks even if I keep my top speed the roughly the same most of the time...

Thanks everyone for replying and giving me your thoughts and perspectives this is my first big performance upgrade and first battery build (times 3) so it helps to have a sounding board to reflect on.... You guys are awesome!

 

[amberwolf]

For the sake of comparison my current pack and controller requires around 600watts to maintain 20mph.

It will almost certainly take about the same power to maintain the same speed. The difference in efficiency between controllers is unlikely to be enough to make a signficant dent in that--it will make *some* difference...just probably not very much.

Also, it will take more power to go faster, and to accelerate more quickly, so unless you limit your system to prevent any usage that is different from your present capabilities, it will end up with enough other power usages to more than make up for any efficiency gains and end up with higher power usage than before.

So here's the part I left out, my current 13s 4p pack has really crummy cells that are rated at 5A max continuous discharge which means that my entire pack can only produce 960 watts max... Assuming an even comparison of 600 watts draw my current pack is running at 62.5% capacity.

But the new pack I'm building 20s 2p is with batteries that have 30A max discharge, which would give me 4320 max watts and assuming 600 watt draw on the new controller (which is probably high, I bet ASI is more efficient at the same wattage and absolutely more efficient at higher wattage) so to assume 600w draw on the new pack would mean it would be running at around 13.9% capacity...

Well, it isn't capacity you're talking about, its' current delivery capability, often listed as C-rate.

That said, it would probably strain the new pack less than the old one, assuming you set the battery current limit and other behaviors and limits of the new controller the same as the old one.

in a perfect world with running a better more efficient controller at a more efficient wattage and running the pack at much lower % of total power capacity I would like to think that my 72v pack with far more advanced cells and a lower mAh capacity could outperform my 48v pack with a higher mAh capacity and crappy cells and outperform in every way in terms of acceleration, raw performance, and distance...

Some thoughts so you understand how this will affect what you expect from your battery:

If you actually use the higher accleration and better performance, you will use more power and the system will be *less* efficient overall than previously, and you will get less range from the pack, assuming that the Wh of each pack is as I previously posted calculations of (that is something you can actually verify using a coulometer or wattmeter, or things like the Cycle Analyst from ebikes.ca...oh, and they also make the Phaserunner family of controllers based on the ASI line, but using their own custom made setup software that is probably easier to use than ASI's).

If you limit the new controller to match the performance of the old controller, the power usages will match, but you will lose the ability to use the higher performance it could give you in those emergency situations.

If you don't limit the new controller to match the old, then even if you manually limit your speeds on the flat roads to match the old speed limit of your system, then every accleration especially from a stop, every ride up a slope of any kind or length, riding against a headwind, etc., will use more power than it did before unless you are very deliberate in preventing that from happening by manually monitoring power usage constantly and self-limiting it.

If the controller supports presets via a display or a switch, then you can use those to create a normal preset for riding with your limits matching the old system, and switch to "unlimited" mode for sitautions you need it and dont' care about extra battery usage.

If it doesn't suport that, you could setup the controller as unlimited, and use the Cycle Analyst as a preset/limiting tool. See the Ebikes.ca CA V3 info page for how the CA works and how you could use it this way if you're interested in the idea.

If it doesn't matter whether or not the new system is actually more efficient than the old, and you really just want the extra performance, then none of that matters, but if you *need* that greater efficiency, you'll have to do something to limit the performance to keep it.

The ebikes.ca simulator is really cool, I did not know about that.... thanks for the heads up..

It can do a lot to help with guesstimates of system usage and modelling...just remember it *is* a simulator so it's output is only as good as the input and the models.

 

[amberwolf]

Some other food for thought:

https://endless-sphere.com/sphere/threads/48v-vs-72v.90540/post-1320121

Potentially the hub and wires can run 33% cooler at 72v . make sure you invest in regen braking these days it's very simple it just requires an ebrakes and a regen controller and it's safer and cheaper than regular brakes.

A commonly shared misconception, including by myself years ago, but the motor actually ends up with exactly the same average current in the winding for any given power and speed. When operating from a higher voltage battery it just means the mosfet is switched on for a shorter period to ramp up the current in the motors winding Inductance before it switches off and let's the current and field flyback into the body diode to complete the buck converter circuit enabling the phase current to be identical despite initial starting pack voltages being different.