Mild Hybrid e-Assist (fuel economy enhancer)

A

A-Spec Reviews

New member
I've been thinking about this for a while - what if you could convert your run-of-the-mill combustion-engine automatic transmission car into a mild hybrid?

My inspiration for this idea is the fact that low-voltage hybrids (typically 48V) are all over the place in European countries, and they provide drivability improvements at low RPM, instant engine start-stop (without the risk of burning up the starter), and as much as a 15% fuel efficiency improvement.
However, I have a few issues that I have found with this concept, and I'd appreciate it if you guys could help me figure some stuff out.

I found this permanent magnet AC motor online: https://evdrives.com/motenergy-me-1114-brushless-dc-permanent-magnet-motor-72v/. Rated 72 volts, around 50 lb-ft peak torque, draws about 400 amps at full load. I've seen @AlexLTDLX video on LiTo battery cells and figured that would be massive overkill for the power of the system. How many would I need for 72-73 volts?

The motor would be attached to the crankshaft via a belt, similar to a traditional supercharger. My main concern would be belt slippage, especially at max assist (about 25kW). Also, "max recommended RPM" is 5000. What would happen to the motor/controller if the shaft was revved to 7000 rpm? Would I need a physical or electrical disconnect above 5000 so I don't kill the motor or controller? I already plan on prohibiting assist above 5000 anyway.

As far as control, an Arduino microcontroller board would be plugged into the OBDII port, a relay hooked into the crankshaft position sensor wiring, and the motor controller. The control board (I'll call it the Hybrid ECU from now on) would calculate how much power assist or regenerative braking was required, based on the calculated engine load and accelerator position values.
When the driver's foot was on the brake, the system would wait until the vehicle was creeping (6 mph or below) before effectively disabling the crankshaft position sensor signal to the vehicle ECU. The ECU would panic, causing the engine to stop. All vehicle systems would continue to run, powered by the 12V battery. A DC-DC converter would charge the 12V battery from the hybrid battery, because this assist system essentially replaces an alternator.

When the driver lifted off the brake or the hybrid battery was low, the assist motor would kick back in, delivering its peak 25kW to spin the engine back up to idle speed quickly. Once the engine was at a workable RPM, the crankshaft position sensor would be reconnected and the combustion engine would take over.
Would about 50 lb-ft be enough for a quick restart of an average 4-cylinder engine? The system would not be tasked to handle cold starts, the 12-volt starter would do that.

Like other hybrids, this system would have specific conditions in which the engine would not stop when the vehicle did. For example, the engine would not stop if it was not at normal operating temperature, because cold starts would not be possible with the assist motor. The engine would also not stop if the ambient temperature was extremely low or extremely high, to ensure cabin comfort.

An additional issue would be motor cooling. If the motor was in place of the alternator, its cooling fan would be sucking in hot air, limiting peak torque and possibly making it impossible to restart the engine after an auto stop. If the motor was mounted on top of the engine, one could use a hood scoop to feed cool air to the motor, so it wouldn't overheat. That would look quite interesting. A bystander would see a hood scoop, suggesting a turbo or supercharger setup. When the hood was lifted, all they would see would be an alternator-looking thing with orange high-voltage cabling running to it.

As far as battery capacity goes, I'd need enough capacity to supply about a minute of maximum assist (of course, current may be limited anyway by motor temp). If the hybrid battery was low, assist would be progressively dialed back and the system would look for an opportunity to charge at the next stoplight (it won't stop the engine anyway, so why not charge the system up while you're stopped?) or while coasting. If the vehicle had to idle for any reason, it would charge the hybrid battery as much as it could. Continuous motor current is rated at 150 amps, so if the motor was at max assist for more than 30 seconds or so, the hybrid ECU would derate the motor to that level.

Do you have any other questions? I'd be happy to answer them!
 
I think you asked a similar question on one of my videos - welcome and thanks for joining the forum!

