Index Home About Blog
From: John De Armond
Subject: Re: Diagnosing a generator problem - HELP!
Date: Sun, 09 Dec 2007 21:05:47 -0500
Message-ID: <>

On Sun, 9 Dec 2007 17:02:07 -0800 (PST), Honkey Lips <> wrote:

>Does anyone know where to start wit hte generator? It's got a quality
>italian alternator on it and wouldn't have been drawing anything like
>a full load. I did try a spare capacitor I had but that didn't make
>any difference. (I don't know however if the capacitor I had was ok)

By "Italian alternator", are you referring to the ones that Harbor Freight and
Northern Tool sell?  The ones with the extruded aluminum frame, no slip-rings and
with a capacitor hooked to an aux winding?  The rest of this post assumes that you

This type of generator is known as a "harmonically excited" generator.  The capacitor
resonates the associated winding at the third harmonic of 60hz.  This coil then
excites the rotor field via transformer action.  The field has a diode across the
coil that half-wave rectifies this current.  Clever design that is fairly new - the
patent is less than 20 years old.

The value of the capacitance controls the voltage regulation point.  More capacitance
== more voltage.  Ergo, the capacitance value matters.

Assuming the blue smoke hasn't leaked out of the windings, about the only things that
can go wrong are the capacitor and the diode on the rotor.

>Q. Anyone know how I can dianose the fault. (The repairers say they
>can't fix until after xmas) on the generator?

The first thing I'd look at is the capacitor.  The replacement must be exact in the
capacity (microfarads) and at least the same voltage rating.  Higher voltage is OK.

How you test the diode depends on what equipment you have.  Do you have a source of
60 hz magnetic field.  A tape degausser, for example?  If you do then hold it next to
a pole piece on the rotor and look at the voltage across the diode.  If you have a
scope, you'll see a half-wave rectified waveform.  If you have only a voltmeter then
you will see some value of DC voltage.  If you see no voltage, try the AC scale.  If
you see AC voltage then the diode is probably open.  No voltage either way and the
diode is probably shorted.

>Q. Does anyone know how I can test a capacitor so see whether or not
>it's blown?

A gross test is to first test it with an ohmmeter for short.  If the cap isn't
shorted, you should see a momentary upscale reading on an analog meter or a momentary
indication of resistance on the megohm scale of a DVM before both meters return to

Then hook it in series with about a 40 watt lightbulb to 120vac.  The bulb should
light at least a little.  You could compute the reactance of the capacitor using
1/2(pi)F*C where pi is 3.1417, F is the frequency (60 hz) and C is the capacitance in
farads.  If you know the resistance of the bulb (120 volts/40 watts = 3 ohms) then
you can quadratically add the values sqrt(R^2+Z^) to compute the composite impedance.
From the 120vac you can compute line voltage and with an AC ammeter, check the
current.  Use ohm's law to compute the resistance necessary to pass that much
current.  This resistance should equal the composite impedance you just calculated.
If they differ by much then the capacitor value is something other than its marking.

A capacitor checker, of course, makes the task trivially easy.  Many DVMs now have
the ability to measure capacity directly.

>Q. Do you think setting the tolerance that wide on my generator would
>have caused the fault?

No.  The poor voltage regulation was an indication of an unhappy generator.  If the
replacement capacitor contained a built-in bleeder resistor (most motor run
capacitors do anymore) then that may be at the root of the initial problem, depending
on the value of the bleeder.  The bleeder would lower the Q of the resonant circuit
and that might not properly saturate the rotor.  An unsaturated rotor would exhibit
poor voltage regulation.  Try to find a replacement cap that does not contain a
bleeder.  They've available, you just have to look and ask.

If you want more info on how this type of generator works, you might want to look up
the patent.  I don't recall the number but the title contains the words "harmonic
voltage regulation".  Hit one of the patent databases and search around.

Oh, one last thing.  Make sure your speed is spot-on.  No more than 61 hz unloaded.
If the speed is off-spec very much then the harmonic circuit operates off-resonance
and the field isn't properly excited.  Set the speed with a tach or a frequency meter
- either method is equally effective.


