DIY Electric Car Forums banner

Regenerative Braking

118932 Views 60 Replies 26 Participants Last post by  MrZion
What is regenerative braking?
Regenerative braking is a way of slowing a vehicle down where some or all of the vehicle's kinetic energy is saved rather than being wasted as heat. Electrically this is achieved through the use of a generator (often the traction motor with the terminals reversed) and energy storage devices such as batteries or capacitors. This generator is usually an AC or permanent magnet DC motor.

What sort of range can be gained from this feature?
The typical stated range gain for regenerative braking is about 10%. AC Propulsion states as high as 30%, US Electricar measured as high as 20+%, Toyota RAV4 owners report as high as 25%. This would obviously be more effective in city driving rather than highway where little braking occurs.

Is regenerative braking possible on a series wound DC-motor?
Yes, but it is difficult and can be dangerous to implement. Some controllers, such as the ZAPI H2 have regen abilities built in but some have questioned the controller's reliability. One successful DIY attempt by Otmar Ebenhoech of Cafe Electric is documented here. Early 90's Soleq brand EV's were DC and had regen built in.

Why is it so hard to use regenerative braking on a series wound motor?
Series wound EV motors have field coils instead of magnets, which are not normally energised when the car isn't being powered forward. No current in the fields results in no current generated when the motor is turned by the inertia of the car coasting/slowing down. Specialised controllers can energize the fields, but even that isn't a simple solution because series motors run above 96V are usually "timed", that is the brush, armature, and field magnet positions are calculated and set in an advanced position to optimize the motor producing torque when it is powered by batteries. Attempting to cause that motor to become a generator will mean that the timing optimization is exactly opposite what it should be for producing power instead of consuming it, and the effective result is that excessive arcing and/or plasma is generated on the armature's commutator, usually ruining it. The motor can be timed in a "neutral" manner, but then it makes a less efficient drive motor, and the regen it might put out will be consumed just making up for the efficiency it lost by being mis-timed.

Is regen possible on a Permanent Magnet DC motor?
Regenerative braking is easier with a permanent magnet motor because the magnets do not need to be energised. Regenerative braking is achieved by having the controller reverse the terminals to the motor so that current flows in the opposite direction. Since these motors are also brushed they suffer from the same advanced timing problems when used at voltages greater than 96V. Typically though PM motors are smaller anyway and therefore run with neutral timing and lower voltages. Thus many motorcycles and small EVs run regen using PM motors at less than 96V.

Can an alternator or generator be attached to the drive train to implement Regen?
Theoretically a generator (or alternator) that is connected to the drive train when the brake pedal is pressed would be a simple way to implement regenerative braking. In practice it could be quite difficult to mount a generator of sufficient size (approximately the size of the motor for similar braking power to your acceleration) or enough 12V alternators to charge the individual batteries in the pack in an already crowded engine compartment. Also the additional complexity of a clutch system to remove unnecessary drag when not in use and controlling the current spike into the batteries further complicates things. All of this for a 10% gain in range is often not deemed to be worth the money, time and additional weight it would take to implement it. That said it would be an interesting experiment and challenge, and serve as a good example if someone was willing to try such a set-up.

Hybrid gas/electric automobiles now use a completely different method of braking at slower speeds. While hybrid cars still use conventional brake pads at highway speeds, electric motors help the car brake during stop-and-go driving. As the driver applies the brakes through a conventional pedal, the electric motors reverse direction. The torque created by this reversal counteracts the forward momentum and eventually stops the car.

But regenerative braking does more than simply stop the car. Electric motors and electric generators (such as a car's alternator) are essentially two sides of the same technology. Both use magnetic fields and coiled wires, but in different configurations. Regenerative braking systems take advantage of this duality. Whenever the electric motor of a hybrid car begins to reverse direction, it becomes an electric generator or dynamo. This generated electricity is fed into a chemical storage battery and used later to power the car at city speeds.

What are AC motors like at providing Regen?
Unlike brushed motors, AC motors (ie induction or perm mag) can provide regen very efficiently. AC motors can usually regen at almost the same efficiency as when motoring. Regen for AC motors also comes at no extra cost to the existing controller, although bare in mind that AC motors and controllers cost more than brushed controllers. Before purchasing an AC drive system it's worth investigating its regen capabilities.
See less See more
1 - 20 of 61 Posts
What kind of range boost can I expect using Regen?