To answer your questions:

1. Belt slip isn't a problem at the power levels you're talking about. A 6 rib serpentine would be adequate with a good tensioner. Heck, I was pulling over 70 hp to drive my Whipple through an 8 rib belt with minimal tension once I built my own tensioner from a Hyabusa steering damper with 50w oil in it, ball bearings and an adjustable spring (the original manual tensioner I initially built required a ton of tension, still slipped and eventually threw a belt so violently it dented my hood and fender from the inside out). Here's the tensioner I ended up with on the dyno:


2. Cooling is an issue, but not a huge one. A small water cooling setup with a small radiator would be adequate. You'd really need to keep the motor under 150 degrees F mostly for the magnets - the neodymium magnets used in most high-performance brushless motors lose their magnetism at elevated temps (but still surprisingly low temps).

3. If you go with LTO cells, I'd do a pack of 30 cells. Why? Because at nominal voltage, that's only 69 volts (I realize that the Lishen cells I have are labelled "2.5 volts" - but in my testing, that's just fudging the numbers and picking a different point in the discharge curve - they self discharge pretty fast above 2.6 volts, and stabilize around 2.55 volts. It's not so much that they'd be overkill (they really wouldn't); but they'd generate less heat internally and are safer than LiPo cells. FWIW, I charge my pack of 28 cells to 68-70 volts. I don't have any more data yet because they're not installed, but all that's left is to put the Anderson connectors on the heavy cables to the packs - everything else is run, tested and buttoned up. Data will come soon.

Now for the big question - why would you do this and not an e turbo? Here's the reasoning. You already have an ICE in your car. You're talking about adding 24Kw - that's only 32 hp. With the controller in place, the best you can hope for is 90% efficiency, so that 32hp is more like 29hp. If your IC engine makes 100 hp, then you'd only need ~4 psi of boost, or about 3kW of e turbo to make the same power. Frankly, that's alot easier, cheaper, smaller and safer. 24Kw of e-turbo power is easily enough to support 800hp (!) on a well-built ICE. Heck, even OEMs are starting to get wise to this - if you haven't seen this thread, check it out:

Looks like we're ahead of our time - Kawasaki just jumped on board the e-boost wagon

I saw this article yesterday - and for months I've been thinking about posting a video on my youtube channel called "All Hybrids Are Built Wrong." Now I'll have to change it to "Almost All Hybrids Are Built Wrong"...
www.electrifiedboost.com www.electrifiedboost.com
I think you're the perfect candidate for a Torqamp.

The only reason why I can think of wanting to do this would be gas mileage; but even that's a fallacy unless you get the regenerative aspect as maximized as it could be; even then you're likely going to be burning gas to charge the batteries (conservation of energy and all that).

A Prius only gets great mileage because it's slower than hell. The old Honda CRXs got the same or better gas mileage (in the real world). Two approaches to the same end goal. Oh, and the CRX is almost 3 seconds faster to 60 mph.

I'll also say this: if you want to do this simply because you want to do this, then in my book that's as good a reason as any. We'll help you any way we can here.
 
Last edited:
Why am I doing this? Well, I'm NOT doing it to add power. The two main reasons for this concept are start-stop and regenerative braking (which can help deliver extra power when needed for acceleration.)
The Prius (I have a 2002 model) gets great mileage because its engine is run at its most thermally-efficient load almost all the time. It's not slow off the line, 0-30 is around 3 seconds because it has crazy (for a 2600-pound car) torque because of the electric motors. 0-60 bis about 12 seconds, but that's adequate IMO.

The conventional vehicles in my family all idle at stoplights. If they could shut off and instantly restart when needed, they could save some fuel and reduce emissions. If they could capture energy during coasting and braking, that energy could be used later to reduce the time spent in an inefficient (power-enrichment) range, and it could reduce downshifts under load (climbing hills, etc). Think F1-car KERS, but more hypermiler.

