fot
[Top] [All Lists]

RE: Alternators and Master Cutout Switches

To: "'Randall Young'" <ryoung@navcomtech.com>, fot@autox.team.net
Subject: RE: Alternators and Master Cutout Switches
From: Bill Babcock <BillB@bnj.com>
Date: Wed, 7 May 2003 15:28:29 -0700
I suppose it makes sense, but I remember an SAE paper talking about
reliability of internal regulators for alternators that showed a maximum
voltage of 140V for a disconnect fault from 6000 rpm that persisted for
less than a millisecond, followed by a decay from 40V to less than 12V
within 20 milliseconds. The paper concluded that any reasonably designed
regulator could handle that for the worst case fault, which is considered
a loose nut on the output allowing vibrational connection and
disconnection. 

These are the kind of things that stick in my mind--but I don't remember
my daughters birthday. I also vividly recall the decay chain for I131. 

I think the voltage curves from this paper are considered the standard for
this kind of transient--I've seen them reproduced in other places. You
could probably find it on the web with a reasonable search. 

If such a fault was truly deadly to the regulator then every time an
output nut got loose an alternator would toast--or at least most of the
times. The output wires from alternators are reasonably heavy and stiff,
making it relatively common for the them to loosen their connectors.
Manufacturers warn against all kinds of abuses of their products that
don't have real effect--it's generally a preemptive strike to minimize
warranty claims. 

A peak of only 140V is not sufficient for an arc across any switch
designed for even 12V. The only electronics that would be affected by a
properly wired straight disconnect would be those on the alternator side
of the switch, which should be only the regulator and the rectifiers.
Certainly there is no problem with having a switch that connects a
resistor before opening the circuit, but it sounds very
belt-and-suspenders to me. You also might be better off with a large
capacitor (something in the range of 1uf at 480V) than a resistor, since
it would be open circuit to steady DC and would be unlikely to present a
hard-to-diagnose switch or wiring failure. You wouldn't't even have to
have a make-before-break switch to use it. It might be what they use
internally to limit transient voltage, though a high voltage zener is more
likely. 

-----Original Message-----
From: Randall Young [mailto:ryoung@navcomtech.com] 
Sent: Wednesday, May 07, 2003 9:24 AM
To: fot@autox.team.net
Subject: RE: Alternators and Master Cutout Switches


> Most modern-day electronics are designed to handle reasonable 
> faults--losing power suddenly is certainly a reasonable fault. Really 
> all you need is to open the output wire of the alternator and open the 
> battery circuit. You can do that with any of the cutout switches that 
> have a heavy pole for the battery disconnect, and a second lighter 
> pole for the alternator disconnect.
>
> You won't get a power spike delivered to the electronics from the 
> alternator if it's disconnected. You might get a higher voltage 
> delivered to the internal regulator of a one-wire alternator, but 
> that's what it's there for and they can handle it.

Bill, I beg to differ.  All modern manufacturers warn against doing this
very thing, because if the alternator is running at full output, the
regulator cannot shut it down fast enough to avoid a very large voltage
spike.  Even with no current through the winding, the magnetic field in
the rotor takes time to decay.  And, if the regulator does fail, the
output voltage can potentially rise to lethal levels, more than enough to
arc across the just-opening contacts in the kill switch and fry that
expensive computer.

Of course, one very rarely operates the cut-off with the alternator at
full output; and generally when that does happen, preserving the
electronics is not the primary concern.  So, you may get away with it for
a long time and never have a problem.  But Steve's idea of shutting down
the alternator (with a "make before break" switch) is sound practice.

Here's a switch I found on the Internet just now.
http://www.racerwholesale.com/Merchant2/merchant.mv?Screen=PROD&Store_Code
=R
W&Product_Code=ELE-28606

If you can't cut&paste that link together, go to
http://www.racerwholesale.com click on "Battery cutoffs" on the left and
then "Battery Cut-Off switch - Alternator kill" P/N ELE-28606

NFI, not even a customer, just something I found on the web.

Randall

<Prev in Thread] Current Thread [Next in Thread>