Mild electric hybrid using a Permanent magnet PMG 132 motor?

[Ryland]

I've read about the motor/alternator that was being marketed to replace a stock alternator to make a mild series hybrid and I think GM is doing something along those lines with the new Buick's and people have talked about building their own that would work the same way, but so far I haven't seen anyone do it, one complexity would of course be getting enough power out of a 12v motor and fitting a 12v motor in the location of the alternator that was powerful enough to do any usable work, so it seemed kind of pointless, of course you could use a higher voltage motor/alternator but then you need a dc to dc converter too.
Then today I was thinking about how my Civic has a large empty spot where the A/C would bolt on, it's right at the front of the engine with threaded holes for the compressor to bolt to the engine block, so why not make a bracket to allow a motor like the E-tek or a Perm PMG 132 to bolt on in that location? specs on the Perm PMG 132 seem to say that it can handle over 12,000 RPM (red line on my Civic VX is 6,000 RPM) to a 2 to 1 reduction would work, spinning the motor at 4,000 RPM while the gas engine was puttering along at 2,000RPM.

My thought for throttle is to try to keep the engine in lean burn, so low load and that is measured by manifold vacuum, as vacuum drops power to the motor increases, when vacuum is high, as in when the ECU is told to cut fuel to the fuel injectors then regen braking should start to ramp up, I haven't yet worked with programming speed controllers, but as I understand it, you can program how fast things like regen ramp up and how strong, if possible it seems like two regen settings would be ideal, one for when engine vacuum is high and another for when the brake lights come on, I don't think most speed controllers are set up to work like that, but that is an area that I need to check in to more.

A lot of speed controllers also have a half power setting, for going in reverse or for letting someone else use the vehicle, it seems like this half power setting could be used as a "long trip" setting, where you want the battery pack to last longer.
I of course thing this would be best as a plug in hybrid, 72v is the max of what those motors are rated for and because it's not powering the car all on it's own that seems like enough to give it an added boost to keep it in lean burn all the time.

Am I missing something? it seems like a project that could be done for a few thousand dollars and have it work well or cobble it together for a grand.
Anyone care to pick it apart or steel the idea and build it?

[bryn]

i like it!

[Daox]

Do you think that your accessory belt can transmit that much power?

What you're thinking of doing is quite similar to Brucey's hybrid setup except he used a low voltage AC motor setup. His gains were alright. I expect yours to be better because of the lean burn engine. If you can get it to keep you in lean burn all the time you should see a decent gain.

It would be much better if you could seperate the motor from the engine though. I was just talking to Ben Nelson last night about this. I think a series wound or PM motor off of the 2nd transmission shaft (not the input shaft) of a manual transmission would work much better because it would allow you go full EV. My idea also incorperated a freewheeling clutch to eliminate brush wear when the motor isn't in use. The downside is that you don't have a transmission to assist with an additional gear reduction. But, when talking with Ben, he normally drives his Metro in 3rd gear which is nearly a 1:1 ratio.

[Ryland]

The A/C has it's own drive belt, so that is a good question, anyone know how much power it takes to turn the A/C compressor?
Power can be limited to the speed controller as well, but the Perm motors are only rated for 10hp and 19hp peek and so I'll see if I can dig up how much power a single V-belt can handle, it seems like most ridding lawn mowers and other devices with 15 to 20 hp gas engines still use V-belts to drive everything right off the engines main drive shaft.

[Ken Fry]

Originally Posted by Ryland

Anyone care to pick it apart or steel the idea and build it?

I stole the idea and built it! My POC prototype uses a couple of very similar PM motors.

The Perm motor has a speed constant of about 45 rpm/v, so at 96 volts goes about 4500 rpm. I doubt that they are able to stay together at 12,000 rpm. That would also equate to 266 volts -- far above the motor's maximum voltage rating, I'd think. I'd evaluated the Perm for my POC (and subsequent production) and it is a great motor, but brushed (a maintenance and RMI issue) and not quite the right voltage constant for my project. The Perm people (in Germany) are straight shooters and helpful, so if you want to pursue this, contact them.

My POC is 540 lb, and a motor of this size works fairly well -- albeit only with a multi-speed gearbox. (With a single reduction, getting a useful top speed means a reduction that does not yield enough wheel torque for good hill climbing.)

(Series motors and AC induction motors are a little better in this regard, but not perfect. Tesla's original idea of using a two speed box was appropriate, even with their AC induction motor. You cannot claim that 120 mph top speed is "supercar" performance: it's less than the top speed of my 2010 Honda Civic. So they went for acceleration alone, (at which it is excellent). The Tesla's peak HP would allow it to go faster than the Lotus equivalent, if it had a two-speed box. )

but as I understand it, you can program how fast things like regen ramp up and how strong, if possible it seems like two regen settings would be ideal, one for when engine vacuum is high and another for when the brake lights come on,

Yes, this stuff is very easy to implement with a PM motor. You can have any amount of regen controlled by a pot, (or a digital pot controlled by a microprocessor). For an aware driver regen under direct control of the driver (press harder on the brake pedal, more regen) is simple and works better than automated "profile" systems (econ, power, etc.) none of which know what's ahead on the road... often the most energy efficient option is coasting, maintaining hard-earned kinetic energy. There is no current system implemented in EV's that can do as well as a good driver with full control over regen amount and timing. Especially annoying to me are the systems the apply regen on trailing throttle: with each undulation in the road, the system toggles between power and regen (which feels clutzy-jerky, but also wastes kinetic energy that an intelligent driver can put to good use.) If you really need to come to a stop, regen is the way to do it, and then you want the maximum amount you can get while still being safe (not locking wheels on slippery surfaces, for example).