If accelerating at a certain rate costs me 200A, how much charging current can I expect from a similar rate of de

Just some things I was wondering. :p
I typically hear that regen gives about a 10% increase in range, obviously that will vary with city/highway driving but 10% is the number that is always thrown around. And I think that the prius regen returned about 30%, so you would get 60A back. These aren't solid numbers from studies or anything, just what most people say when talking regen... hope that helps.
Can regen create more electricity than the batteries can handle? Or should I say does some of the electricity created by regen get wasted because batteries can only accept so much current all at once?
Can regen create more electricity than the batteries can handle? Or should I say does some of the electricity created by regen get wasted because batteries can only accept so much current all at once?
I guess so. Much depends on the system. Like the particular battery. And then the application. Like most often, the scenario used is that the EV is fully charged by the guy who lives at the top of a hill. So he goes down that hill and tries to regen into a fully charged battery. What happens? The system is smart enough not to regen. So there is no "electricity" wasted. Electricity is not generated and therefore no braking torque. What's the guy to do? Duh, use the original braking system. That is what it was designed to do. And if the guy was smart enough to figure out his driving mission ahead of time, maybe he would not fully charge his battery so he could take advantage of the potential energy available at his location.

And yes, some of the electricity used to charge batteries is wasted, whether the charge current comes from a battery charger or from regen. Batteries, like most things, are not perfect. So there are losses associated with charging and with discharging batteries.


See less See more
I typically hear that regen gives about a 10% increase in range, obviously that will vary with city/highway driving but 10% is the number that is always thrown around. And I think that the prius regen returned about 30%, so you would get 60A back. These aren't solid numbers from studies or anything, just what most people say when talking regen... hope that helps.
I suspect the Prius 30% figure is solely for optimistic regenerative cycle efficiency: how much of the power slowing down can be returned in later traction effort if everything is perfect.

A quick back-of-the-envelope calculation shows this to be likely. The motor and controller are each about 85% efficient. A very good battery is about 70% efficient at return charge energy. Multiplying these together, remembering that the power is making two trips through the motor and controller, gives you 0.85*0.85*0.70*0.85*0.85 = 36%. Factor in the drivetrain losses, at 90% efficient (although a dyno operator will give you a 130% factor for wheel-to-shaft horsepower) and you get very slightly under 30%.

Considering that each of these numbers is optimistic (e.g. a cold, loaded differential can absorb 10HP at highway speeds, the regenerated voltage will be far from optimal for charging), I suspect the 30% number is for the traction/power unit only, not as-installed: you won't be able to climb a 300ft hill from the power going down a 1000ft hill.

Consider regenerated power a tiny extra bump in efficiency, with a bigger bump in braking effectiveness. It's not a major win in range, despite what the press stories would have you believe.
See less See more
what efficiency does a pm dc motor get in regen? you mentioned that AC motors are quite good but what about pm dc?

might as well ask about SepEx while i'm here.

thanks for your time.
it always depends on the particular motor, and only the manufacturer will know that.
I hope it's ok if I updated this part of the wiki on regen. It's very valuable. Soleq's use sep ex to generate regen. I've personally measured 20-40% regen return on my vehicle.

Hi Mikep
Those numbers sound incredibly high - they imply that 30% of your energy is wasted in the brakes
I have only seen numbers of that level in inner city buses with frequent stops
Hi Mikep
Those numbers sound incredibly high - they imply that 30% of your energy is wasted in the brakes
I have only seen numbers of that level in inner city buses with frequent stops
I get that a lot. I've just observed it in other vehicles from ACP, RAV4, and my own. Of course the number can drop very low if the drive is all freeway. But my work commute was 20-40%.

Hi Mike,

Those numbers seemed high so I did a little calculating
I looked at the EPA cycles - do you guys drive like that?
City has 23 stops in 11 miles!
Even the highway is only over 10 miles!

