Old Generator power low

[Soloboss]

Even if the rings really weren't glazed, they certainly weren't bright copper. And the rings are really thick copper so a good sanding didn't hurt anything. Certainly brush contact is not the issue now. I did go through all of the spade connections and look at each under a magnifying glass. I now see bright metal at the points where the blade contacts the receptacle at each connection. All of the wire to terminal connections are soldered. I visually inspected each but did not resolder them. I'll test run the generator tomorrow. If it works, it's likely the brushes/rings had a glaze.
I didn't do the solder thing because of the possibility of creating a cold solder joint in a connection that didn't need rework.

I can't run it without the bell end bearing because the generator does not have an idle circuit in the carb. Cleaning the rings was simple enough, using many small strips of sandpaper so as not to spread whatever contaminant might be on the rings. If I still had the cylinder head off, I'd connect an electric drill to the engine and spin it that way. I do need to ensure the rings are very smooth so I don't grind the brushes to dust in 5 minutes run time. I still believe that this little unit will be a nice power source when it's done. For the new digital equipment it may not be the answer, but for good old fashioned portable power, it's got the plastic housing imports beat.

 

[Soloboss]

Just a thought - is this mechanical generator thread in the right area of this forum. I used to be very active in an RC Helicopter forum and the guys were pretty sensitive about who was posting where. If I'm in the wrong spot, I'll change (and hope I don't get lost!).

 

[john1]

Hi Mark,

Your thinking that a magnetic charge gets induced within the coil,
is closer than you think, and not far off base.
The pole pieces opon which the exciter coils are mounted,
will hold a small amount of residue magnetism.
When the generator is run up to speed, this gives enough feed out
to start the generators electrics.

Sometimes this residue fades too low to kick off the generator,
that is the reason for "flashing" a couple of times.

This is best done with the diodes and capacitor disconnected,
because the sparks involved are a result of very high voltages
and surges in places which would not normally see such high voltages
or surges in the usual conditions.
They are not rated to cope with that.
It is only the field winding that may need this treatment, which on this
type appears to be the rotating part (on some types it is stationary)

*********************

As to the appearance of the mating surface of the commutator rings,
they will assume a darker colour than freshly cleaned copper after a
little use, this is normal and does not affect the contact with the brushes.
However cleaning them is a good idea and will do no harm especially if
the unit has been idle for a time.

Best of luck with it,
John

PS. I think your postings are Ok here.

 

[Soloboss]

If I read your answer correctly, allowing the generator to produce power for some period of time may help to restore the coil's magnetic properties? I'd be happy to allow it to run for a bit and operate a heater in the house. Just Tuesday it was running perfectly and offering a full 15 amps with very little effort. Then it all went away. I'm hoping that a bit of grease on the rings caused the generator to lose its ability to produce power and that my cleaning restored contact. Next time the generator actually generates some power I'll let it power something for 30 minutes or so. Perhaps it just needs some exercise.
When I flashed the capacitor, I touched the cap with 12 volts for perhaps a half second. The generator was not running. Following the flash, I did start it within one minute and it operated perfectly except for the gasoline leak and the rich carb setting. This is interesting, although I'm not certain that I'm learning anything - at least not until it all works!
Thanks for the reply
Mark

 

[john1]

Hi Mark,

Sorry to be a little late in replying.
I was not expecting an answer tonight, its three in the morning here (UK).

I realise that electrics is not your field, so i will try to be direct and
clear, and i will try not to be misleading in anything i may write.
As there have been over a hundred and fifty views on this thread, i am sure
that somebody will soon speak up if anything looks confusing.
*************************

Does running the generator restore any of the residue magnetism?
.
There are differing opinions on this.
Some say it does.
I dont think it does,
but i do think that striking or hitting metal while its being magnetised
can cause it to retain some magnetism.
So if the unit were knocked or dropped or struck in some way while the field
windings were energised, then maybe that would account for some people who
say that running it helped it retain some magnetism.

I have never known running a generator to help with residue magnetism.
The only (sensible) way i know of to impart magnetism to the field poles
is a couple of short sharp pulses of high current through the field windings.

