B&C Permanent Magnet Alternator - Compass Interference?

[Canadian_JOY]

A local EAA chapter member has installed a B&C permanent magnet alternator in his aircraft. This unit is similar to the SD-8 model popular on many all-electric RV's, however it is a 30-amp rated unit for installation on a Continental engine (gear drive, mounted on rear of engine case). The alternator works as advertised, except that it seems to cause his compass to literally spin at certain engine RPM settings. Since I've never had a chance to work with B&C permanent magnet alternators I thought I'd throw the question out to the more experienced users here.

Has anybody had similar problems with SD-8 or other PM alternators? If so, what's the cure?

These questions were put to Bill at B&C but we were surprised to find he had no response short of suggesting surrounding the alternator with a tin can or mu metal.

Thanks in advance for your considerate replies.

[elippse]

[quote=Canadian_JOY;246999]A local EAA chapter member has installed a B&C permanent magnet alternator in his aircraft. This unit is similar to the SD-8 model popular on many all-electric RV's, however it is a 30-amp rated unit for installation on a Continental engine (gear drive, mounted on rear of engine case). The alternator works as advertised, except that it seems to cause his compass to literally spin at certain engine RPM settings. Since I've never had a chance to work with B&C permanent magnet alternators I thought I'd throw the question out to the more experienced users here. QUOTE]

I'm not sure about the alternator, but I can tell you that cables carrying lots of current to loads can effect the compass by several degrees if not separated by a good distance. I had a cable running along the bottom of my instrument panel which was a good 8" below the compass and it would cause up to 7 degrees error. I then moved it down to the bottom of the fuselage where it would still pull the compass about 2 degrees. If the compass spins at certain engine rpm, that could be a result of the rotating magnetic field. I have one of these alternators on my bench I've been using for tests of a buck-boost regulator to get more output at low rpm, and I wonder if it might be something associated with the phase-controlled regulator rather than the alternator. Is the regulator or its cabling near the compass?

[Doug]

Hi CanadianJoy(),
Mu metal is very effective in this situation. Forget the tin can. Perhaps the tin can comment related to how to form up the mu-metal.

I have to say I am very surprised that the stray magnetic field had not already been managed in the SD-8(?) alternator itself.

Can this compass behaviour be attributed entirely to the new alternator?

I suggest removing the alternator to confirm this since it is conceivable that the surrounding metalwork (canopy frame, engine mount etc) may have indvertently been magnetised during installation.

De-magnetising the metalwork is not a big deal.

Doug Gray

[Canadian_JOY]

Thanks for your input so far, folks. I should mention the alternator in use is the BC433-H, which of course is just a bigger (30 amp) version of the same kind of technology used in the SD-8.

Paul - Compass is located atop the aluminum glareshield (where it's been for the last 20 years), and the power wiring to the buss enters the cockpit at the top right corner of the firewall, about 2 feet away. Alternator, regulator, OV protection etc are all located on the engine side of the firewall, approximately 2 feet forward and 18" below the compass location, with the firewall between them. Engine cowling is composite. Your comment has prompted me to suggest an overall braided shield on the B+ lead from the regulator to prevent more steady-state compass deviations and, of course, reduce or eliminate any EMI coming from this lead. Unfortunately the compass spinning at a several-RPM rate is probably not going to be improved by this shielding if my understanding of magnetism is anything close to reality.

If you're testing the SD-8 on the bench, would it be possible to take a hand-held compass and see what happens as you spin the SD-8? Would be interesting to see if a similar effect is noted.

Doug - compass behaviour is entirely attributed to new alternator as the previous Delco generator was flown for approximately 20 years with the compass in the same position. The B&C alternator is a direct bolt-on replacement for the Delco generator, and is in fact a 3-phase dynamo.

Given that the compass swirls around wildly at specific engine RPM (1200-1400 RPM) I suspect it's not a case of having the engine mount or similar metal structure magnetized as that would tend to provide a steady-state compass error, rather than a "spinning" compass. I could well be way off base on that assumption...

[elippse]

Originally Posted by Canadian_JOY

If you're testing the SD-8 on the bench, would it be possible to take a hand-held compass and see what happens as you spin the SD-8? Would be interesting to see if a similar effect is noted.
...

I'll give that bench test a try.
In order to get total magnetic shielding, there has to be a complete, closed path around the conductor. Consider a circuit in a metal box that connects to a circuit inside another metal box that has the connecting conductor inside a tube (shield) that is connected to the metal boxes at each end. Everything is totally enclosed! 'Get the picture? That way any incident varying magnetic field will cause a current to be induced in the shield which will in turn oppose the incident field. This shielding does not require ferrous materials, only materials with high conductivity such as copper to complete the current path with minimum IR drop.
Most shielding is only for electrostatic (capacitive) coupling. Twisted pair conductors provide both electrostatic and magnetic interference rejection since the current generating the magnetic field flows in opposite directions, effectively cancelling any magnetic radiation. The same can be said for an incident magnetic field illuminating the twisted pair, since the currents induced in the two conductors will cancel at the load end (common mode). Since it is possible for one of the two twisted conductors to be a little closer to a sensitive load, capacitively coupling to it, the TWP is often encased in an overall shield to further eliminate this coupling.