From: John De Armond
Subject: Re: Diagnosing a generator problem - HELP!
Date: Mon, 10 Dec 2007 11:07:17 -0500
Message-ID: <>

On Sun, 9 Dec 2007 20:33:20 -0800 (PST), Honkey Lips <> wrote:

>But just getting to the capacitor. People have been suggesting to
>stick a screwdriver across the contacts before measuring with a DVM
>(with an insulated screwdriver) does this seem ok??????

The capacitor is unlikely to be charged since the resonant winding is a DC short
across the cap, but if it is, shorting it with a screwdriver will burn pits in the
driver and, with internally fused caps, sometimes burn out the fuse, rendering the
cap junk.  I much prefer to keep a lightbulb screwed into a pigtail socket handy on
the bench for such tasks.  A 100 watt lightbulb will discharge any reasonable sized
cap in just an instant.


From: John De Armond
Subject: Re: Diagnosing a generator problem - HELP!
Date: Mon, 10 Dec 2007 15:32:43 -0500
Message-ID: <>

On Mon, 10 Dec 2007 10:39:01 -0800, "Ulysses" <> wrote:

>Is it a Mecc Alte Spa?  I have two of those and they are very dependable.  I
>did, however, exactly short the output while installing one and it fried one
>of the diodes.  The result sounds like the problem you are having.  I could
>not find the exact diode so I replaced it with the next higher number.  It's
>been fine ever since.  If you read the resistance of the rotor with an
>ohmmeter it should say zero if it's shorted but if it's open you really need
>to unsolder it to test it.  You can compare the resistance of the two rotor
>coils and see if they are the same.  This *might* give you a clue.

Bingo! That's the name I was trying to think of.

>I'm going to add Neon John's response to my archives.  How can anyone know
>so much about everything?

<tongue in cheek> After being on the net for a couple of decades, I marvel that so
many people can navigate life knowing so little... </tongue in cheek>  Of course, for
me, a few other things seem quite difficult.  Things like wardrobe selection, the
concept of owning more than one pair of shoes and remembering to sleep....  I guess
it all balances out.

Seriously, my mom likes to tell people that I was the only kid she knew about who got
more enjoyment out of taking something apart than I did using it.  She nailed me....

The OP got lucky and hit one of my interests.  A few years ago I built this 10kW

(bottom of the page)  It's the only generator of that output capacity I've seen that
can be moved and loaded into a truck single-handedly.  I used one of those MECC
alternator heads.  It is relatively light weight and I scored it for about $200.

Of course I had to take it apart before using it.  And I had to look at it with a
'scope after I got it lashed to the engine.  What I saw - no apparent means of field
excitation and a high frequency signal across the cap - made me get on the net and
start looking.  I didn't have much of any idea what to look for but after combing the
patent database with a few terms related to brushless fields, I hit on the patent
that describes this kind of generator exactly.

Ahhh, I could sleep well again.  It REALLY bugs me to have something in my presence
that I don't understand completely.  Alas, some aspects of computing are leaving me
behind.  Frustrating to have some electronic lump in front of me and not completely
understand it.

BTW, you can trim the voltage upward by adding a little capacitance.  Say, for
compensating for a very long cord.  I installed a switch on mine that parallels
another cap across the existing one.  5uF I think.  I selected it to give me a 10
volt boost.  That was very handy for my concession stand operating at the other end
of a hundred feet of 12/3 SO cord, overloading the generator a little.


From: John De Armond
Subject: Re: How to flash an AC generator
Date: Sun, 30 Dec 2007 20:24:58 -0500
Message-ID: <>

On Sun, 30 Dec 2007 22:35:09 GMT, Gordon Richmond <> wrote:

>Check your diodes with the ohmmeter. They should show a fairly low value with the leads
>connected one way, and a very high value (ohms) with the leads reversed. Very high (both
>ways) means an open diode; very low (both ways) mean a shorted one.

The diode is connected directly across the field winding.  It'll measure less than an
ohm either way.  Normally the diode is soldered and cemented down securely because it
rotates at 3600 rpm.  Practically impossible to remove without destroying the diode.

My method of checking the diode with the rotor out is to hold a tape degausser over a
pole piece, setting up a transformer, and measure the DC voltage across the diode.
Pretty much any DC means that the diode is probably OK.