I dont think running it will help at all with the residue magnetism.
(this is not a suggestion to knock or strike the unit in any way)

The Flashing.
-------------
This is not something that will need to be done often.
Many generators will retain their residue magnetism for years.
The coils are copper and do not retain any magnetism, it is the ferrous metalwork
that retains the residue magnetism, that is - the pole pieces of the field assy.

The flash or spark is due to the sudden interruption of the current which is
temporarily applied to the field winding/s, usually by a person touching it's
connection with the tip of a length of wire.
This is a function of the inductance of the winding, and is not associated with
a capacitor in any way, other than maybe their shared connection.
Having the capacitor involved by remaining connected can only reduce the desired
effect of flashing the field winding.

I was shown to 'flick' the temporary connection off of the terminal which resulted
in a distinct flash as the wire 'sprung' away from the terminal.
This seems dificult to describe, but very easy to do.
I would hold the wire on the terminal or connector near it's edge, then push it
past the edge in a flicking motion, so that it would break contact while moving
quickly. This would usually give a disinct flash.
It only needs to be done a couple of times.

The residue magnetism is needed only for the starting up of the generators
electrics, once its running the field would be fed in some other way from the
unit's output.

If the unit started to fail when it was running, then this could be as you say,
the rings getting a bit gummy with oil.

It would be best not to have the capacitor or the diodes conected to the field
while flashing as they could be damaged quite easily by it.
Those sort of sparks and voltages wil not do the capacitor or the diodes any
good at all.

Best of luck with the petrol leak and the mixture settings,
John

 

[RCinFLA]

Check the frequency on output. Higher then normal voltage on these older inductive pickup field winding generator usually means the engine is running higher then 3600 rpm. You will probably be higher then 60 Hz.

 

[colin55]

Get some LEDs and 4k7 resistors and put them on the exciter coils and see if the illuminate and if they show any faulty connection.

 

[Soloboss]

I'm building a great database in the generator theory area! Thanks everyone for the replies. Here's the latest.

The Ag-Tronic again runs perfectly. In my excitement, I put a load on it soon after it was warm enough to run without choke. I had not set the engine RPM at this point, but I did see 8 amps output. UREKA!!
I restored the engine to 3600 per my digital tachometer. At proper RPM, I had 12 amps at 117vac output with 62vdc at the brushes. Cool. Feeling pretty smug, I sat back and just watched it run – then watched the voltage fall off to nil at about 5 minutes into the test. So what happened?
Investigation leads me to a theory. In a word, BRUSH WEAR. The newly cleaned rings show polished brush contact areas with one distinct feature. The outer ring is about 12” in circumference. There is a gap in the polished area that is about 2” long. It appears that the ring surface is not PERFECTLY perpendicular to the centerline of the rotor. The spring should keep the brush in contact with the ring despite the very minor imperfection. UNLESS – the brush is worn in the mounting flange.
See the attached examples.
Example A shows the design intent. The brush should be a sliding fit in the holder. The brush would move easily on it’s axial centerline and would not bind.
See Example B showing worn brushes that can cock and stick in the holder. Certainly the example is overstated, but noting my brush wear, I believe that it’s valid in theory. See the photos and note the slight cuts indicating the end of the location of the brush holder. This indicates that the brush is supported at the outer edges of the holder and not along the length of the inner sides of the holder. Also note that the end of the brush is not worn square. The angle of the contact face indicates that the brush is cocked in the holder.

When the bell is mounted, the brushes are compressed. The rotation of the rings could certainly cock the brush just slightly. As the generator runs, the corner of the brush in contact with the ring would wear off, but the side thrust across the brush face would prevent the brush from extending by spring pressure. The contact face wears away just a few mills and contact is lost, befuddling the entire AC power generation cycle. When I shut the generator down and fiddle with it, the brush resets against the ring and the generation resumes for five more minutes. Each time I run the generator it starts strong and quits at about 5 minutes into the test. I think I'm shopping for new brushes.

Thoughts?

