Winkelwagen
Je winkelwagen is leeg
On a direct-drive hub with a regen-capable controller, regenerative braking typically returns a few per cent of your energy: about 9–13 Wh per 100 m descended on a 5% hill in our worked example, and under 1 Wh per stop. Geared hubs with a freewheel clutch recover nothing.

MOTORS & CONTROLLERS RECOVERY IN WATT-HOURS, NOT PROMISES UK + US

On a direct-drive hub motor with a controller that supports it, regenerative braking usually gives back a few per cent of the energy you use, not a third of your range. Grin Technologies puts it at roughly 3 to 8% on a typical ride with its direct-drive setups, close to zero on flat roads where you rarely brake, and higher on hilly or stop-start routes (Grin).
In watt-hours, our illustrative calculation below gives about 9 to 13 Wh back for every 100 m (328 ft) you descend on a 5% gradient, and well under 1 Wh for each stop from 25 km/h (15.5 mph). A geared hub with a freewheel clutch recovers nothing, because the wheel cannot drive the motor.
When you ask the controller to brake, it runs the hub motor as a generator. The wheel turns the motor, the motor makes current, and the controller pushes that current back into the battery. The resistance you feel is the braking force. Energy that would have gone into your brake pads as heat goes partly back into the pack instead.
Two things must be true first: the wheel has to be able to drive the motor, and the controller has to support regen and be set up for it.
| Motor layout | Can it regenerate? | Why |
|---|---|---|
| Direct-drive hub | Yes, with a regen-capable controller | Motor and wheel are one rotating part; nothing lets them turn separately |
| Geared hub with freewheel clutch | No | The clutch lets the wheel overrun the motor, so the wheel cannot drive it |
| Geared hub with a locked clutch | Yes, but it no longer coasts freely | Specialist motors and modified hubs remove the clutch on purpose |
| Mid-drive | Essentially no | The motor drives through the chain and the rear freehub, which freewheels |
From Grin Technologies’ regen guide, hub motor comparison and locked-clutch geared motor page, November 2026.
A geared hub is a small, fast motor driving the wheel through a planetary reduction. Between the gears and the hub shell sits a one-way clutch. It lets the shell overrun the gears, which is why a geared hub coasts with almost no drag, and why it cannot regenerate: braking means the wheel driving the motor, exactly the direction the clutch lets slip. Grin, who build a geared hub with the clutch locked for this reason, say the usual clutch prevents regenerative braking and reverse running (Grin). The trade is that a locked-clutch geared hub has more drag when you pedal unpowered: Grin quotes 1.0 to 1.5 Nm for theirs against 0.6 to 1.0 Nm for a typical direct-drive hub.
Our 250W front-wheel kits use geared hub motors, according to their product pages, while the 52V 2000W, 60V 2500W–3000W and 72V 4000W MTX kits use direct-drive rear hubs. See our geared hub vs direct drive guide.
Grin point out that many budget controllers do not support regen, or do it poorly. The controller also needs a brake signal. The commonest is an e-brake lever switch, which gives a fixed amount of regen; other systems use a second throttle, part of the throttle signal, or a speed limit that brings regen in on descents.
On KirbEbike’s smart controller, the Ride Power app lists “Regen.Braking” among its settings, and the product page FAQ says that on braking the motor feeds charge back to the pack, most noticeably on long descents. The 52V, 60V and 72V MTX kit pages each list regenerative braking among the app settings, and each kit ships with e-brake levers. As our app settings guide says, confirm it for your exact motor and controller pairing.

Regen cannot return more than your brakes would otherwise have turned into heat. That energy has two sources: height you lose on descents and speed you scrub off before stops.
Potential energy is mass × gravity × height (m × g × h). For 120 kg all-up (rider, bike, kit and battery) dropping 100 m, that is 120 × 9.81 × 100 = 117,720 joules, or 32.7 Wh (joules ÷ 3,600).