If the engine had to idle for any reason (battery low, ambient temp, engine not at operating temp, transmission in reverse) the system would put a little load on it to charge the battery and make sure the engine isn't just running and doing nothing. The goal would be to never load the engine down while driving, unless the driver is completely off the throttle.

An e-turbo would require even more fuel, and couldn't capture and re-deploy energy, not to mention start-stop. It also couldn't help optimize engine load to be more efficient.

RPM question: max recommended is 5000 for that motor. The 4-cylinder K series in my parents' Accord red-lines at 7000. If the motor was spun to 7000, what would happen? Would I need a clutch to keep the motor from blowing to bits, or an electrical disconnect to make sure the controller wouldn't pick up the back-EMF spikes and fry itself?
 
Last edited:
1)So the point of this is to add auto stop start at the lights?
2) Add re gen

I have some thoughts.
1)A. Typically the brakes are vacuum assisted if the engine is off then they are no longer assisted once the brake booster runs out.
1)B. Why bother adding your electric motor just get a much better battery and use the already there starter motor.


2) In your example the electric motor is connected to the Engine.. Im some what puzzled as to how effective regen will be when connected to the motor directly because that implies that you are engine braking 100% of the time.. I feel like this is a difficult thing to to in an Auto.

3)The RPM limit the faster you spin it the more likely it will explode .. So the manufacture is saying dont spin it faster than that.. If you spin it too 7000 it might last 10 years it might last 10 days. 7000 RPM is a long way over 5k i wouldn't do it.. but you could say how often do you really rev it over 5k so maybe it's not really an issue.
 
@MkngStffAwesome the point of this system is
A: Start-stop smoothly and quickly, without using a traditional starter.
B: Generate power during engine braking, and re-deploy it to improve efficiency and torque.

Answer to 1A: Start/stop would only engage when creeping at 6mph or less with foot on the brake pedal. There would be enough brake boost to hold at a stop with the engine off.

Answer to 1B: You would hear and feel the starter whirring every time you took your foot off the brake, you'd spend way more money replacing starters than what you'd save on fuel, and there would be no torque boost or re-gen braking.

Answer to 2: Automatics can engine brake. I'm just supplementing engine braking with re-gen drag on the crankshaft. I'm only doing it while coasting, so not all the time.

Answer to 3: Revving over 5K would not be very likely because of the torque boost of the motor, but I just want to know if the system could take a 7000-rpm burst. If someone really steps on it I just want to make sure the motor doesn't go BANG and send pieces flying everywhere, and the controller doesn't fry.
 
Last edited:
I don't have answer of your questions because I don't think I fully understand. Anyway! for the engine RPM. I'm a machiniste so I have seen some things. will your engine will take burst of 7000rpm? probably. but there's a lot more then that. If I was at your place, I would not try it. because, like mkgstuff said, the gap between 5k to 7k is HUGE. I think more of the "fatigue" of your engine. the fatigue of the parts will be your biggest enemy. IF your engine can go up to 7000 rpm 1 time, it will be different after 100 times
 
I wrote this whole long dissertation (and deleted it) about how start-stop technology doesn't help my fuel economy but hurts it (since 2014, I've owned 3 Jeep Grand Cherokees - all with the same engine, two had start-stop, 1 didn't), along with after-start enrichment strategies, IAT sensor heat soak, (I've been programming/tuning car ECUs since DOS was a thing); "gaming the mileage test," etc, etc. There are places, I'm sure, that you would see a fuel mileage gain. But not around these parts - I see a loss. Why? Because we're rarely stopped for longer than a few seconds, but at the same time, traffic crawls abysmally slowly - a good drive to the DC studio (14 miles door to door) takes 75 minutes; a bad drive takes over 130 minutes.

There's even this clown, who freely mentions being stopped for several minutes in his start-stop test when he's testing with the system off, but not when it's on and then brags about his 10% fuel gain:

Maybe, in an "urban" place with virtually no traffic - like his test site of Birmingham - you might save fuel. But here, in a real urban environment, you don't.