 

[john1]

Thoughts?

Before going much further along this path,
i think you should attach a couple of leads to the brush connectors,
and take them out to a voltmeter.

You can then watch the voltmeter next time you start the unit.
I think you may see the voltage fall away after five minutes or so.

The more information you can get at this point,
the more likely that you can identify the problem.

John

 

[Soloboss]

John, At this point I think I'm getting to the mechanical cause of the electrical problem. I've tried not to offer a persuasive arguement, just offering what I know to be true and what that would lead to. Your thought on the leads and voltmeter would be very informative. I found brushes at an online supplier. My brushes are 14.34mm long. New brushes are 22mm long. The shortened length will allow them to turn in the mount. I didn't really want to put another $20 into this rig. I purchased it for reconditioning and resale and I paid too much for it. If I can recover most of my money and deliver a solid performing unit I can write off the loss as experience. I must say that I have a MUCH better understanding of the workings of these little gadgets! And it's been fun.
Off to bed,
Mark

 

[RCinFLA]

After the brushes, I vote to replace cap.

 

[Soloboss]

I'll send you the specs and then wait by the mailbox! LOL
I agree. With all of the variables, the mechanical wear parts would be first and the cap would be next.

 

[Reloadron]

I think I have read all the post in the thread. I have had similar problems twice with generators. Once with a large 90 KW 480 Volt unit and once with a small 8 HP Briggs and Stratton driving a 4 KW Coleman generator. The units would run and display their rated voltage but as soon as a load was applied they would just fold over with voltage and current dropping to about nothing.

I had checked the diodes using an old (and trusty) Simpson 260 VOM reverse and forward biasing them and they looked good. However, looks were deceiving. The large unit used a full wave bridge of stud mount diodes and what was happening was one diode in the bridge was actually breaking down when it ran. Looked great with a static test using the VOM but... My problem in both cases was a diode that had breakdown. Replaced the diodes and things worked fine. I still have the little Coleman Power Mate and to this day it runs fine.

Ron

 

[Soloboss]

You're aware that I have just a diode pair, but that's not the issue. Your thoughts of thermal breakdown are well taken. In order, I figured that a 35 year old electrolytic cap that's been stored every way but correctly would be the first component to go. Then the diodes.
At this point, I'm into the brushes. My father in law suggested finding very thin plastic shim stock and shimming the old bushes so they run true and tight (not binding tight) in their holder bores. I offered tape. He laughed at my 'cheap way out'. But packing tape is 0.002 thick, adhesive backed polyethylene plastic sheet. Funny how the description changed the validity of the suggestion. Knife cut and applied tightly under a work magnifier the fix looks good. The brushes are now 0.003 undersize in width and 0.007 undersize in height. The load of the ring wiping motion is in the width direction and that's as tight as I can get it without binding.
This is a temporary fix. If my power output becomes stable and reliable, I'll replace the brushes as a set for $20.

 

[Soloboss]

Test result with the brushes tightened. It initiated and took a load perfectly for 3 minutes with 135vac and 75vdc at the rotr measured at the brushes. Then the amp meter started to fluctuate and output dropped to 3-4 amps and the DC voltage at the rotor dropped from 75vdc when the output was strong to 32 vdc.
I shut off the engine and fiddled with the brushes by reaching in through an access port. I basically just poked at the brush holder with my finger. Within one minute of shutting off the generator I restarted it and it picked up the strong output again, 135vac, 75vdc at the rotor and 8 amps. I've never seen it run for more than just 3 or 4 minutes with strong output. This time at 8 minutes of run time it was still strong when I shut it off. We agreed that it seemed to be running fine. Gotta test it just one more time - I restarted it and the vdc at the rotor was back to 35 volts with the resulting very low output again.

Just to be clear, when it drops to the Low Power situation it's not dead. There are 35vdc on the rotor. The AC is about 85 volts and the current meter reads about 3 amps. It has never been DEAD in any of my tests.
Perhaps corrosion at the diode threads? I'll re-seat the diodes. And I'll solder all of the connections. I'll pull the brushes and start looking locally for replacements.
So the story has changed again. When something interesting happens I'll post. Until then I'll keep trying to nail down a repeatable test situation to remove some of the variables.
Thanks for the ongoing support. I'll beat this thing! I just hope I don't have the cost of a new generator invested.

