Resurrecting a dead generator

It does sound high, but alternators have a highly inductive stator where the source voltage is proportional to alternator speed/frequency and where the impedance is also dependent on speed/frequency. This means that at output current is near proportional to rotor field strength. Obviously where terminal voltage is a significant proportion of source voltage this argument breaks down.

I'm presuming your inverter is smart enough to limit voltage by shorting terminals when the rectified voltage is above a prescribed threshold, otherwise, as you say it would be instant destruction of the inverter and its caps.

My first suspicion would be a chinese manufacturer cutting corners when they were short of the right caps.

NT

Yes, I think my initial view of how this works was a bit simplistic. I've been reading up on synchronous reactance. I need to do some measurements with a load applied to characterise this at various speeds. It may be that the inverter doesn't need to draw much current to bring the voltage down to something reasonable. It may even be that the inverter's own power requirements are sufficient.

Sylvia.

I don't think even the Chinese would believe that running electros above their rated voltage was a good idea.

Sylvia

I'm a little confused, is the field generated by a permanent magnet alone or does it have a field winding in addition? I'm trying to picture how and alternator has both.

Some do have separate excitation generators.

It has magnets and a field winding. Bear in mind that this thing has no battery, so without the magnets it would have to rely on any residual magnetic field to get itself up and running, and that might not be sufficient, let alone reliable.

The field winding allows the generator to control the output voltage over a wide range of currents.

Sylvia.

Oh I do. It does work as well.

NT

Well, it would, if it were a field winding. I'm now rather doubting that, having found a couple of exploded diagrams for Kipor alternators that make no mention of a field winding, and show nowhere to put it.

The white wires are for a "sub-winding", and it appears that such a winding is sometimes used to power the inverter in inverter generators.

A different way of handling the varying load is just to change the mark-space ratio for the pulse-width output that's used to produce the sinewave output after filtering. But it does leave the problem of the potentially very high voltages that have to be dealt with. The main wires from the alternator are marked as being rated at 600V, which still doesn't seem enough.

Sylvia.

Sylvia-

This is your fist mention of a brand. It might help if you provided the model number as well.

Alternator Voltage is proportional to speed. If your inverter is not working, there may be nothing to stop the engine from running flat-out full speed. So the voltage you see may not be the same as if the inverter electronics were functioning.

Some alternators have a permanent magnet rotating field, but that is rare. Often a permanent magnet is used to maintain a residual field required to start generating.

Your alternator produces variable-frequency 3 phase AC. It must first be rectified to DC, since 60 Hz AC output voltage is generated from a DC-to-AC inverter.

I understand pulse-width modulation is how output voltage is controlled in inverter generators, but there are some designs I am not familiar with.

Fred

As Fred has suggested a model number would help.

Is there an "exciter" winding or otherwise mentioned in any diagram?

The problem relates to what happens in the worst case where the generator is running with a rated load, which is abruptly disconnected. To support the load, the generator will have been running at a high RPM. In response to the loss of the load, the control circuitry will close the throttle, but this, and the actual slowing of the engine, both take time. So there's an interval of perhaps a few seconds during which the engine is running fast, but there is no significant load on the alternator. The voltage from the alternator will inevitably rise. I had thought this situation was managed by changing the excitation current, but the absence of an excitation winding precludes this.

Using capacitors to limit the rise is out of the question - the required capacity (at a high voltage at that) would be prohibitively expensive, as well as occupy a huge amount of space.

I'm thinking now that a dummy load must be involved that can be switched in as required to keep the voltage down, and dissipate the resulting heat. Presumably there would be some protection circuitry to shut the generator down if there were repeated rapid excursions of load that would cause the dummy load temperature to rise too high.

It seems rather inelegant compared with a field winding control approach, and I have to wonder whether it's really cheaper.

This is a link for a service manual for a slightly different model, but which uses the same engine and alternator combination.

Page 63 in the PDF (numbered 62 in the document) shows an exploded view of the alternator. The rotor contains permanent magnets. There are no slip-rings for a field winding, nor windings for a slip-ring free excitation system.

Don't this type of engines use a governor precisely to control the speed to a reasonable range? When the load increased the throttle opens and produces more power, but the speed stays the same within limits, no?

I think this assumes the load produces an appreciable voltage drop in the windings. Do you know this is correct?

I would expect that as long as the speed of the rotor is maintained, the output voltage would be maintained. That requires more mechanical power which provides the higher electrical power. If the mechanical power is not there the rotor speed will drop. I don't see the voltage varying wildly unless the rotor speed varies wildly. Rather than capacitors, I see the voltage maintained by a flywheel.

Conversely, if the load is removed the excess power provided by the engine will ramp up the RPM and therefore the voltage. But this will be slow compared to a purely electrical effect and the flywheel will mitigate the voltage increase until the governor reacts and drops the power to the rotor.

I've been around many generators and they don't ramp up speed wildly when the load is removed or drop to idle speed when there is no load. But the throttle runs the power up and down dramatically with changes in the load.

Is this possible with your unit?

Rick-

Looking at Sylvia's manual, it appears that there is a "throttle control system" that includes a control motor. However precise engine speed is not as important as in generators where frequency is determined by speed. Two speeds are mentioned: 3200 and 4300 RPM, depending on the position of a "Smart" switch.

For this generator, 60 Hz output frequency is determined by a quartz crystal oscillator.

For the case where a load is suddenly changed, the engine speed should not have a violent change because the DC power supply's capacitors will buffer changes in load.

On manual page 25, it states the phase-to-phase voltage is in the order of 30 Volts for 120 VAC output. Sylvia's 600 Volt AC measurement suggests an instrumentation problem. Perhaps there is an odd waveform causing the meter to read very high. A different meter may give different results.

Fred

Yes, it's correct. It arises (mostly!) because the current in the armature produces a magnetic field that opposes, and thus reduces, the field produced by the magnets, and hence the underlying emf. The effect is known as synchronous reactance.

No, for the reason given above.

Before it caught fire, it behaved as I would expect - applying a load makes it slow down for a few seconds while it opened its throttle. Removing a load has the opposite effect. I wouldn't call it "wildly" but the effect was there.

Sylvia.

That's when the engine is not running, and is just turned over by pulling the starting rope.

But I'll get a scope on it when it's running, just in case.

Sylvia.

hell, if it's way too fast, just short the points, short the output winding too if there's too much output, generaters are inheriently current limited, no heroic effort is needed to shhort one safely.

it follows directly from what a generator is,

you need to think about it some more: what is magnetic field proportional to?

They are inherently current limited, but it's less clear that that current is always low enough to avoid burning out the windings.

Of course I could, and indeed will anyway, put fuses in to protect the windings, but having the fuses blow as a result of what is actually normal use doesn't seem like a sensible option.

Given that this thing worked for a fair while in total before it caught fire, and given that Honda small inverter-generators also use permanent magnet alternators with no field winding, there must be a reasonable solution.

Sylvia.