Not all of it reaches the brakes. Air drag and rolling resistance absorb part of it; the brakes only take the surplus above the speed you want. The worked examples below hold 25 km/h (15.5 mph) and use these stated assumptions: frontal drag area (CdA) 0.6 m² for an upright rider, air density 1.225 kg/m³, rolling resistance coefficient 0.008. At 25 km/h that gives about 17.7 N of air drag and 9.4 N of rolling resistance, 27.1 N in total.
| Illustrative descent, 120 kg, 25 km/h | 2% gradient | 5% gradient | 10% gradient |
|---|---|---|---|
| Distance to lose 100 m | 5.0 km (3.1 miles) | 2.0 km (1.2 miles) | 1.0 km (0.6 miles) |
| Gravity force along the slope | 23.5 N | 58.9 N | 117.7 N |
| Braking force needed (gravity minus 27.1 N drag) | None | 31.7 N | 90.6 N |
| Energy into the brakes | 0 Wh | 17.6 Wh | 25.2 Wh |
| Braking power while descending | 0 W | 220 W | 629 W |
| Back in the battery at 74% / 50% chain efficiency | 0 Wh | 13.0 / 8.8 Wh | 18.5 / 12.6 Wh |
Illustrative calculation, not a KirbEbike test. Braking energy = (m·g·grade − drag − rolling resistance) × distance. Small-angle approximation used; 32.7 Wh of potential energy in every column.
Look at the 2% column. Gravity is weaker than drag at 25 km/h, so you never touch the brakes and regen has nothing to collect. Shallow hills feed your speed, not your battery. On a 10% hill, more of the same 32.7 Wh reaches the brakes because you lose less to the air over a shorter distance.
Kinetic energy is ½ × m × v². At 25 km/h (6.94 m/s), 120 kg carries 0.5 × 120 × 6.94² = 2,890 J, or 0.80 Wh. That is the most a full stop can give you, before losses. At 40 km/h (25 mph) it is 2.06 Wh, because speed is squared. At 150 kg and 25 km/h, 1.00 Wh.
Braking energy passes through the motor as a generator, the controller and the cells, losing some at each step. For the examples I have assumed 80% for the motor, 95% for the controller and 97% for the battery, which multiply to about 74%. Battery University gives about 97% energy efficiency for lithium-ion charged at 0.5C, which is where the battery figure comes from (BU-808c). The motor and controller figures are planning assumptions, not measurements, so I have also shown a pessimistic 50% case.
Regen also weakens as the wheel slows. A 2021 bench study of an ebike hub motor (Candan and Mamur) spun the motor up, applied regen and measured the energy returned to the battery. It recovered 6.37% at 320 rpm, 5.15% at 237 rpm, 0.57% at 160 rpm, and nothing at 84 or 45 rpm. Those percentages describe that rig and its lead-acid battery, not a road ride, but the trend suggests the slow end of every stop returns little.
Put the two sources together and Grin’s pattern falls out of the arithmetic. These use the figures above, plus an assumed consumption of 20 to 25 Wh per mile (about 12 to 16 Wh/km).
| Illustrative ride, 120 kg | Braking energy available | Recovered (50–74%) | Share of ride energy |
|---|---|---|---|
| Flat, steady, 20 miles (32 km), few stops | Close to 0 Wh | Close to 0 Wh | Close to 0% |
| 10-mile (16 km) city commute, 20 stops from 25 km/h | Up to 16 Wh | 8–12 Wh | 4–6% of 200 Wh |
| 20-mile (32 km) hilly loop, 300 m descended at 5% | 53 Wh | 26–39 Wh | 5–8% of 500 Wh |
| Long descent, 500 m down a 10% road | 126 Wh | 63–93 Wh | Depends on the climb that preceded it |
Illustrative calculation. Stop energy assumes every stop is fully braked from 25 km/h, which overstates it. Consumption figures are assumptions for the example; measure your own.
The commute and the hilly loop both land inside Grin’s 3 to 8% band. Grin also note that heavier vehicles that brake a lot do better: one of their staff gets about 20% back on a recumbent trike.
The trap is the round trip. Climbing 100 m at 120 kg takes at least 32.7 Wh from your battery and legs combined, plus motor losses and drag on the way up. Coming back down returns 9 to 13 Wh of it in our 5% example. Regen softens the cost of a hill. It never refunds it.
Divide recovered Wh by your consumption. At 25 Wh per mile, 13 Wh is about half a mile (0.8 km) of extra riding. The 26 to 39 Wh from the hilly loop is roughly 1 to 1.6 miles (1.7 to 2.5 km). Useful, but no reason to buy a smaller battery.
