I've thought about this many times over the years, but I doubt I'll ever have time to really complete something like this. Here's what I came up with:
Alternator conversions are a bad idea, even though GM ecoassist uses it. It's tempting because it already fits the engine and the windings are pretty beefy.
Alternators have a lot of poles and also spin at high speed. That means a 100k erpm controller might not even be able to keep up.
Alternators also have thick laminations for cost savings, since they have a weak rotor field. With a permanent magnet, you'll be incurring very high iron losses.
The battery pack will get expensive fast as you try to add capacity, so I think the best solution is to buy used lithium titanate packs for maximum power per dollar.
With LTO, you can do full cycles all the time without worrying about degradation, you can use it in the winter without heating it up first. These are a really great start, IMO:
https://www.greentecauto.com/hybrid-...te-li-ion-scib
You can squeeze 10kW out of a single one of these no problem. Double up and it might even not need cooling.
Drive options:
You can add a separate toothed pulley, or add a second tensioner to the existing serpentine belt. I haven't thought much about it, but I assume you just need a bracket to hold your motor, a ball bearing that can take ~100lbs, some underdrive alternator pulley from any car that you adapt to your motor shaft, and you're good to go.
The load on the alternator pulley with the alternator field current at its maximum is something like 4Nm (that's like a little over 2hp at 2000rpm at the alternator). With a 60mm alternator pulley diameter, that's around 70lbs of tension on the belt from the OEM.
If we use a larger pulley that's say 90mm with a fairly typical 135mm crank pulley diameter for a 1.5 reduction ratio, then 6Nm at the "alternator" or 9Nm at the crank is a similar load on the belt in tension. That's pretty conservative though, because the amount of belt contacting the alternator is now quite a bit greater, and belts usually don't fail from snapping. We could probably go up to 15Nm at the crank, which is a pretty respectable amount of torque on a small 4 cylinder.
Most modern starter-generator setups use dual tensioners. I think if the length of belt between the crank pulley and alternator on the side that doesn't already have a tensioner is short, you might be able to get away without a second tensioner.
As far as the motor itself, I think you want a sensored brushless for cost. In terms of motor options, there are really cheap RC motors from Hobbyking, but I found these that look really nice and are not expensive:
https://alienpowersystem.com/product...shless-motors/
You'd want the motor to be able to spin to ~10-15000rpm like an alternator does, max current of 200A seems pretty adequate. The biggest watercooled inrunner they offer looks pretty suitable for mild hybrid. You'd want around 150-220kv to match the 52-68V range of the 24S LTO pack.
As far as controllers go, I'm not very familiar with what's on the market, but Kelly was the name that kept coming up years ago. The prices seem pretty reasonable to me, but I wouldn't know any better.
Finally, in terms of actually hooking the controls up to the car, this is what I was thinking:
-Regenerative braking is controlled by engine vacuum. High vacuum = more brake, low vacuum = less brake.
-Install either a second throttle attached to the throttle pedal, or intercept the existing throttle position sensor and turn that into a signal to the motor controller. You only need to engage the electric power at higher throttle positions because you want the gas engine doing more work to improve its efficiency.
-Tap the brake lights for a "maximum regen" switch.
-(optional): have the motor provide some assistance to the starter when cranking. You may be able to crank a warm smaller engine by stabbing the throttle pedal and engaging the electric assist alone.
Today
Other popular topics in this forum...
)
