Everyone seen this? More efficient alternator

[serialk11r]

Originally Posted by ConnClark

Yes, basically I am making something that will allow you to improve your alternators efficiency.

Bump, @ConnClark what happened with the synchronous rectifier?

I see some driver ICs out there that make it pretty easy to build, but does anyone know of any synchronous bridge rectifiers you can just straight up buy ready to go?

I'm thinking, add 10% more efficiency to the alternator, you get an extra volt or so to play with at idle, so a 10% underdriven crank pulley (maybe just machine the original a little) can be used to slow the water pump a tad (0.9^3 is 0.729, 27% less power!) and the average savings should be something like 200W crank power saved at cruise between the alternator and water pump, which is a solid gain for a reasonable cost vs. buying overdrive alternator pulley, underdrive main pulley, etc. or electric pumps and such.

[ConnClark]

I got caught in a layoff of with about 1/4 of the company I worked for. I haven't had the means to pursue it further.

Playing with an under drive pulley would throw off some of the benefits of a lower field coil current. There might be some gains to be had though. Its not something thats going to be a straight forward adjustment.

[serialk11r]

Originally Posted by ConnClark

I got caught in a layoff of with about 1/4 of the company I worked for. I haven't had the means to pursue it further.

Playing with an under drive pulley would throw off some of the benefits of a lower field coil current. There might be some gains to be had though. Its not something thats going to be a straight forward adjustment.

Aw, sorry to hear. What driver and FETs were you using? Mind sharing the schematic? I think a Denso square conductor alternator (~20% higher fill factor) and synchronous rectifier could return a reasonable 70% efficiency which is not too bad.

Yea I realize that, but it's not like reducing load at idle saves much fuel anyways, and even a small reduction in the speed of the alternator and water pump helps a lot for power savings under typical usage and high rpm since power is a cubic function of speed.

[serialk11r]

Can any people who are good with circuits verify that all you need to build an active rectifier bridge would be comparators and MOSFETs? I really want to try this! Haha.

EDIT: Okay my research brings me to the following specs for a comparator:
Minimum supply voltage <= 6V (so it still functions when the engine is cranking under low battery), max supply voltage >16V (maybe good idea to pick the higher voltage ones to prevent damage from surges?), open drain or push-pull output, and 125C temperature rating.

http://www.st.com/web/catalog/sense_...S1089/PF254080
STMicroelectronics TSX393 fits the bill? 2nA output seems too small, is push-pull better since it has more output current?

[serialk11r]

@ConnClark

Was thinking about this again recently, and I came across this after many hours of reading:
Diodes, Rectifiers - Single | Discrete Semiconductor Products | DigiKey

Product sheet: http://www.ti.com/lit/ds/symlink/sm74611.pdf

26mV forward voltage drop, 0.3uA reverse leakage, 15A recommended continuous current diodes for $3.69 each! They can take 24A but the voltage drop increases a bit. They're synchronous rectifiers in a single package! The Microsemi chips that are the next lowest rated Vf on Digikey are said to be active rectifiers (LX2400ILG, out of production replaced with the higher rated LX2410A). Those chips happen to also be rated for 225C which is good for an engine bay.

I think my car has a 130A alternator, each phase is putting out 130A at its peak over a brief moment, so it would take 5*2 per phase * 3 or 30 of them I don't really see this car consuming 130A but it could probably get close thanks to the electric power steering, though I've probably shaved off around 10A with LED bulbs around the car. With my MR2 and its 80A alternator, using 18 of the chips seems a little more sane.

Anyone want to try these out?

[oil pan 4]

Seems like a lot of maze for very little cheese.

You can get 100 volt 160 amp rated pieces for, not real cheap.
STPS160H100TV

Problem with Schottky diodes is they really don't like to be hot.
Between 50'C and 100'C they lose half their power carrying capacity.

These 160 amp diodes really are only good for 140 amps, if I can keep them at 50'C (as in they can not go in the hot engine compartment).

Typically alternators electronics are rated for up to 200'C.

Why stop at Schottkying the main rectifier, why not replace the half wave voltage rectifier for the regulator?
That should decrease the RPM at which the alternator comes up to the desired voltage and increase its output.
I like this mod for those reasons.

[serialk11r]

Originally Posted by oil pan 4

Why stop at Schottkying the main rectifier, why not replace the half wave voltage rectifier for the regulator?
That should decrease the RPM at which the alternator comes up to the desired voltage and increase its output.
I like this mod for those reasons.

Those aren't Schottkys, they're active rectifiers like I said! Look at the spec sheet.

And yes, more power from the same alternator particularly at low speed is the strength of this mod.

I took a look at the alternator on my car, it seems to be the 127mm stator 130A Denso "hairpin" wind that is fairly common nowadays, with higher copper fill but normal diodes and giant heatsinks to cool them. All Toyotas use it.

On German cars the standard these days seems to be a Bosch "High Efficiency Diode" (aka Schottky) alternator. A Passat for example has a very short black casing around the end of its alternator due to the lower heat output from the Schottkys.

The benefit of the Denso approach is that at high amperages, copper losses are cut significantly. However at idle, I don't think the alt is pumping out anywhere near half its rated amps, and the voltage drop due to stator winding resistance can't be that high, but the drop across the diodes is huge.

I plan on switching my FRS for something else, so I'm not going to touch the alternator, but I've been looking into hypothetically hybridizing a car and I was thinking that keeping a smaller alternator around for redundancy and extra low rpm charging capability might be a good idea, while shaving off a bit of high up mass from the engine.

[pete c]

Seems to me that the only was an alternator with a substantially higher weight makes any sense is if you have...

a. increased battery capacity with the ability to charge at a very high rate

and...

b. a smart charging controller which would allow you to go to a high charge rate when you wanted to slow. MAybe even have the ability to go to a low level of power output when acceleration was needed for short periods

[serialk11r]

Originally Posted by pete c

Seems to me that the only was an alternator with a substantially higher weight makes any sense is if you have...

a. increased battery capacity with the ability to charge at a very high rate

and...

b. a smart charging controller which would allow you to go to a high charge rate when you wanted to slow. MAybe even have the ability to go to a low level of power output when acceleration was needed for short periods

Yep, agreed. I think nowadays they just slap a big alternator onto everything so they don't have to make as many different ones to save cost, and so that the idle speed doesn't have to increase as much for turning on electrical components.

Since there's a battery in the system anyhow to provide intermittent power, it seems more logical to size the alternator for typical operation, which is generally in the 30-50A range I think. If I weren't so concerned about resale value I would definitely try to experiment with a 6-7lb 80A alternator or something like that, because as far as shedding weight from a car goes, 100 bucks to drop 7 pounds off the top of the engine is pretty darn good. Lighter wheels, lighter battery, and lighter exhaust are all around that much or more per pound lost.

[Xist]

If the manufacturers save money using fewer different alternators, are the replacements any cheaper?