That is why the KirbEbike range calculator does not add any regen credit. Its published methodology lists modifiers for terrain, weight, speed, tyres, wind, temperature, stop-start riding and drive type, and it applies a small consumption penalty to direct-drive hubs (×1.05, or ×1.12 on hilly and off-road routes). Plan your battery on that band. Treat whatever regen returns as margin at the end of the ride.
If you fall well short of the band, start with our range loss diagnosis guide, and size the pack with the Ah and Wh guide.
Grin call range the smallest reason to use regen. The bigger gains are mechanical:
Regen is not your emergency brake. It fades to nothing at walking pace, and on a 2000W to 4000W build the mechanical brakes must still stop the bike from full speed on their own. Our brake and wheel upgrade guide covers what those builds need.
Connect the Ride Power app, find Regen.Braking under Settings, and start at the lowest level. Try it somewhere quiet first: a strong setting on a rear hub can step the back tyre out on loose or wet ground, like grabbing the rear brake. Save a known-good profile; our 2000W tuning guide explains why.
Regen reverses the torque on the axle. Grin designed a no-play splined torque arm for their regen motor for exactly this reason. All three KirbEbike MTX kits include a torque arm. Fit it with no slack, and recheck the axle nuts after the first rides. See our torque arm guide.
A full pack has nowhere to put recovered charge. Battery University is plain that lithium-ion cannot absorb overcharge, and charge current must stop at full (BU-409). A KirbEbike 52V Taishan pack is full at 58.8V. If you live at the top of a long hill, expect little or no regen on the first descent of the day, and ask support how your controller and BMS handle it.
The 52V 2000W, 60V 2500W–3000W and 72V 4000W MTX kits use direct-drive rear hubs and list regenerative braking among their Ride Power app settings, with e-brake levers and a torque arm in the box. Already have a direct-drive hub? The smart controller comes in ratings from 25A to 100A.
Regen needs a hub the wheel can drive and a controller that supports it. On a direct-drive kit it returns a few per cent on most rides: nothing on the flat, about 9 to 13 Wh per 100 m descended on a 5% hill in our worked example, and under 1 Wh per stop.
Plan your battery on the range calculator with no regen at all. Then use regen for what it does best: saving pads and holding speed downhill.
Motor type → controller support → route → watt-hours.
Typically a few per cent. Grin estimate 3 to 8% of energy on a typical ride with their direct-drive systems, near zero on flat steady roads, and more on hilly or stop-start routes and heavy vehicles.
Not normally. A geared hub’s freewheel clutch lets the wheel overrun the motor, so the wheel cannot drive it as a generator. A few specialist geared hubs have the clutch locked to allow regen, at the cost of drag when coasting.
At 120 kg and 25 km/h (15.5 mph) the bike carries 0.80 Wh of kinetic energy. After losses, expect roughly 0.4 to 0.6 Wh at best, and less in practice because regen fades at low speed.
The practical limit is a full pack. Battery University notes that lithium-ion cannot absorb overcharge, so charging must stop at full voltage. Do not count on regen at the start of a ride on a freshly charged pack, and ask support how your controller and BMS handle it.
No. It has no regen credit, and it applies a small consumption penalty to direct-drive hubs. Treat any recovery as a bonus.
Bronnen
Author: Andy Kirby. Last reviewed: November 2026. KirbEbike has not run its own regenerative braking test. Every recovery figure here is either published by the source named beside it or an illustrative calculation shown in the text, using the stated assumptions (120 kg all-up, CdA 0.6 m², rolling resistance 0.008, 74% or 50% recovery chain). Product details come from KirbEbike’s live smart controller and MTX kit pages, November 2026.
Stay in the know
Conversion guides, fitting advice and the deals worth knowing about. No spam.
Door je aan te melden ga je akkoord met ons Privacybeleid.
How To & InstallEbike Controller App Settings: Three Profiles for One 2000W SystemTwo settings do most of the work. DC current sets the ceiling, acceleration strength sets how fast you approach it — and because they are independent, one kit...
Batteries & ChargingWhy a 2000W Ebike Cuts Out Under Load: BMS, Voltage Sag, Connectors and HeatStart with the pattern of the cut-out, not the part you suspect. Whether the display goes dark, whether power returns on its own, and whether it happens only...
Batteries & ChargingWill an Ebike Battery Fit My Frame? How to Measure the Case, Rail and Removal ClearanceThree checks decide it, and most buyers only make the first. The case has to fit the frame, the rail has to sit where your bosses are, and...