What's new
What's new

Semi OT: Exciter circut capacitor on a generator? Help!

John in MA

Titanium
Joined
Apr 14, 2005
Location
MA, USA
Another Round Tuit project. Several years ago I bought an '80s vintage Yamaha EC2800 120V generator from guy. He said he had bought it from a strange old fellow who had the thing hardwired into some kind of weird house-powering charging system. As a result the frame and electrical control box had been stripped off.

It took me a long time to find the schematics, but once I did I was able to test it. Jurry rigged up some stuff and it seemed to function OK.

It's a brushless self-exciting generator. The exciter winding uses a capacitor in its little circut. The problem: I have no clue what the value of this cap is, and Yammie won't tell me. They want $50 for a replacement--fat chance. How does the capacitor's value effect the generator's output? I looked at a newer EF2600 an it was a 24mfd. A fellow I talked to said that he thought 26-28 would be good for mine. But I tested it with a 45 and it did OK. I just didn't test it under load.

Any EE types care to help me with this? I'd really like to get this thing into full operation because I could sure use a nice generator.

Schematic, although it's not really needed:

ec2800.jpg
 
My best guess is that the capacitor (or condensor as shown on the schematic) is there to eliminate AC ripple and dampen the voltage changes on the exciter circuit.

A self exciting brushless alternator requires a DC voltage on the exciter stator (the sub coil on your schematic), which induces a voltage in the exciter rotor, which is rectified by the diode bridge, and then feeds DC into the main rotor, which induces a volage in the main stator which is the power you can use at the outlet.

My best guess is that if the capacitor is too small, it will allow some AC ripple onto the exciter and the AC output will not be as clean as it could be. I would also suspect that a capacitor that is too big would likely dampen the voltage regulator action since the capacitor would take some extra time to charge or discharge as required. The scehmatic doesn't show the regulator details, but normally the regulator takes a sample of the output voltage (120V or 240V) and uses that to power the regulator as well as determine if the exciter needs to be driven harder as a load is applied.

I would try a 24-28 mfd and if possible, borrow an oscilloscope to see the AC waveform under varying loads. If it looks good (no spikes or weird shapes) you should be good.
 
What are the two cylinder shaped symbols on either side of the cap? It seems like the cap would be wired shunt instead of series to smooth out DC if that was it's purpose?
 
Thanks guys. The two cylinders are just connectors, as far as I know. From looking at similar generators in person there's no other wiring. It's hard to see into the generator housing itself but I don't think there's any more circutry in there.

No need to borrow an oscope, I have a couple.

Anyone else?
 
060731-1504 ERST USA

John in MA:

Here are some guesses if the circuit diagram is accurate and complete and no other components exist.

The capacitor is not a filter capacitor based on my above assumptions. Rather it is a resonating capacitor that in combination with its associated magnetic circuit works to regulate the output voltage. This means the the capacitor must be an AC capacitor. This circuit might work like a Sola constant voltage transformer.

I assume there are no slip rings for rotor excitation and that induced current in the rotor coils, allowed to flow primarily in only one direction by the diodes, provides the DC magnetic field required in the rotor.

Capacitance value will be important, and it may require a fairly high voltage rating. More or less than the optimum value will produce poor results.

.
 
I am going to agree with gar.....somewhat.

Since this appears NOT to have any separate exciter section, I am going to assume that the residual magnetic field is used at startup to "pump" the resonator system and it in turn provides an excitation by induction.

Otherwise it is a little unclear to me where the excitation "buildup" comes from.

There may be a regulation function also.

It appears somewhat like an AC induction generator, which often use capacitors as a source of excitation load. They tend to collapse under overloads, due loss of excitation, since excitation is in the same circuit as the load.

In this case, I suspect the discrete resonator system provides a somewhat independent excitation. It may be isolated similarly to a Sola as gar references... in which case it may indeed have the dual purpose of excitation and regulation.

In any case, his observations about the capacitor are right on, as I see it.

I'd like to see what PeterH has to say about it.... He may have an additional input.... But his message box is full!
 
The whole thing is somewhat of a mystery with various worthwhile comments. My own is that in looking at it with an oscilloscope you may very well see a more or less trapezoidal waveform with great big slot ripple notches. If so there is not much you can do about it. I'm only mentioning this because I had a generator which showed this sort of waveform.
Ken
 
My first guess might be to measure the
inductance of the "sub coil" and then pick
a value for C that resonates it at 60 cycles.

If however that winds up being a nonsensical
value for C, I'd have to do some headscratching.

Jim
 
Do you think it would be worth buying the OEM part? I'm rather reluctant to since I haven't even given it a test under load. Didn't want to until I knew what using the wrong cap would do to me. Looking at catalogs it seem motor run caps in this range are usually 25 or 30 mfd. What exactly would the difference be? Wrong frequency 120V output? Voltage change?

Residual magnetism is correct. When I got it, it hasn't been run in years and years. I had to connect it to a 120V outlet with a current limiting lightbulb before it started generating voltage. Worked fine after that.
 
Frequency should be set by RPM.

I suspect that wrong cap would give poor excitation, with lower than expected max power out, and probably poor regulation, voltage sag.

Probably wrong either direction would give similar results, so.........

I'd be tempted to resonate it at 60, probably using a small transformer and a series resistor across it as a source so as to check resonance. A few hundred ohms and 12V should give an indication.
 