This guys speaks the truth:

And he followed it up with this (also very true - click on the pdf link in his description):

The other thing around here is if you're stopped too long, getting shot or carjacked is a real concern. Baltimore is 45 minutes away, and I've been there 3 times in the last 7 years, all for work. If you want to live, stay the hell out of Baltimore. And I've lived for years in NYC (Brooklyn, to be precise) - things are much worse here. That's why I want to leave so badly.

Anyway, can you clarify your goals? Is it to save money? You'll spend way more than you'll save, and you'll shorten the life of other critical components of your car. Is it to get better gas mileage? You might, but see my answer to the first question. Is it to make your car more environmentally friendly? You'll wear out more parts faster, and your engine as well - I don't believe replacing parts more frequently, adding weight to the car and shortening the car's life is more eco-friendly. Is it to do something different? This one I can't argue, and kind of want to see it done, to be honest. I do understand wanting more low-end torque, which and electric motor has in spades.

And no, you're not likely to blow up the electric motor at 7,000 rpm. You could always replace the bearings with higher quality ones if you'd like; and perhaps even balance the rotor to a higher tolerance.

I'll also posit this notion: your alternator is already a regen device. It just needs a load to regen into (get a big 200-350 amp alternator, and have a contactor activate a boost converter only on deceleration). You've given me an idea for my own e turbo setup...
 
@AlexLTDLX my goal with this system is to save fuel and reduce emissions while improving drivability and performance. At low RPM, the motor would fill in to prevent the engine from needing to go open-loop and run super-rich to extract as much power as possible. It may also help avoid transmission downshifts under load. Assisting the engine is one way this would differ from an average start/stop system. On auto-restart, the system could do double duty, spinning up the engine to idle speed and then continuing to assist in propelling the vehicle.

Speaking of restarts, is about 50 lb-ft at the crank enough to restart an average engine? I already plan on re-gen dragging the engine to a stop to avoid vibration, but I'm not so sure about the restart and how fast it would be.

Another question about RPM: what would excessive assist motor revs do to the controller? Would I need an electrical disconnect to prevent the controller from frying, and how would I do that?
 
Last edited:
You can calculate approximately how much torque you'd need to apply to the crankshaft to restart (I specify crankshaft, because pulley ratio comes into play). Let's say you need at least 160 rpm cranking:


And let's say current draw is a consistent 250 amps (a little high, but close); and with voltage drop you've got 10 volts. That's 2,500 watts or about 3.35 hp (2,500/746=hp). To get to torque from hp and rpm, the easiest thing to do is use an online calculator (the formula is: hp = tq x rpm / 5250).

I use this one a lot:
spicerparts.com

Spicer Horsepower and Torque Calculator

The Spicer Horsepower and Torque Calculator will quickly calculate approximate horsepower or torque, depending on your selections.
spicerparts.com spicerparts.com

That gives you 110 ft-lbs of torque. So you're less than halfway there.

As far as the motor assist revs/controller question, it depends on the controller. The one I just bought:

APD UHV 20S BRUSHLESS ESC - NeuMotors Brushless Motors

neumotors.com neumotors.com

Gives you up to 110% of your power voltage and 40 amps max as the limits. So at 7,000 rpm you'd likely be over the voltage limit on that controller; but the current is easily limited. Other controllers might have more, but keep in mind mine is a $1,700 unit.

To disconnect high-power stuff, look at EV contactors. I'm using one from a Tesla in my current controller box. To disconnect a motor, you'd need at least two.
 
Hyundai Sonata Alt is all you need. They are good for 20kw and 100+lb-ft of torque. Oakridge labs did the evaluation for the gov and there are several public docs that test hybrid system components and their ratings.
Hard to beat at under $100 to your door with pulley and brackets already installed. For the motor inverter/controller I would go with a prius gen 2/3 and an openinverter.org control board.
 
