How permanent-magnet charging systems really work.
There seem to be quite a few misconceptions about how vfr charging systems operate. Everybody understands that they "throw away" excess power, but not many seem to realize how
little power is thrown away, at least when things work right.
Permanent magnet AC generators have some rather counterintuitive properties: The most important is that the generator tends, at normal operating speed, to be a source of constant current, not a source of voltage. Constant
voltage sources, (wall plugs and batteries) are more familiar, so the VFR charging system is a bit alien. When you short out the stator of a VFR it's doing
less work than it would be otherwise.
Hydraulic analogies are helpful in describing electrical circuits, but in this case they fall short. There's no combination of pumps, valves and pipes that acts exactly like a permanent magnet alternator. Certainly not with corresponding parts.
Still, there are enlightening similarities. For example, voltage regulation on a permanent magnet alternator is a bit like the unloader in a constant speed air compressor. The compressor is driven steadily, far in excess of the air system's needs. An extra valve in the cylinder head can "unload" (relieve) pressure in the cylinder when pressure in a storage tank gets high enough to satisfy demand. In this analogy the compressor's intake and exhaust valves are like the rectifier diodes, the unloader valve is similar to the regulator and the storage tank plays the same role as the battery.
If the unloader really vents all the excess air, with no pressure buildup, the compressor does no work when it's not needed. The usual setup is to let the accumulator cycle between low and high pressures; unloader opens when pressure hits maximum, unloader closes when pressure in the accumulator hits minimum.
There are lots of things wrong with this analogy: First, flow from a compressor does depend on speed, the current from the VFR's stator does
not depend on speed once a little past idle. Second, air is compressible and electrons, at least for purposes of the present topic, are _not_ compressible. They're more like hydraulic fluid. This makes the battery similar to an accumulator, a limitless volume at constant pressure. I'll stick with pneumatics for rhetorical purposes.
A couple of things are right in my analogy: When the unloader (regulator) "turns on" it dumps excess flow, with no pressure (voltage) buildup and hence no work (power) wasted. The vfr's regulator is upstream (before) the rectifiers and can be thought of as an extra valve in the compressor cylinder head.
Of course, neither the unloader nor the RR are perfect, and there is some waste. In the unloader case the waste is the pressure built up to push air through the valve. In the RR's case it's similar: Regulator and rectifier diodes are rather like check valves, they need some pressure (voltage) to open them. In most cases, about 0.7 volts each. About five percent of the operating voltage, so about five percent of the total output power.
Here is where things get interesting: The regulator and the rectifier
should cause exactly the same amount of power to be turned into heat. Deliver current to the electrical system, spend 5% on the rectifier, deliver current to regulator, spend 5% loss in the regulator.
When I was measuring the "unloaded" voltage on my VFR's alternator the drop (pressure loss) was more than you'd expect based on my analogy. Not a lot more, 2 volts instead of 1.4, but somewhat. This means an extra load on the electrical system will in fact reduce the heating of the RR, something other folks have noted as well.
Finally, a somewhat subtle point: Stators are designed to deliver a constant current; it's a property of the magnetic design, not number of windings. At various times folks observed what looked like overcurrent damage to RRs and attributed this to "shorted turns" in the stator. Shorted turns can't do that: they will lower the open circuit voltage but can't raise the output current, which is controlled by a ratio of two quantities; inductance and coupling between magnetic and electric fields. Shorting turns leaves the ratio unchanged and simply reduces open circuit voltage.
Replacing windings with fewer turns of lower DC resistance
will increase a stator's output current, but that's not the scenario at hand.
There's no doubt that overcurrent damage occurs on charging systems, but the only source of extra current is the battery: A leaky (not shorted, just leaky) positive output rectifier will allow battery current to flow back through the stator windings and then to ground through the negative output rectifiers. The heating effect of reverse leakage on the diode is much greater than that of forward conduction, because of the larger (~13 volts vs .6 V) voltage drop in the reverse direction. Heat tends to increase diode leakage, resulting in thermal runaway. The fractional-ohm DC resistance of the stator windings and harness connections serve to limit the current, often to a value that does not blow the main (30 amp or so) fuse. However, 30 amps through two legs of the stator is more than the windings can take.
When the leak becomes big enough (comparable to the output current of a single phase) it can latch one or more regulating thyristors. This is not an easily-achieved situation, since the electrical system voltage is likely to be lower than the regulating point by this time.
If by some twist of bad luck the regulator kicks in, even if only for a fraction of a second, the active regulating thyristor then provides a good, clean short to ground, bypassing the stator windings. This is where you can get smoke and burst encapsulating resin from the RR. There was an alternative viewpoint expressed
in an Electrex technical paper. It seems to have gone away, I can't find a replacement on the Electrosport website. Perhaps they agree
. bob prohaska
