Is it normal that the pedalling assistance gets significantly less on an e-bike with low battery?

Question

I recently bought my first e-bike (Coboc Montreal Seven), which is advertised to have a ‘natural riding feeling’ by measuring torque sensor signals and translating them to the ‘right’ amount of support/pedalling assistance.

I noticed that with 20% battery power left I have a significant reduction in pedalling assistance and with 15% or less battery power left I have almost no pedalling assistance anymore, when I expected to have the usual assistance until the battery runs out (or is really low, e.g. 10%-5%). What I find surprising is that neither the integrated LED's (indication for the battery level) switch off, or the app showing any kind of warning for a low battery and resulting reduction in pedalling assistance.

I checked the product site and its specs, but I couldn’t find any information regarding the remaining battery power impacting the pedalling assistance.

Does anyone know if this is expected behaviour for e-bikes of this type or if there is something wrong with mine?

Answer 1

I think is is about the underlying behaviour of batteries and motors, and could be experienced with a crude controller, even a simple on-off switch.

That's a far higher spec e-bike than the one I've ridden the most, but I'm familiar with the feeling. My suspicion was always (I no longer have access to that bike) that it had to do with the available voltage from the battery sagging under load at high discharge currents when partially discharged.

Torque delivered to the wheel is what you feel in terms of assistance. That will vary with the battery's state of charge. How can I determine LI-ion battery charge while varying the load? at electronics.se has some detail that may be of interest. Looking at the specifications, (352Wh battery, 250-500W motor), you're discharging at up to about 1.4C (where 1C would drain the battery in an hour). That's probably only labouring up hills, but even at 1C the voltage drops off fast once you've only got about 20% capacity left - beyond 1C losses mean you don't even get the full capacity, according to the curves in the question I linked. From the equation for DC motor torque (this probably isn't a simple DC motor, but without more details, I'll treat it as one) the torque is linear with the motor voltage minus the back-EMF the motor generates in its own windings. From the battery voltage we also have to subtract that lost to the motor controller (the on-resistance of the MOSFETs or other power transistors). So the available torque will fall off faster than the battery voltage, in percentage terms.

This drop in battery output voltage will reduce the power available to the assistance via the motor control circuit, but the circuit probably isn't sophisticated enough to know that. The display of remaining battery is based on some assumptions that don't necessarily hold very well, due to the difficulty of measuring battery capacity under load.

In practical terms, when I used to occasionally borrow an e-bike, if the battery didn't start the day full, I'd turn assistance right off on the flat to save the battery for the last hill. This is likely to get worse when the battery ages, but even now it might be a good idea to plan your charge pattern to ensure you don't have a difficult bit when the battery is low.

Answer 2

Does anyone know if this is expected behaviour for e-bikes of this type or if there is something wrong with mine?

The best place for information would be the manufacturer.

As a general rule, no this would not be "expected behavior". But, it can certainly depend on the exact implementation of the motor controller for each bike.

I have ridden a wide variety of e-bikes, and have ridden bikes with two different types of controllers with the battery at a low, nearly-discharged state of charge. So I can say how those motor controllers work.

One used a Bafang motor and controller. As it reached the discharged state, I noticed no decline in power, until the motor just started "cutting out" intermittently. That is, power was cut completely, until the battery recovered enough to provide a little more power. As you can imagine, this was really annoying.

The other used a Bosch Performance Line CX motor and controller. On that bike, the battery did not actually get to a state of discharge where it wouldn't drive the bike at all, and I did not experience any decline in power, in spite of the fact that the indicated range was 0 (i.e. battery level was "zero bars", and state-of-charge low enough that the display wasn't willing to provide any range estimate).

Based on the official reply from a Bosch employee in this online forum, it appears that had I managed to discharge the battery even more, it eventually would have just "cut out", the same way the Bafang implementation does.

So, in both cases, the available power provided as long as the battery had some charge available, was the usual amount.

But that's not to say that a manufacturer couldn't include a feature that reduced power as the state-of-charge declined for the battery. And frankly, given the stated specs on your bike — 70-100 km on just 352 Wh of energy, which seems highly optimistic to me, except at the lowest assist levels — it would not surprise me at all if they achieve this by significantly reducing the amount of assist (and thus the amount of battery drain) as you approach a lower state-of-charge.

If you're concerned about it, you really should just contact the dealer and/or manufacturer and ask them what the expected behavior is. If it's not working as designed, you're going to have to contact them anyway, so you might as well start there.

(As an aside: in the most basic implementation of an e-bike, it's even possible that the power provided is limited by the battery voltage, which declines as the battery discharges. But all the motor controllers I've seen are "constant power" controllers, i.e. they draw a certain amount of current from the battery to achieve the desired output, and on a bike such as yours I can't imagine they would use something different from that. So if you're getting less power, I'm confident it's the motor controller that's causing that, not just a reduction in battery voltage.)