 

[Soloboss]

I think I have read all the post in the thread. I have had similar problems twice with generators. Once with a large 90 KW 480 Volt unit and once with a small 8 HP Briggs and Stratton driving a 4 KW Coleman generator. The units would run and display their rated voltage but as soon as a load was applied they would just fold over with voltage and current dropping to about nothing.

I had checked the diodes using an old (and trusty) Simpson 260 VOM reverse and forward biasing them and they looked good. However, looks were deceiving. The large unit used a full wave bridge of stud mount diodes and what was happening was one diode in the bridge was actually breaking down when it ran. Looked great with a static test using the VOM but... My problem in both cases was a diode that had breakdown. Replaced the diodes and things worked fine. I still have the little Coleman Power Mate and to this day it runs fine.

Ron

OK Ron, I'm getting lost in the details of this thing. The things that seem obvious make no difference to the test resutls. As I've noted this unit takes a load fine initially. After two minutes it drops from 130vac to 90vac. The DC on the field drops from 75vdc when all is well to 35vdc which yeilds 95vac. If I was losing a diode, I'd lose my field as yours did. Ground the field, no AC output. Open the diode, no AC output. I'm starting to think that I could replace the diodes, the brushes and the cap and I'd still have the same problem. I need a better test.

 

[john1]

Hi Mark,

I think your problem lies with the capacitor or the diodes. (or both)

After two minutes it drops from 130vac to 90vac.
The DC on the field drops from 75vdc when all is well to 35vdc which yeilds 95vac.

You could individually test the diodes and the capacitor with a testmeter,
most general faults will show up, but not always.
Some faults only show up in operation.
As Ron has mentioned, looks can be deceiving. His diodes showed alright with a test-meter
( a Simpson VOA is a testmeter )

The definitive test is of course that the components operate in service.
I would like to see that they work Ok in conditions a little more severe,
than they would meet in service.

It might be easier for you to replace them than to arrange for such testing.

Unless the testmeter shows up the problem.
Checking with the testmeter should be done with the items disconnected at least at one end.
If the testmeter shows up a problem, then changing all three components might be your best bet.

Best of luck with it, John

 

[tcmtech]

If one of the diodes is failing after it warms up slightly you wont loose all of your exciter power but only about half of it. Diodes can go "semi bad" Its rare but being yours are 35 years old its possible that one has a micro crack in the glass seal that has allowed just enough of something to get in and contaminate the semi conductive materials.

 

[Reloadron]

In my case neither generator started out good, they had the problem from start up. What I found in each case was a diode that looked good with a basic VOM test but was breaking down as was later seen on a curve tracer. The overall symptoms looked like those in the original post. My field voltage was bad and resulted in a bad output which made sense to me. Just like tcmtech called it.

Not saying that is for sure your problem but the options and possibilities are running out.

Ron

 

[Soloboss]

I can build a test situation and work quickly after startup. I'll monitor only the field voltage. My initial test will involve only the diodes. Both will be in-circuit at startup using test leads. I'll "open" each diode briefly and watch the filed voltage. Restoring each good diode should reinstate the full power. When the output drops due to it's typical failure, removing the bad diode from the circuit will have no effect. If I remove the working diode from the circuit and the other diode has failed, I'll lose the field entirely.

Correct me if I'm wrong. If one diode goes open, I'll have REDUCED field voltage. If one diode shorts to ground, I'll have NO field voltage. TEST ONE: Test one diode then the other diode in circuit and out of circuit.
IF there is no useful information from that test, I'll try TEST TWO: one diode, then the other diode will be shorted to ground. I expect that either the open test or the short test will reproduce the lower field voltage.
I'll look for suggestions before continuing with the test. I'm trying to make this work better, not blow up a good component or winding.
Thanks all and good night. To be continued . . .