Suggestions on a Honda charging system

OK I've got one here that's just got me stumped. Maybe some
old-timey Honda mechanic might be able to help

1980 Honda 750K.

I put a new battery in it start of the season, and I ride it most
every day. The problem is that after about a week of riding,
the battey has very little charge in it - just enough to start the
engine if the engine starts in about 3 revolutions. If the engine
doesen't the starter gets slower and slower and stops. Then
the bike will not start.

If I pull the battery and put a fresh one in, the bike starts
immediately. If I leave the fresh battery in, the bike works
fine for a week then the battery starts getting lower and lower
charge in it and the same thing happens.

I've traced the problem down to the bike isn't charging the
battery unless the engine is 2500rpm or more. I see the
battery terminals jump about a half-volt when the bike is
started, to about 12.5 volts, but I don't see the voltage hit
13.8 at the battery terminals unless the engine is revved up.

At idle, I measure almost a full 12 volts going into the field coil
and about 9 volts AC leg-to-leg on all 3 legs coming out of the
alternator.

I've replaced the regulator/rectifier with no change in
behavior. On both units, the diodes passed the ohmmeter
check.

Both rotor and stator pass the resistance checks in the
FSM. There are no shorts to ground on either one. This
is true whether the bike is cold or warm. I've cleaned all
the electrical contacts. The rotor brushes both have plenty
of material left on them. Power output of the alternator
does not change whether the bike is warm or cold.

My guess is that either the rotor or stator has some windings
that have broken down. Unfortunately, the FSM has no
listing of what voltage is supposed to be on the field, and
stator, at a given RPM.

Ted

 

You might get more suggestions if you describe what type of alternator it
has. There's plenty of electrical people about who don't know your
particular bike.

 

I like that chart, somewhat similar to the diagnosis instructions at

http://home.earthlink.net/~trinomial/DOHCcharge.html

which links from http://www.motorcycleproject.com/

The author, Mike Nixon, helped me with a similar problem on a 750C. I
notice both his write-up and the electrexusa.com chart say to check the
rotor before the stator. In my case, swapping the rotor fixed it,
luckily we had a parts bike. We also swapped the aluminum cover - you
could see the scuff marks on the bad rotor, it seems that the cover
housing on these engines can easily distort even if the bike is just
leaned against a wall, let alone a tip-over, causing rotor mis-alignment
and scraping.


Another tip I found somewhere is that the rear axle is just the right
size to use as a puller to get the rotor out. In my case, I put a cheap
electric torque wrench on it. Seem to remember that we put the bike in
gear to torque it back on it, not sure if that's a risky thing to do.

 

I fixed a drop in charging by replacing a set of brushes
even though the old ones had only about 50% wear.
Not sure if the brush was beginning to hang up or if
it was just the increased pressure but at $9 for new
ones, it was well worth swapping them as an experiment.

Even if they're not obviously bad, replace the brushes if
they're near 50% gone. Might not hurt to swap in a new
regulator as well. You could check this by jumpering
out the old one to constantly energize the rotor. On mine,
there's a standard 3 wire automotive regulator and it's
a simple jumper across two connections.

There should be a known resistance value for both
the rotor and stator that you could check. Resistance
values should be available on an owners list.

You could also check the A.C. on each leg from the
alternator and on the rectified output of the diode board.
A difference in voltage between the A.C. legs would
point up a problem. A significant A.C. current after
the diode board would indicate a bad diode.

 

Sounds like an energized rotor energized by a regulator,
probably a three wire stator, and diodes to rectify the
A.C. output. What other info did you have in mind ?

 

I am an old Honda mechanic and my best guess is that the rotor is going
bad. A quick check is to hold a long (8 inch or more) feeler guage so
that the lower end of it is about 1 inch away from the alternator cover.
Turn on the key and the feeler guage should move over to the cover. My
guess is that it will only move a little. Check the resistance of the
rotor. If it is about 4 to 5 ohms it is OK, if it is below 2 it is bad,
I have probably changed 50 of these "back in the day". If you are gong
to buy a used one be sure to check it out.
Bruce

 

Some alternators have a brushless energised rotor - which would obviously
rule out worn brushes.

 

The OP had mentioned brush length. Which bikes use
brushless energised alternators ?

Agreed that it might have been good to put all that up
front instead of burying snippets all over the posting.

 

Things were getting a bit hazy by the sixth paragraph.

Which bikes use

2 that I can think of off hand are the Honda CB400F and the Yamaha XS250,
the alternator had a rotor with interleaved pole pieces but no coil, one set
of 3 poles went to the center spindle the other 3 went to the outer ring
thus making a co-axial magnetic coupling to a static field coil. It was
popular for a while but most manufacturers seem to have reverted to putting
the field coil in the rotor and using brushes again.

 

Thanks, Bruce.

