What is considered 'good' and 'bad' performance retention?

So the nature of electric vehicles is that performance will drop along with the voltage.

But what do you consider a good and bad performance retention rate? Has anybody ever compared and quantified data on this?

For instance, I have a board that tops out at around 25 MPH at 100% charge. When it reaches 50%, the top speed is closer to 22 MPH. At around 20%, I’m down to 18 MPH. Would that be considered ‘healthy’ numbers or would it suggest that the cells I’m using probably aren’t the best?

Is there a theoretical way to predict performance drop-off based on, say, a discharge capacity test as seen here:

1 Like

Welcome to the ESK8 forum, @Pstrobel1991. That sounds fairly typical. I’d recommend stopping your ride at 30% or higher if possible for longer battery life. If you’re looking for a bit less voltage sag the thing to do is add more cells in parallel. Does your speed controller have data logging?

1 Like

That would depend on the programming of the esc. However, depending on battery, after reaching their 3.6v discharge, the power degradation will be more noticeable. So for battery, you would use one with a high rated discharge rate. For example p42a. It says 45, but the safe line would be 35A. This will still give you the speed without gaining voltage sag. The more pgroup you have, the longer it takes to get the sag.

For esc, there is a cut off start and end programmed into the esc, which also slows down your board dramatically. You can decrease the cut off timer, but make sure your cut off end isn’t below 3.3v (for lipo). For Li-ion, about 3.0v since the lowest is 2.5v. Make sure your voltage is .3v away. For example: a 10s set up would be 33v cut off start, and end at 30v (li-ion)

Other than this, I have no information about other things. All I know is that if your battery can’t maintain 4.2v in full charge and it goes as low as 4.0, it’s no longer healthy.

Note: multiply these numbers by your batteries series, which are usually labeled [XS]XP in the bracket.


Thanks guys, extremely useful. This is on a Chinese board so I don’t actually have much configuration options (nor the know-how) to get completely accurate measurements (the numbers in OP are just guesstimates after testing for some time). Very insightful replies. One additional question - towards the very end of the charge (sub-10%), I start experiencing slight “jolts” when holding the throttle down. The lower the battery level gets, be more often these small “jolts” or “shakes” in the board occur. Do I understand correctly that this is voltage sag - I know it lowers accel and top speed but didn’t read anything about jolts/shakes. It’s a 10S2P battery with Samsung 22P cells by the way.

Edit: I’ll definitely try to end my trips before getting below 20-30% but with a 36V 4.4Ah battery, it isn’t a lot of range until you run low on juice if you like to go heavy on the throttle.

1 Like

Not sure where you’re getting these percentages. Sub 10% would be damaging to li-ion cells. Whats your resting voltage after a ride?

Ideally you should have a way of monitoring pack voltage if not individual cell voltages during use.

Gearing your vehicle so your preferred top speed is always attainable would alleviate this effect. Needing to always drive the board at its top speed seems like a design error. What if your car’s top speed was 100km/h (62mph)? I’d say the design should be blamed.

That being said, keeping the top speed as low as possible will buy you range, at the expense of, of course, top speed. If range is what you desire, then this could be what you want. It all depends what you optimize the design for.

1 Like

LiFePO4 cells give a lot flatter discharge curve, with far less “sag”. They top at at nearly the same speed throughout most of their discharge cycle.

They’re also safer and last a lot longer then 4.2V li-ion cells.

The only downside is they have a higher weight per range, so your battery would need to be twice as heavy (and bigger).

I do use LiFePO4 batteries on many of my skates. They work great for “tool” skates like commuter boards but they are not great for “toy” boards like big electric mountainboards.

1 Like