Neosophist wrote:Computer power supplies have big fans and big heatsinks inside them, they still generate and burn off excess heat.

Yes, thats implied when I said SMPSes are generally 70-94% efficient.

Neosophist wrote:The altenator is generating AC voltage that varies between 50 - few hundred volts

Its true that the alternator on a bike has a wider voltage range than household electricity. But many computer SMPSes can take from 90-250V input. Thats a fairly wide range.

Neosophist wrote:you still need to regulate this variable voltage to a stable voltage.

Thats what a regulated SMPS does.

Neosophist wrote:Adding further circuitary to make it switched mode would be pointless

Its not pointless.
The regulator output on many VFRs can do 350W.
Which means the alternator is putting out even more power than that.
So lets be conservative and say the alternator is putting out 400W of energy.
As I measured a few days ago, my NC24 requires around 4A to run (headlight off)
Thats 57W (at 14.3v)
So 400W-57W is being wasted. 343W, thats 46% of 1hp wasted with a linear (normal) bike regulator.
Lets be conservative and say the SMPS regulator would be 80% efficient providing 4A.
So it would draw 71.5 to provide 57W meaning it wastes 14.5W
Net power saving of the SMPS would be 343W-14.5W = 328.5W which is 44% of 1hp.

So using a SMPS regulator would save 44% of a horsepower. So the bike will be faster and use less fuel.
As far as I know 99% of people who ride bikes would like those benefits.
And the extra cost is irrelevant if the owner can spare a few extra quid temporarily because the savings in fuel would pay for it pretty quickly.

To reduce the complexity & cost of the SMPS: a combination of linear and SMPS techniques could be used in the power supply.
For example lets say with no load the alternator floats at a few hundred volts as you said.
If you pull 4A from it it would be at a much more managable voltage, lets just say for argument purposes 50V.
A linear regulator (like some kind of a shunt circuit) could be used to cap the output of the alternator to 50V DC.
So most of the time the shunt would not even be operating, but it would protect the SMPS from input voltages higher than 50V reducing the complexity and cost of the SMPS part.

In conclusion its not complicated, and theres significant power to be saved. Power that costs fuel, power that could be used for acceleration.

Hi,
i did look into building a SMPS when my rec/reg went on my nc30, you can actually buy linear regulators that can handle 10amp with a 50v input and 12v output, but they cost about £15 and you'd need about 4, and for the £25 you can get a R6 rec/reg one that'll work without having to pratt about.
I do think they would be the way forward from a effiencey point of view there much better than the old tech shunt regulators we all use,(and you wouldnt need much( if any) smoothing on the output side) but i think size is always going to be the major problem, and if its not broke then dont fix it so i dont see them been used in the near future.
Maybe if i ever get some time i will build one just to see if it'll work, but i need to get my cbr project home before i can do anything!!!
Cheers
Ian

Morespeedvicar wrote:Hi,
i did look into building a SMPS when my rec/reg went on my nc30, you can actually buy linear regulators that can handle 10amp with a 50v input and 12v output, but they cost about £15 and you'd need about 4, and for the £25 you can get a R6 rec/reg one that'll work without having to pratt about.
I do think they would be the way forward from a effiencey point of view there much better than the old tech shunt regulators we all use,(and you wouldnt need much( if any) smoothing on the output side) but i think size is always going to be the major problem, and if its not broke then dont fix it so i dont see them been used in the near future.
Maybe if i ever get some time i will build one just to see if it'll work, but i need to get my cbr project home before i can do anything!!!
Cheers
Ian

The Generator generates several hundred volts at high RPM though.

Building a switching regulator that can handle the rapid changes in voltage you get with a motorcycle generator, has a low rfi and half decent power factor would be huge and cost a lot, switching regs don't handle rapid changes in input too well.

A liner-reg/rec may be inefficient (circa 40%) However its relativly cheap and effective at the job it performs.

If you want an extra .5hp remove the lash gears from your cams :D

Are you basing this fuel saving on the effects of the smaller magnetic field created in the alt windings when less power is drawn ?

The magnetic field causing drag on the crank. Less drag, higher efficiency ?

Old Painless: When there is no load on the coils the magnets can pass by freely, but when there is a load on the coils (more current flowing) the magnetic field from the coil resists the movement of the magnets.
So it creates more load on the engine.

I've even wondered for a long time if it would be cool having a 'boost' switch in cars and bikes. When you're about to have a road race, you can temporarily disconnect your reg :)
I would have liked that the last time I raced an NC30 and we were neck in neck hehe.

The alt makes as much power as the bike needs, i think you may be over-thinking things slightly.