Are you talking about using one as a motor? Do you have links to anything about how that works? I'm also interested in the prius inverter controller options and how that works with the control board you mentioned. Clearly, you've looked into this :)
 
Here is the gov doc that goes into great detail on it's capabilities. In the OEM system it's used as a motor/generator for stop/start functionality, charging of hvdc battery and adding 20kw into the belt system under full load.


A prius gen 3 inverter has 2 seperate 3 phase motor control sections capable of 500A total and 600V, a 20kw boost converter, and 12V 150a dc-dc.
EVBMW sells a board HERE to control 2 separate motors independently and this is what I plan on using.
If a single motor and charging are required this board HERE. This allows for control of a motor with 1 of the 2 inverters and the second inverter is used to charge directly from the grid at "level 2" rates.

Openinverter.org would make a great resource for the site as there are a whole manner of motors on there with no purpose in particular other than utilizing them in EV projects
 
I've got to fix the hyperlink color on this forum. It's hard to tell what's actually a hyperlink. I'll try and get to it tomorrow. Thanks for the info. I'll look it over.

I'm actually working on my own setup in the evenings now. I really want to make a test pull this weekend and hit the track next weekend if all goes well.
 
AES said:
Hyundai Sonata Alt is all you need. They are good for 20kw and 100+lb-ft of torque. Oakridge labs did the evaluation for the gov and there are several public docs that test hybrid system components and their ratings.
Hard to beat at under $100 to your door with pulley and brackets already installed. For the motor inverter/controller I would go with a prius gen 2/3 and an openinverter.org control board.
Click to expand...
How many volts would I need though? The G2 Prius runs 500V, which is INSANELY dangerous. I have a G1 Prius that runs 275V, also very dangerous. The motors I'm looking at run 72V, which is an order of magnitude safer.
 
A-Spec Reviews said:
How many volts would I need though? The G2 Prius runs 500V, which is INSANELY dangerous. I have a G1 Prius that runs 275V, also very dangerous. The motors I'm looking at run 72V, which is an order of magnitude safer.
Click to expand...
If you want an e bike 72v is fine, anything more and you need more voltage. The 200v in your prius pack is the exact same as touching 120vac.
I have the v/rpm written on one at home but full pack voltage was 300v or so.
 
48-volt hybrid systems are becoming more and more popular, especially in Europe. If you put enough amps behind a 48-volt system, you can get 10kW of assist. With a 72-volt system, the same amperage would get you 15-17kW. Unfortunately, you can't just buy a 48-volt assist motor. I have seen some from the V8 Hemi RAM trucks (yes, they're hybrids now.) but that would likely require the ECU, engine, and battery from the Ram truck to get it to work (and a 5.7 Hemi won't make a car more efficient, LOL.)
 
10 kw is only 13hp. That would be barely noticeable in any decently-sized car unless it was woefully underpowered. I know in mine, even 50 hp is barely noticeable, and it's not that heavy @ 3,400lbs (probably a bit lighter now).

A-Spec Reviews brings up a good point about danger. Look at the damage a mere 70 volts caused to the robust aluminum case of my ESC in seconds: https://www.electrifiedboost.com/threads/another-esc-bites-the-dust-in-a-hellfire-inferno.67/ Imagine what 4-7 times the power is capable of.
 
Yes, it's only 13hp, but electric motors have massive torque. The OEM systems from manufacturers like Mercedes pull 184 lb-ft (at 0 rpm) out of a 48-volt system. Current is unknown, but I'd guess 500+ amps.
The most torquey 72-volt permanent-magnet motor I could find could do 60 lb-ft at 0 rpm, and max (probably only for a few seconds) output was about 25kW or 33hp.