I have a parts bike, and a second bike, both with the same engine, and a
rotor puller.
(I bought the rotor puller when I discovered a slide hammer was no good for
removing these rotors - fortunately, Mr. JB Weld, Mr. Soldering Gun, and I
were able to keep the rotor usable, this is -not- the rotor in the bike with
the
problem, BTW)

The rotor resistance when I first measured it (measuring at the connector
which
included the resistance of the brushes) was more like 20 ohms. I then
turned over
the bike and the resistance fell to something like 10 ohms. I will measure
it again
with and without brush resistance, as well as measure the resistance of the
other
rotors in the other bikes that I have for comparison.

Brand new, the Honda dealership wants $452.20 and it is a special order
part.
(yes, that's four hundred fifty two)

There's a guy advertising on Ebay that rebuilds these for about $100. I
probably
can also get one rebuilt at a local automotive electrical shop (we have a
number
of them in town). I doubt I will find a used one, every parts bike I've
seen over
the years in the various bike wrecking yards has been missing it's rotor - I
had
always thought it was because of people dropping them and destroying the
rotor,
side cover and such, but now I know better.

Ted

 

They aren't. I just replaced them last year, and the old ones
were indeed worn out. But I will check the springs, perhaps
the ones I put in had weaker springs. I did save the old
brushes just for use in identifying new ones.

Already tried.

There is. But, DC resistance values are usually bullshit in this
application. Your dealing with an AC system.

A perfect rotor with superconducting wire would have
a static resistance of zero. A few dozen loops or regular wire shorted
against each other inside the rotor won't even move the needle - but the
loss
in magnetic field is considerable.

When the rotor is energized and spinning around, there's
a very large resistance/reactance that it puts against the regulator. It
it really was only 4 ohms when spinning around, the current
through it would burn it up.

Generator output is a function of RPM and load. A properly
running generator at such and such RPM produces a fixed
amount of power. Power being both current and voltage.

If Honda had bothered to do their job with the FSM they would have
listed the generator AC output in volts and amps at several given RPM
speeds and a given field coil voltage (ie: rotor voltage). Very easy to
measure (any clamp-on ammeter can do it, such as the $10 Harbor
Freight special) and very easy to calculate the correct output if the
bike's owner has done something like, for example, loaded on a bunch
of extra power-sucking crap. And if the generator output is not up
to snuff, that eliminates all the downstream nonsense.

Instead, the FSM has a couple pages devoted to this Rube Goldberg
arraingement whereby you can use 2 batteries and test the regulator
for proper operation. Well the regulator is a solid state device, the
stator and rotor are part of a mechanical moving-parts device, which
is more likely to fail? Duh!

That is an excellent suggestion and I did indeed do that. The
voltages are the same. Incidentally, that is a good check for
the -stator-.

Unfortunately the only way you can see this is by an
oscilloscope and that's the one electronic testing tool
that I -don't- have (although I really wish I did)
A multimeter won't show it, unless so many diodes
are shorted that it isn't rectifying anything. (in which
case you probably would fry the regulator)

The biggest problem is Honda's idiocy in not providing
output voltages for the system when it is running in the
factory service manual. They do provide CURRENT
measurements at such and such RPM but they don't
provide the resistance load the bike presents, so if the
current is off you just know there's a problem, but you
don't know if the problem is because the bike's resistance
load is much less than normal because of some other
problem, or because the alternator output is down.

I put it down to a FSM written by people used to permanent
magnet bike magnetos.

Ted

 

Fer Gawds sake !!! There's a resistance value both for the
rotor and for resistance between phases in the stator.
For my bike, these are approximately 3.4-3.7 ohms and
6.2-7.5 ohms respectively. This really ain't rocket science.
It's a part spec. Parts outside that spec are suspect.

I've been running a bike with a very similar charging
system for about 85,000 miles and have worked through
rotor, stator, brush, regulator, short and ground problems.
Please spare me the lecture on alternator theory. My bike
runs fine and yours has a problem.

One other question: How carefully have you checked
your ground and +12V paths back to the battery ? Crappy
grounds or other major path problems can manifest
themselves in very strange ways. Both my diode
board and front casing have an extra wire back to
the main engine casing where the battery grounds.
I've been bitten by bad grounds before.

Best of luck.

 

I have put in re wound rotors with no problems. You might see if you can
get one from a dealer that can order them from K&L Supply in California.
I know that they used to have re wound ones. Good luck.
Bruce

 

Yes, I am aware of that. As I said in the original post,

"...Both rotor and stator pass the resistance checks in the
FSM..."

FSM = Factory Service Manual.

The Honda factory service manual gives the DC resistance.

I should have really clarified what I was meaning, though,
when I said that these checks are usually bullshit.

The reason they are misleading, is that IF they
pass, that is NO guarentee that the part is good. Windings
fail in weird ways. I've seen transformers where they
overheated and in one part of the winding the coils
melted off their insulation and shorted - thus dropping
the resistance - and in other parts they merely melted
and didn't short, but instead burned through, yet enough
wire-to-wire contact remained that there was some
conductivity left - thus raising the resistance. As a result
the static resistance was similar as before, it is just
that the transformer doesen't work. An alternator is
the same deal, it's just windings.