If the rotor is in the generator then one can turn it until the rotor poles are
centered in the stator poles and then apply some AC to the output leads.  120vac
through a 100 watt lightbulb works.  Measure the DC across the diodes as before.

You'll need a high quality meter like a Fluke that isn't affected by AC on the DC

Most of the time, if changing the cap doesn't bring the voltage back up, I just snip
out the diodes and replace 'em.  Easier than doing a lot of testing.  Not much else
to go wrong other than maybe a shorted or open main winding and I exclude that before
I crack the case.


From: John De Armond
Subject: Re: How to flash an AC generator
Date: Sun, 30 Dec 2007 20:18:59 -0500
Message-ID: <>

On Sun, 30 Dec 2007 22:19:06 GMT, "Classic-Car-World Ltd"
<> wrote:

>Thanks John, I suspect the capacitor at the moment. I've removed this and
>checked it with a DVM I get a momentary short circuit reading and then it
>returns to about 56 Ohms. I would have thought it should have returned to a
>value much greater than this? I've ordered a replacement capacitor of the
>same value and voltage to see if this makes any difference.

That cap sounds maybe OK, if you meant 56k instead of 56 ohms.  Motor run caps like
that commonly have an internal bleeder resistor.  You should see the ohmmeter kick
toward zero ohms and then drift back to the value of the bleeder.  56kohms is about
right for a bleeder.

Can you run the generator?  If you can, measure the voltage across the cap.  If the
cap is bad or leaky, the voltage will be below about 150 volts, usually way below. If
the cap is good then the voltage is typically 250 volts or more.  If you can measure
frequency, the frequency will be the 3rd harmonic of whatever speed your generator is
running at.  If the cap is bad the frequency will be primarily the fundamental, 50 or
60 hz, depending on where you live.

The generator may not build any voltage at all if the cap is bad.  If it doesn't then
hook a 12 volt car battery through a 120 volt, 100 watt lightbulb to the generator's
output.  That'll supply enough field to get it to come up and if it happens to come
up to full voltage, the 100 watt bulb will protect the battery.

BTW, that's a standard motor run cap so you can pick up one at any appliance parts
store, HVAC supplier or electric motor repair shop.  If they see you coming, it'll be
$30-40.  If they give you the normal "wholesale" price, figure $5 to 10.

You can substitute a plastic cap, polypropylene or whatever, of the same voltage and
capacity.  Motor caps are no longer oil and paper.  Inside the can IS a plastic cap.
They just put 'em in the cans so as not to confuse the HVAC parts changers.  The can
is mostly hollow or oil filled.


From: John De Armond
Subject: Re: How to flash an AC generator
Date: Mon, 31 Dec 2007 11:58:48 -0500
Message-ID: <>

I can't get your schematic to load any farther than the bottom of the rotor diagram
so if there are any notes there I didn't see 'em.  Tried Opera and Firefox.

That's odd behavior.  Rotor diodes are a good suspect but I might also suspect at
this point a shorted turn in the stator.  I'd do a couple of things.  First, increase
the current input.  When I recommended a 100 watt light bulb, I was thinking around 1
amp for our normal 120vac.  If you had a 500 watt heater that you could put in
series, so much the better.

My other thought is, can you remove the stator from the rotor and test it separately?

In the motor shop I'd now move to the Baker Surge Tester

Unfortunately they've password protected their user manuals (!) so I can't send you
there to see how this test works.  In a nutshell, the instrument "rings" the winding
with a high voltage pulse and analyzes its decay behavior.  An OK winding will ring
for 3 or 4 cycles while a compromised/shorted winding will either not ring at all or
will damp down in one cycle.

If you have an O'scope then you can somewhat simulate this test by passing DC through
the winding and snapping open the circuit with a cap-shunted switch (same as old
points-type car ignition) while monitoring the winding with the scope.

We have another gadget called a core loss tester.  This involves passing a couple of
loops of heavy cable through the stator and passing high current AC through it.  An
AC buzz-box welder will do.  Then we scan each stator bar looking for differences.
The instrument has a probe coil connected to a meter and headphones.  The probe is
held over the stator and gradually moved around.  Any variation in intensity
registers.  A shorted turn will essentially eliminate the magnetic field.  That area
of the stator will also get slightly warm from the circulating current.

