Index Home About Blog
From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: Generator problem
Date: Sun, 06 Aug 2000 05:36:06 -0400

James Thorne wrote:

> Ok,  the saga continues....
> First, let me thank everyone for the help, especially Alan for
> the detailed 'how to's'.  Please do not take the rest of this
> email in any negative way what so ever.  It is not meant that
> way at all.  Just want to understand what I am seeing.
> Got the gen set out of the Bounder and disassembled the gen end
> per Alan's through instructions.  Did not find what we expected
> when we got the housing off.  To look things over, the rotor
> looks fine.  No obvious defects.  Still measures 2.8 ohm through
> the windings.  Now the new rotor measures 7.8 ohms.  This is far
> below what Alan said it should read but still 4 times the
> existing rotor.  Anyone care to help me understand?  Is the new
> one defective too?  Looks good, obviously new from the flywheel
> to the bearing, pretty and shiney.

Let's do some math.

The generator output voltage is a function of the rotor speed,
amper-turns (number of turns times the amps through the turns) on
the rotor and stator, among others.  Since the rotor speed remains
constant, we can drop them from our figuring.

Your old rotor measures 2.8 ohms and the new one 7.8.  You had
between 35 and 50 VAC with the old rotor.  let's assume a partially
shorted winding.  Let's further assume that the regulator was
ramping up and down fairly smoothly (makes the math easier.)  We'll
therefore average the output to 42.5 volts.  If there is a
turn-to-turn short, then the field current will bypass the turns
that are shorted.  The resistance of the rotor is roughly
proportional to the number of turns for a given wire size so the
resistance should vary linearly with the turns present in the
circuit.  Therefore we should be able to estimate the number of
turns shorted by the ratio of the resistances of the good and bad
rotors.  That is 7.8/2.8 = 2.79.  Since the AC output is
proportional to the amper-turns, everything else being constant, we
should be able to use this ratio to compute what the new rotor would
output.  42.5 volts * 2.79 = 118.4 volts.  Hmm, pretty close to
120!  This looks good.

But!  We have a problem.  Alan said that the full load field
excitation should be around 85 volts.  85 volts into 7.8 ohms is
(v-squared over r) 926 watts and 11 amps.  This seems way too high
to me.  That's almost 13% of the generator's (and engine's) output
at full load. Indeed, that would take over a horsepower just to
generate the power the field uses.  I have seen alternators that
would go as high as 10% field excitation but never anything more and
then generally on old, low speed, open frame units. That also seems
like a LOT of heat to get rid of in the poorly cooled rotor.  My
generator's stator frame barely gets warm to the touch at full load
so I doubt that almost a kilowatt of heat is being produced in the
field alone.

So.  Using the rotor resistance ratios, I gain confidence in the new
rotor being good.  Alan also stated that the rotor should have about
25 ohms.  85 volts excitation into 25 ohms 289 watts and 3.4 amps.
That sounds about right.

So, we have a conundrum.  Your actual numbers check out with the
math and yet they don't match Alan's numbers, which themselves seem
OK.  My guess is that you're both correct - that the new rotor is
good and Alan's looking in a book for a different generator than
what yours actually is.

The Onan AJ in my rig excites the field with around 35 volts
unregulated, if my foggy memory serves me.  35 volts into 7.8 ohms
is 157 watts and 4.5 amps.  Amazingly enough, that's about the same
amps my field pulls!  On your generator, that amount of field power
would be a much more reasonable 2.5% of output. I'm thinking that
you may have a low voltage excitation system in your generator for
which the new rotor resistance would seem appropriate.  I bet Alan's
looking at the service manual for a high voltage, high rotor
resistance generator.  At least that's my theory.

> That's not all we found.  The stator is not in such good shape
> and looks like a real problem.  The face of the windings near
> the flywheel where all the wire leads are connected to the
> windings inside insulated sleeves and tied up with nylon string
> is black and crispy.