Based on a few simulations I ran, the electric assist from that motor was noticeable, even in a 3500-lb vehicle with an extra 500 lbs of cargo. I could go right up hills that previously required a downshift to climb, and I could accelerate to pass without having to rev the heck out of the engine.
I also did some scientific tests. I started at the spawn location of a racetrack (to keep the starting points equal.) A 2-liter 4-cylinder without and with an electric assist motor was set up for this test. I shifted at 3000 rpm during each run. Without e-assist, the test vehicle (which weighed about 3200 lbs) would hit 51mph before slamming into the tire wall. With the assist, it would hit 57. Accelerating from 25 to 45 mph (about 1700 to about 4500 rpm) on a flat surface took 7 seconds without assist, and 4 seconds with it.
 
I'd dispute "massive torque" - more accurately, declining torque. Which is the opposite of an internal combustion engine - torque typically rises with rpm up to a point. Here are a couple of good articles:

things-in-motion.blogspot.com

How to estimate the torque of a BLDC (PMSM) electric motor using only its Kv and current draw

There exists a fundamental relationship between an electric motors velocity constant `(K_{V}),` armature* current `(I_{A})` and torque `...
things-in-motion.blogspot.com things-in-motion.blogspot.com

My current TP Power motor generates 2 ft-lbs of torque as calculated by the kV method (in the second link above). That's relatively accurate and how I'm able to get away with my hex drive using only a 6mm bolt. In your simulations, if you wanted to go faster without the electric motor, simply shift it higher with more throttle. I think we're looking at things from two different perspectives: you're looking at it under extremely light load conditions, when you're barely using any power from the ICE in the first place. Then yes, of course you'll notice a difference. Adding 50 ft-lbs to 30 ft-lbs is significant. But not any better then the ICE can do at WOT. I'm looking at it from a balls to the wall, wide open throttle pull - 50 ft-lbs on top of my 500+ ft-lbs really isn't much; and with the torque converter flashing well past the torque peak anyway, HP rules to day (acceleration is the key). The Whipple I had on the car before would leave on 13.2 psi of boost at 3,200 rpm - making well over 750 ft-lbs at the crank. With the torque converter doubling that, and a further gear multiplication of 5.54, it was jamming over 8,000 ft-lbs to the track. And it would dead hook on slicks. The only way I can describe it is being stopped and having a semi truck rear-end you at 80 mph. The first time I released the trans brake in anger, I swear I thought the entire driveline fell out of the car, and the garage door opener clipped to my visor flew off and bounced off the rear window. You don't even realize the front wheels are off the ground at that point. Greatest feeling in the world.

LTDWheelUp1.jpg

Alas, this was only a baby wheelstand, but it's the only photo I've got. I expect the electric turbo to do the same or better than the Whipple used to, if for no other reason than it's lighter and the weight distribution is better.

But as far as electric motors/torque/performance is concerned, a Tesla Plaid doesn't do any better than an ICE car based purely on HP calculations (torque is actually irrelevant). Here's a calculator that I've used for years - it's pretty much spot on based on literally thousands of trips down the track: http://www.wallaceracing.com/hpcalculatorquarter.php

Put in a Plaid's specs, and you'll see. Keep in mind an EV's hp rating is the same as the "rear wheel" rating in the calculator.

In essence I think we're both right. But I think that my "right" is a helluva lot more fun :)
 
In my acceleration simulations, I was flooring the accelerator, shifting at 3000 rpm each gear This was with a conventional manual, so no torque converter was involved.
As for the more "real world tests" I did, the traffic AI in the simulator I use likes to brake unpredictably, dropping from the speed limit to 20 under in one second, and then slowly getting back up to speed. Without the electric assist, I had to drop from fourth to second and floor it to get around a dawdling traffic car on a two-lane road. With the assist, I only had to drop to third, and if I had enough room, I could leave it in fourth.
I simulated also the torque curve of the electric assist. If I really revved-up the engine, I could feel the assist fall off as the revs climbed past 5000. In the real world, that would be acceptable, because this system is meant to "fill in" where a gas engine has trouble providing adequate torque (down low in the rev range.)
 
You must log in or register to reply here.
Top