That is what I meant when I said the checks are usually bullshit.
I will repeat - static DC resistance checks on stuff like this
can prove the part is bad - BUT they CANNOT prove that
the part is good. Only a correctly running system can prove
that all parts in it are good.

Well then please spare me the lecture on testing something I
already said I tested and said passed, OK?

Yes, and it still has a problem when ALL the USUAL checks
have been done and it's passed them.

HOWEVER, I -think- that I may have found the issue,
and your tip on the brushes plus Bruce's comment on
the rotor PLUS reflecting on the theory of alternators
led me to what I think the problem was.

During my reflecting of alternator theory, it occurred to me
that something didn't add up. I was measuring 12 volts at
the rotor, yet I was getting low output
from the stator. Since the rotor is low DC resistance,
it should have been pulling a large current and thus creating
a voltage drop across the regulator and I should not see 12 volts.

If the rotor had shorted windings, it would draw even more
current, and I should definitely not see 12 volts at the rotor
terminals. (although, with a short it would also not create a
strong mag field which is why Bruce suggested the feeler guage
test)

In other words, 12 volts measured at the rotor connection
means the rotor is drawing very little current. Low current
means low mag field, means low generator output. And
the only thing in between the rotor terminals and the rotor
itself were - the brushes. And when you said you fixed a
setup by replacing the brushes, I decided to take a closer
look at them.

Well, the brushes were fine. Testing from a probe tip on
the tip of the brush to the rotor terminal I got 1 ohm.
Testing from ring to ring on the rotor I got 5 ohms.

Bolting the assembly together and testing the rotor terminals
I got 40 ohms. Ah ha! Bumping the crank I got that resistance
to drop - but not to 6 ohms.

The problem wasn't the brushes. The problem was the
COPPER SLIP RING on the rotor itself.

As I said, the rotor met spec on DC resistance checks. BUT,
when I tested it, I tested it by pulling the rotor cover and touching
the tips of the meter probes to the slip rings on the rotor.
Meter tips that are sharp, as any good tips should be.

Meter tips that apparently had no problems stabbing down through the
layer of tarnish and copper oxidization on the slip rings,
and making good contact with the rotor rings - without me
realizing the implications.

The brushes, by contrast, don't have this. Instead they are
designed to ride on top of the slip rings without digging gouges
in them. Unfortunately, they also ride very nicely on the layer
of tarnish and copper oxidization on the rotor slip rings.
(or something else - for all I know the brushes themselves
were releasing carbon onto the rings)

15 minutes of work with the wife's copper polish that she
uses to polish the bottoms of the Revereware in the kitchen,
followed by a wipe with some mineral spirits to clean the
polish off, and now, at 1000 rpm, I am measuring 8 volts
at the rotor, 10 volts at each stator leg, and full battery
charging voltage at the battery terminals. And, I'm measuring
6 ohms at the rotor terminals, which includes the brush
resistance.

My other 2 bikes (the other rider and the parts bike) obviously
need the same treatment since their rotor DC resistance is double and
triple the 6 ohms (but not as high as 40 ohms)

I guess what do I expect with a 28 year old bike.

It's easy to see how this is one of those problems that
creeps up. In operation as the copper slip rings get more and
more oxidization and tarnish over the years, the voltage
regulator compensates until it finally runs out of headroom
and then generator output starts to drop off.

As have I, as well as chafed wires, etc.

Ted

 

Its certainly true that a single shorted turn would make no visible
difference to the DC resistance - but a big difference to the alternating
current output of the winding.

For testing motor armatures, there is a test jig known as a "growler", its
basically just a shaped laminated core with a mains coil wound on it, the
poles of the armature are offered up to the poles of the growler which
buzzes in a different way when its in contact with a pole that has shorted
turns. Not much use with an alternator stator with inside pole pieces but
the basic concept might inspire something.

If you can get hold of a variac (rotary variable auto-transformer to give
its full technical name) feed this to a normal step down transformer of at
least 12V secondary, and an AC ammeter. With this you can set a convenient
current to measure through one arm of the 3-phase winding and compare any
one with the other two. This trick is no good for the field winding though,
the excitation voltage is DC, so it doesn't have a laminated core - the
solid core behaves like a shorted turn to AC.

 

If anyone is interested, the followup is that dirty slip rings were
the culprit. The battery has stayed properly charged now ever
since polishing them up.

Ted

 

Was there much trace of oil deposits in the alternator housing?

You might want to think about the crank case seal and possibly the main
bearing.

 

Ted Mittelstaedt wrote:
....

That's happy news. Just curious, what did you 'polish' them with?

 

Yeah. Resistance problems are pretty sneaky
and usually escape notice until they really hit
you hard with strange and confusing symptoms.

Last one I had was a main ground that started to
work its way loose. When you put on the turn signal
the engine would cough.