I think that the residual magnetism is too low to get full voltage, so they increase the rotors magnetic field, by allowing to take current via the capacitor. While DC excitation is common, when the o/p volts are zero, there is no need for any excitation current! So the capacitor must be above some value or full o/p volts will not be achieved. As it gets bigger the o/p volts will tend to rise, but may be limited by the saturation of the iron in the rotar, but the rotar will be taking more current so getting hotter. So there is some maximum value set by over heating the rotar coil. If it is run in the saturation mode, the calculated capacitor might be 20Mfd, so allowing for a +_ 20% tolerance, 24 would be the commercially suitable value.
Frank
 
Chuckey,
Not sure what o/p means but I think you're onto it: are you saying the cap is just storing a charge to flash the field on startup? I'll buy that though I haven't seen it before.
 
o/p = output, most likely.

I don't think it can store a charge....

The cap is on the stator, and any charge would be drained thru the winding.....

My assumption is that, if it is resonant at 60 Hz, similarly to a voltage regulating transformer, the residual magnetism in the rotor will set up a 60 Hz weak pulsating field when the unit comes up to speed.

That will "pump" the resonator, inducing a current. The resonator will have a larger current in it, due to resonant energy storage, than you would expect just from the weak residual field.

That current must then induce a current in the rotating field windings, which pumps the resonator harder, inducing more in the rotating windings. That cycle continues until the resonator 'clips" by local saturation (assuming it is like a Sola as gar suggests).

The rotating field also induces voltage in the output windings. Since the induced field current is dependent on the resonator, the limiting of resonator energy also regulates the output.

I would expect the output windings are coupled magnetically to the resonator, which provides the feedback of "volts per turn" necessary to regulate the output voltage.

At full voltage, the "clipping" of the resonator limits the induced voltage in the field, and so the output voltage.

It sounds like a perpetual motion machine, until you realize that the input is the mechanical energy of turning the rotor. That rotates the field and pumps the resonator.

Because the resonator is somewhat decoupled from the output, it isn't so dependent on the output load as say a shunt DC generator. Those won't "build up" with too heavy a load on them because the load steals too much current and the field does not get enough current to have positive feedback and build up.

That all means there will be a "best value" for the cap. Any other value will give lesser performance.

I said the frequency is set by RPM, and that would be so.... but per my thoughts above, the wrong value cap would tend to make the whole system work best at a different frequency....

So too big would cause the best frequency to drop, and too small causes it to go up, by the square root of the cap value error.

Thus, too big a cap, and it could be a 50, or 55 Hz generator, with max output at a lower RPM and lower frequency than 60.
 
Well said J Tiers, I think the judges'll buy that. By your description it seems like for another dollar or two for a coil they could have made a really accurate resonating circuit but that would be less profit I suppose.
I should've studied harder in school.
 
There *is* a coil there. It's the stator
winding. Hence my suggestion to just measure
that inductance with an inexpensive L meter.

Also you don't want that ckt to have too high
a Q. Because the speed on the motor varies,
it should probably be less then 5 or so.

Jim
 
Gar is absolutely correct. That would be a "resonating" cap and is undoubtedly used for voltage regulation.

The value is going to be critical--I have no idea how to calculate, other than trial and error

It needs to be an AC cap, such as a run cap

I doubt that it has anything to do with excitation
 
That works, but so will setting it for resonance (largest voltage) using a small transformer and resistor as a source.

Or trying values and checking for full output at correct RPM/ frequency. An electric clock with a sweep second hand will quickly ( a couple minutes) show if frequency is far enough out to be a problem.

The "Q" will be already set, since it is strongly dependent on the cap value. The cap in turn can have only one value to resonate at the correct frequency with the existing coil.
 
1, Apologies for my first post, what I said is a load of b*****ks, should have looked at the circuit closer. Another thought, I have heard that cheap generators tend to produce a wave form that is a series of steps, rather then a sinewave. Could be a tuned circuit to notch out the third harmonic of 60 hz in an attempt to improve the waveform. According to my maths 24Mfd will tune with .3H at 60 hz, so it will take .3/9 = .033 H to tune to 180hz. To measure the inductance without a bridge, get a bell transformer and a known resistor,say 10k ohms, connect the reistor in series with the bell transformer secondary and the sub coil. Energise the transformer. Measure the voltage across the subcoil (= x ) and the resistor (=y). Providing y is much bigger the x, the impedance of the coil is x/y times 10k ohms. The inductance (at 60Hz) is the impedance / 2 x 3.14 x 60. If the voltage measured at x is too small for your meter, use a 100 ohm resistor.
Frank
 
What you have there is a "mase" generator. The capacitor is there to provide feedback/resonate the auxiliary windings and in turn induce a magnetic field in the rotor windings, and hence usable power in the stator windings.

The following comments are taken from "The Marine Electrical & Electronics Bible" by John C. Payne:
1) While rotating, the residual magnetism and permanent magnets induce a voltage into the auxiliary windings for excitation. This voltage is fed to the capacitor, which generates capacitive current in the circuit.
2) The capacitive current creates a magnetic field, which is rectified by the diode, supplying a DC current to the induction winding. This generates a rotating magnetic field for generation of output.


Varing the size of the capacitor may give you some measure of control over voltage regulation, especially loaded vs unloaded.

YMMV
Graham
 
I have an EC2800 from 1986-86 I was looking for information on how to field flash the unit, as it has laid around for so long it's lost it's magnetism, when I saw your artical. The capacitor in my unit was made by ShiZuki and has the following markings. SH-P
36 uF
250 VAC
RS-X
1986

I hope this will be of help to you.
If anyone has any info on feild flashing this unit it would be appreciated.
 








 
Back
Top