Alternator modifications

[ConnClark]

Ummm... mechanical loses from studies actually show mechanical loads only making up about 5 to 10% of the total load an alternator places on the engine.

I don't buy disconnecting the belt from the alternator getting someone %5 more mileage over a kill switch. The facts just don't support it.

Once again this is a thread on alternator modification and not an alternator disconnect, kill switch, or solar augmentation. There are other threads on ecomodder for that.

[ConnClark]

Since my simulations keep running into jacobain errors I decided to look up a few patents. here is a circuit from patent application US 2004/0232538 A1

You would need one of these circuits for each phase of your alternator .

[George Tyler]

as that document shows, the diodes are a big loss at high currents, at low current there is the magnetic losses and also the fan etc. I have been thinking of doing the mosfet rectifier thing , but using an international rectifier chip that does the same thing.
de-energising the alternator removes the magnetic losses, but the windage losses are still there. Some cars now have electromagnetic clutches on their alternators, they would not do this if it did not reduce consumption. Failing this, how about removing the alternator fan and using a controlled electric fan instead, and also disconnecting the feild winding when it would help?

[oil pan 4]

Some alternators have internal fans. I was working on a denso alternator on a mitsu 4cylinder diesel yesterday, noticed it had an internal fan.
AC delco style with the external fan can be removed very easy.

Looks like I am suffering big diode losses in my truck.

[t vago]

Originally Posted by ConnClark

Since my simulations keep running into jacobain errors I decided to look up a few patents. here is a circuit from patent application US 2004/0232538 A1

You would need one of these circuits for each phase of your alternator .

How do you propose to test your alternator circuitry? Are you going to have a test stand like what was in one of the papers of this thread? Or are you going to do an on-vehicle test?

[ConnClark]

Originally Posted by George Tyler

as that document shows, the diodes are a big loss at high currents, at low current there is the magnetic losses and also the fan etc. I have been thinking of doing the mosfet rectifier thing , but using an international rectifier chip that does the same thing.

Do you have a part number for this chip? The only thing I have seen is reference to something they designed for another company which isn't for sale to others.

de-energising the alternator removes the magnetic losses, but the windage losses are still there. Some cars now have electromagnetic clutches on their alternators, they would not do this if it did not reduce consumption. Failing this, how about removing the alternator fan and using a controlled electric fan instead, and also disconnecting the feild winding when it would help?

Actually most alternator designs use Pulse Width Modulation changing the percentage of on time to off time to vary the field strength. This way it controls the output voltage of the alternator and creates only as much electrical/magnetic related drag as needed. This is why an alternator is more efficient than a generator.

From the papers I have read so far mechanical losses and fan losses only make up 5 to 10% of total losses in an alternator. However I agree with you and do think that there is room for improvement. I'm not so sure an electrical fan is the way to go. You want it running almost all the time if not all the time to keep the copper coils cool and thus keep their electrical resistance low. An electric fan is going to suffer an penalty efficiency drop equal to the drop of the alternator loss itself vs a directly driven one. Looking at most alternator fans its obvious that their shape could be improved quite a bit.

You mentioned some cars now have electrical clutches on their alternators. Do you know of a particular model? I might be interested in getting one or two from a junkyard or maybe new if I have too for experimental A/B testing.

[George Tyler]

there are a few cars with clutches on the alternators, one is the GM opel/holden Vectra, not sure which year. An Alternator generates a constant 15V or so, and Flux density of the field depends on this and is inversly proportional to the speed, so it decreases with increasing speed. However losses at a constant flux increase with speed, so the relationship is complicated as these are not linear. Magnetic losses don't change with load if everything else is constant.
My thinking with the electric fan is that it is controlled with the rotor current, maybe just driven off the AUX terminal in the alternator that powers the voltage regulator, then when it is charging it will run, that supply to both regulator and fan is disconnected via a relay to disable the alternator.
I have tried mechanically disconnecting the alternator with the thermo fan and P/S with good results, in another post. I have not found a practical way of supplying power to the battery though.

[George Tyler]

they have a range of chips for this:

http://www.irf.com/technical-info/appnotes/an-1087.pdf

https://ec.irf.com/v6/en/US/adirect/...N=0+4294841695

[ConnClark]

Well after a little research into these clutched alternators I found that they are just free wheeling clutches. They only disconnect the alternator when its spinning faster than the pulley, This was done to save wear and tear on the belts and tensioners not to save fuel.

[ConnClark]

Originally Posted by George Tyler

they have a range of chips for this:

http://www.irf.com/technical-info/appnotes/an-1087.pdf

https://ec.irf.com/v6/en/US/adirect/...N=0+4294841695

Thanks.

I have found the Linear LTC4354 IC. LTC4354 - Negative Voltage Diode-OR Controller and Monitor - Linear Technology

I figured out a way for it to do the high and lowside control. Right now I'm working on a way to get it to handle driving a higher gate capacitance.