I'm still not happy with that capacitor reading.  56 ohms is far too low for a
bleeder.  I also don't think the meter would kick much initially with that little
resistance across it.  Are you sure your meter is OK?

I normally don't suspect test instruments because they are so reliable but every
reading you reported below is highly abnormal.  Do you have another meter you can
try?  I've never seen a shorted rotor diode cause the excitation winding voltage to
be that low.

At this point, if your meter is OK then I'd probably pull the rotor and stator apart
and test each separately.


On Mon, 31 Dec 2007 13:47:52 GMT, "Classic-Car-World Ltd"
<> wrote:

>Hi Guys, I've done some testing this morning on the generator with a 240V
>60W lamp in series with the stator windings to try and induce an EMF in the
>I then took the following measurements:
>Exciter windings with the Capacitor removed (Test points A - B) = 2.46V AC
>Exciter windings with the Capacitor installed (Test points A - B) = 0.93V AC
>Rotor Windings with the Capacitor removed = 153mV DC (Test points A1 - B1)
>and 174mV DC (Test points A2 - B2)
>Rotor windings with the Capacitor installed = 1.42mV DC (Test points A1 -
>B1) (1.94V AC) and 1.42mV DC (Test points A2 - B2) (up to 10V AC changing).
>I suspect one of the Rotor diodes may be faulty as I get the same changing
>readings on the same winding if I rotate the rotor 180deg.
>I've also rechecked the capacitor with the DVM and it is only reading 56
>ohms after the initial charge kick.
>The following is a link to the test set up and how I believe the windings
>relate to each other.
>I'm not sure about the orientation of the diodes as these will have to be
>disconnected before I can check them.
>As soon as I have disconnected and checked the diodes I will update the

From: John De Armond
Newsgroups: rec.outdoors.rv-travel
Subject: Re: I want Your take on the best Portable Generator
Date: Sun, 21 Sep 2008 12:32:18 -0400
Message-ID: <>

On Sun, 21 Sep 2008 02:01:23 -0700 (PDT), "Hustlin' Hank" <>

>Alan and All,
>     Let me be the first person on this NG to apologize and admit I
>may have been a LITTLE in error. :-)

Thanks.  Takes a big man to do that.

>    I called TSC but due to the wind storms we have had they didn't
>have any Generators in stock. I didn't ask them what kind they carried
>since they were out of stock. I didn't try Lowes.....yet.
>   Since most stores are without ANY generators due to this wind
>storm, I am waiting for a while before I buy. I will not shy away from
>the Champion brand.
>    Thanks to all for your input.

You're welcome.  I spent some time on the net last night price shopping.
Amazing how cheap those sell for.  Most of the e-stores indicated that they
had merchandise in stock.

On the champion page, notice that they helpfully supply both the manual
(fairly poor) and the schematic.  I downloaded all the schematics last night
(another "issue"? :-) and looked them over.  All but the two cheapest "light
duty" ones have actual electronic voltage regulators and slip rings on the
rotor for field excitation.  This is GOOD.

"Brushless" is NOT good in a small generator.  The brushless harmonically
regulated generators work but the waveform output is lousy plus sometimes the
generator can come up on the wrong harmonic and output excessive voltage.  My
homemade 10kW diesel generator will sometimes do that if I start it under
load.  That's why it has a voltmeter on it and why I never start under load.

The one thing that the harmonically regulated generators are good for that
electronic regulated ones sometimes aren't is driving a welder.  My homemade
generator does that wonderfully.  I smoked the regulator board on my Yamaha
with my MIG welder.  Apparently it tried too hard to follow the wildly varying
load that the welder presents and overheated the output transistor in the
process.  Fortunately, not potted so I could fix it.

>To John: I too, will spend all day fixing a $10 appliance. Maybe you
>and I have some "issues". :-)

Of course.  Many many issues.  Ain't they fun? :-)  Like one of my .sig lines
says, "I'm not suffering from mental illness.  I'm enjoying every second."


Index Home About Blog