Anyway I took all of the decelerations in each cycle and made a best case that ALL of the energy for these decelerations would go into re-gen

I took two extreme cars as examples

2003 Hummer -------- weight 3000Kg -- CdA 2.44 m2
1999 Honda Insight --- weight 840Kg -- CdA 0.474 m2

City Cycle
- deceleration energy
Hummer -- 2.1Kwhrs
Insight --- 0.59 Kwhrs

- Aerodynamic energy (just using average speeds)
Hummer --- 0.68Kwhrs
Insight --- 0.13Kwhrs

I assumed that 90% of the re-gen was captured

To my surprise I got
Hummer -- 2.78Kwhrs total - 1,89Kwhrs regen = 68%
Insight ---0.72Kwhrs total -- 0.53Kwhrs regen = 73%

Highway Cycle
- deceleration energy
Hummer -- 0.58Kwhrs
Insight --- 0.16 Kwhrs

- Aerodynamic energy (just using average speeds)
Hummer --- 3.32Kwhrs
Insight --- 0.64Kwhrs

I assumed that 90% of the re-gen was captured

I got
Hummer -- 3.90Kwhrs total - 0.52Kwhrs regen = 13%
Insight ---0.80Kwhrs total -- 0.14Kwhrs regen = 18%

These are theoretical maximums assuming no deceleration by just letting the car slow down and no bearing/tire drag but still they are much higher than I expected,

Looks like re-gen is more useful than I thought
Not so much for around here (Southland) but in busier areas .....
See less See more
Hi Duncan,

Nice piece of math. You'd be surprised at the return regen will give you. It's also a great way to charge your pack should you run out. Just get a tow!

Max regen values are equal to max output values. Mine puts out 300 amps, and it regen's 300 amps. Well, at least it does if it's not raining. But the gating factor is always how the motor controller is programmed. ACP adjusts theirs with a slide pot on the dash in the older cars. I don't remember if they still use that on the new stuff. Mine is setup in what ever 'gear' I select on the 'shifter'. The manual transmission is locked in second gear so the shifter is for the computer. Drive=0%, 2=50%, 1=100%.


Can you elaborate on your calculations? Are you estimating energy used per mile at some constant speed by estimating energy lost to rolling resistance and air drag, and adding this to the energy used to accelerate the vehicle to that speed to determine total energy used over some interval in the EPA cycle before stopping? Then assuming 90% of the vehicle kinetic energy is converted to potential energy in the battery pack as a upper bound for regen, and repeating this for each interval in the EPA cycle to determine the ratio of regen to total energy used?

I expected to get somewhat higher regen on my car, especially going down some of the large hills around here, but like I said I've not seen over 20% down a hill, and usually more like 10-12% in flatland driving. But then I don't drive like the EPA cycles, more on secondary roads with more like 8 -12 stops per 15 miles, less when I drive on the highway. The EPA cycle is about 3x more stops per mile than this, so I might expect something like 70/3 = 23% based on your calculation? I would expect a hybrid like the Prius to do better since it has a separate generator, optimized for generation, whereas the drive motor is not.
See less See more
Hi Tomofreno

I thought that the numbers Mikep was quoting seemed high so I did a "best case" calculation

I looked at the EPA city and highway cycles,
I took all of the decelerations and noted the initial speed and the final speed,
for each deceleration I calculated a V squared term and summed them for the cycle

With this I could calculate the total possible re-gen energy for the Hummer and Insight,
I took an arbitrary 90% for max possible re-gen energy

To get the total energy I added a term for the aerodynamic losses - I ignored the rolling resistance

The aerodynamic term will be too low as I just used average speed and energy used goes up with the cube of speed,

These were just to give an idea of what could be done - I was expecting much lower numbers

I could add rolling resistance to the numbers - but I am not sure where to get numbers for rolling resistance?

CdA and mass were from Wikipedia

I would expect a hybrid like the Prius to do better since it has a separate generator, optimized for generation, whereas the drive motor is not.

I used 90% - which is almost certainly far too high even with a dedicated generator as it needs to include storage losses as the batteries are charged as well as generator efficiency

These were just rough numbers to see how useful re-gen could be

They represent absolute Maximum possible - plus a bit

A lot of the numbers quoted on the forum have been very low -5% - not sure if that is
"my DC motor won't do re-gen but its not worth it anyway!"