The idea of a series hybrid is to keep the engine running at peak efficiency or not at all. Lean burn operates only under light load, so you really don't want to help the engine stay in a light load scenario. If it is trending into that area, it should be off entirely. Most Hondas are less efficient in lean burn mode than in normal mode... in other words the actual bsfc is worse at 10 hp in lean burn that at 50 hp in stoich mode. So you want the engine always running at 50 hp or not at all. This requires series hybrid architecture, so that when the engine is running at less than 50 hp (meaning the battery charging load has dropped off) you shut it off.

The beauty of a simple series hybrid (unlike the inefficient Volt multimode system) is that the engine runs at peak BSFC or not at all. Any system that drifts away from that concept can quickly become inefficient (thus the crummy Volt gasoline fuel efficiency numbers relative to the Prius). A series hybrid does energy conversions in which efficiency is lost, so getting high efficiency requires optimizing everything, to make up for the gasoline-to-electricity-to-chemical-to-electric-through-motor losses.

Just try to model everything as closely as you can before laying out money. It is hard to match a really thoroughly-engineered hybrid (like the original Insight or the Current Prius) in terms of carrying ability vs fuel usage.

and because it's not powering the car all on it's own that seems like enough to give it an added boost to keep it in lean burn all the time

Sounds like you are thinking of a parallel hybrid. In a true series hybrid, only the electric motor powers the car. In a parallel hybird, the ICE and electric motr can work together, in parallel.

BTW, be aware that if you drive the motor at twice engine speed you will have halved its torque for propulsion purposes.

[Ken Fry]

Originally Posted by Ryland

I'll see if I can dig up how much power a single V-belt can handle, it seems like most ridding lawn mowers and other devices with 15 to 20 hp gas engines still use V-belts to drive everything right off the engines main drive shaft.

Gates has a V-belt calculator online that can come up with recommendations from their whole line. V-belts can handle hundreds of HP. Toothed belt can handle loads of HP, too, will not slip, and are more efficient than almost any alternative -- they can easily be 98% efficient.

[Ryland]

Sorry, my mistake with the title, would this be a mild hybrid then?
Yes, I realize that I would use less gas if the engine was not running at all, but without pulling the stock drive train or adding a 5th wheel or doing some funky set up with bolt on wheel motors, it seems like there is no good way to do an ad on hybrid drive train, also Honda Civic's are common enough that if this does work then there are a ton of people who could copy it.
There is a guy on EV album who has a civic cx converted to electric that getting 205 watt hours per mile under ideal conditions on pure electric power, I assume that is at 55mph or so, that would be over 11,000 watts at a steady speed or about 15hp, more then these little 10hp motors can handle, but being able to provide half of that would be a large load taken off the engine and my goal is not to build a pure electric car, I have a pure electric car already, but being able to offset my gasoline use on longer trips or when I need to haul more stuff or people then my two seat electric car will handle is the goal.
I also agree with the idea of plotting it all out on paper before putting any cash in to parts, of course all of the parts that would be used would also work well for an electric motorcycle, go cart, powered bicycle trailer or any number of other electric vehicle projects and I don't see any way that this could hurt my current gas mileage other then the added weight and V-belt, so it's only going to improve and if my other options for getting 70mpg is to spend $5,000 on a Honda Insight and be back where I don't have much more space then my two seat electric car then it seems like a worth while project to think about and well within a lot of peoples price range and skill set.

[Daox]

Any worries about brush wear when the electric motor isn't in use?

[mort]

Hi Ryland,
I don't mean to be discouraging, I'd like to see this work. But here's what I see, you are going to add a fairly large weight of batteries and a motor in order to run your engine a lower and less economical power level. The efficient motor will make up for the inefficient engine. What you do about batteries will make all the difference. If the batteries increase the car's weight by 5%, then the engine efficiency must not fall more than 10% (assuming rolling and wind resistance are equal at cruise). Without a map of the engine nobody knows. But consider that at cruise the engine might be 25% efficient, but at 1/2 cruise power it could drop to 22.5% efficient.
-mort

On reading this I think I sound too pessimistic. The engine will be making 1/2 as much power too, so efficiency could drop by 20% and you could still be ahead. It looks better to me the more I think about it.

[mort]

Originally Posted by Ryland

anyone know how much power it takes to turn the A/C compressor?

Hi Ryland,
A small car A/C takes about 5 hp at the worst, and half that in more moderate conditions. Automotive belts are in the "L" series. 3L is 3/8 wide across the wide face. 5L (5/8") are the widest Vee belts you'll see on cars. Vee belts are rated by power and load capacity. For instance the water pump is running all the time the engine is, and the pump power is related to rpm so the water pump belt has a load capacity of 1. The belt must be rated for the full pump power to have a normal service life. The alternator belt only sees maximum load 1% of the time, so it can be down rated to a fairly "weak" belt and still have a long service life. Your motor might see maximum power a few percent of the time; 10 or 20? So a belt rated at 1 HP, (I think 5L just get to 1 HP) might work OK. Before wide multi groove belts got popular some cars used 2 parallel belts for the A/C.

-mort