That may or may not mean anything.  I'm seeing a lot of windings
coming from the factory, especially the Class H and above high
temperature ones, looking burned.  The varnish is deep brown to
black and the varnish makes the tie string brittle.  Does the
winding have a burned smell?

> Most of the tie strings are burned away
> and the windings are black in several areas.  There is also a
> spot on the flywheel where an arc has obviouly struck between
> the windings in the stator and the flywheel.

This is interesting.  How far away is the flywheel from the
winding?  Unless there can be physical contact, that sounds like
lightning damage.  Especially since you say that there's a SPOT on
the flywheel.  That indicates that it was not turning when the
flashover occurred.  Had it been turning and come in contact with
the stator, you should see a ring, or at least an arc of burn marks
around the flywheel.

> Measuring the
> stator windings, they all measure ok with no shorts to each
> other or to ground.

That really doesn't tell you it's good per se.  It only tells you
that it might not be bad.  If it has an inter-turn short, it will
smoke quickly when started.  A megger check from coil to ground
would tell you more.  Only way to tell for sure is with a Baker
Surge Tester.  This thing hits the winding with a high voltage pulse
and then lets you look at the ringing on a built-in O-scope.  A
shorted turn will greatly dampen the ringing.  This is a standard
piece of equipment in a motor shop.

> From the looks of things, if it is not bad
> now, it will be very shortly.  I ordered a new stator.  Southern
> Plains Cummins did not have that part in stock but is getting it
> for drop shipment on Monday at a cost of $445 including the
> express delivery charge.


> My course of action is pretty clear here.  Replace the stator
> along with the rotor, carefully reassemble and enjoy a fully
> rebuilt generator.  There are not any other parts to go bad on
> the electrical side.

I think that's the correct route at this point.

> What I am looking for here is some help in understanding what is
> going on and a confirmation that the new rotor is likely ok.  Is
> this an overheating problem?

From my calculations above, I'm gaining confidence that the new
rotor is correct and OK.  I suspect that if the stator winding is
burned (and not just black varnish), that the scenario might be as
follows:  You took a lightning hit at some point which flashed the
stator to the flywheel.  This flash made a weak spotweld between
turns.  When the generator was next cranked, the shorted turns did
two things. One, they heated vigorously from the heavy circulating
current.  Two, the reduced output voltage caused the regulator to
over-drive the field, damaging it.  At some point either the heating
or the vibration or both broke loose the spotweld and the generator
continued working.  However, the insulation on the field was damaged
and eventually it failed, allowing a short between turns to form in
the rotor.  That's when you lost output.

Had the RV been in a thunderstorm sometime in the fairly recent
past?  Nearby lightning strike?  And perhaps did the generator seem
to have low output (slow fan on the AC, etc) at some time in the
recent past?  If you can recall either or both of these events
happening, then I'm conformable with my theory.  You might not have
noticed the low output voltage if the stator short only involved a
few turns.

> If so any ideas on how to keep it
> from happening again.  Or is it just insulation breakdown from
> years of disuse, wet/humid environment etc.  I know one thing
> for sure, next time I do an oil change, I am not going to clean
> the genset compartment with engine cleaner.  This could very
> well have been what set this off.  Likely would have failed
> eventually on its own but.....

I don't think that the engine cleaner could possibly cause the
flashover but it very well could degrade the windings.  Most (all?)
of these new non-petroleum-based degreasers have both a very strong
detergent and caustic soda in them.  Either would play hell on the
varnish of the windings.  Indeed, at the motor shop, we use a
boiling solution of concentrated caustic soda to strip the windings
of insulation on certain classes of motors in preparation for
rewinding (most just get burnt out in a furnace).

> Thanks and will follow up as things progress.

Please do.  I really enjoy these remote diagnostics sessions.  I'd
love to know how it turns out.  I'd particularly like to know the
field voltage at no load and full load.

Finally, I really, really would like to take a look at the old
parts.  I'm really interested in looking at the stator where it
flashed.  The motor shop is certified to do insurance examinations
for lightning damage so we should be able to make a pretty positive


Index Home About Blog