It looks as if 30% in city and 10% general would be achievable re-gen
See less See more

Bob Brant gives F = C*W*cos(phi) for rolling resistance, with C = 0.012(1+v/100), v is vehicle speed. He adds 0.003 brake resistance and steering/suspension drag to this (Build Your Own Electric Vehicle, Bob Brant). Using this, and the equation he gives for still air drag force, Cd*A*v*v/391, I estimate that the combined rolling resistance plus brakes/suspension is 50 lb, and drag force is 53 lb at 60 mph for my car. Rolling resistance force is larger at lower speeds, for example 43 lb and 18 lb at 35 mph. I think 90% efficiency into the pack is optimistic for a generator and would not expect that for a motor. Edit: Bob references the Bosch Handbook for his equations.
I went through a similar exercise using the EPA city cycle here:

I already had power at constant speed and acceleration energy calculated for my car (2250 lb, Cd = 0.32, A = 18 square ft) in a spreadsheet. I estimated time at each rough average speed in the EPA cycle and took the product of that and power to estimate energy used at constant speed. Acceleration energy was estimated from the equation for force given by Bob Brant, Ci*m*a, where Ci depends on the moment of inertia of rotating parts such as wheels and flywheel, and overall gear ratio (final drive and transmission), m is vehicle mass, and a is vehicle acceleration. I used an acceleration of 4 mph/sec (about 13 sec from 0 to 50 mph). I made the simple assumption that regen was 90% of acceleration energy. With these I got total energy used at constant speed of 2393 Wh, total acceleration energy of 772 Wh, total regen energy of 695 Wh, and ratio of regen to total energy used of 0.22 for my car, which falls right about in the middle of your estimate of 10 to 30%. I would expect it to be lower for a vehicle with higher drag force, as well as for less stop/starts per mile. For example, at half the duty cycle (half the number of stop/starts same overall run time), the energy used at constant speed would about double, and the acceleration and regen energies would halve, giving about 0.07 ratio of regen energy to total energy used.
See less See more
I finally found my notebook. I had spend something like 6 weeks focusing on regen vs no regen on work commute for my last job in 2007. Going to work with and without regen netted an 19% gain in efficiency using regen. Going home, with and without regen netted a 18% efficiency gain using regen. This was a no highway commute. City streets only with 40mph max. Not too congested. Total commute distance one way was 5.7miles. Some elevation change. Fairly flat. Some long stretches with no stops. Truck weighs 4000 lbs. Summer time daylight temps. These numbers were taken with the hardware set to 250 amps max on output and regen. It's currently set to 300a and 300a. This commute resembles an EPA highway driving cycle, except the speed never exceeds 40mph.

I have a 5 mile test loop mapped out here that I call the San Jose loop. I'll test drive it to get more regen data. Now that it's winter here, it will be interesting to see if the percentage changes. The energy required is definitely higher in cold months. I could also, given the time, record the drive with the data logger to see how it compares to the EPA cycles.

See less See more
FWIW, Here is a quote from my blog where I just posted results of driving my Gizmo for 1 year and 5500 miles.

Also, the difference in the CA total of 12,544Ah and my total of 12,365Ah is due to the fact that the CA records total Ah that the batteries deliver whereas my figure is from adding up all the individual trip Ah which include regenerative braking. From this you can calculate that regenerative braking gained me 1.4% range over this time period. [edit: I just realized that 7597Ah came from back calculating for the time I did not have the CA and are based on the kWh used to charge from the wall. A more accurate representative result would be a 3.7% gain from regen.] Not a whole lot, however I have gone over three times farther on this set of brake pads than I did before I had variable regenerative braking so it is definitely a benefit. Not to mention the ability to come down steep long hills and not have brake fade and the much quicker panic stops than without.
Some may argue that my numbers are off because I used Ah and not Wh but since all the discharges were on the level part of the discharge curve there is very little, if any, measurable difference with my setup.
i guys,
i'm new in this field , so i'm going to ask some simple question just to understand better the problem. I have a brushless motor has a three phase winding on the stator and permanent magnets on the rotorals; when i'm going to brake in output from my motor i will have a three phase AC,so if i want to recharge a battery i need to convert it in DC, isn't it?
an other question is, which is the dependence between the regenerative voltage and corrent, and the rotation speed of the rotor? for example,i don't know, the voltage is constant and only the corrent changes.
Sorry if i made stupid question....:)
1 - 20 of 61 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.