Inverters shutting down - why?

[ecogen]

As you say the neutral earth is still common to both. This could be relevant if the inverter incorporates an earth monitoring method which involves a test signal between neutral and earth?

Taken from the link that h1267 kindly supplied.

(There is one important note you should be aware off: Transformerless inverters have very strict ground-fault detection build into them (as required by the UL and CSA norms). When the inverter is placed behind an isolation transformer this mechanism will not work properly, and erroneously indicate a ground fault. This will happen to the slave inverter in the wiring diagram, behind the isolation transformer. For that reason it is necessary to switch off the 'Riso Measurement' option inside the inverter. This is done through the 'Advanced' menu choice of the AuroraInstaller program and involves an inverter-specific access code. If you are planning to use inverters in a stacked configuration please let us know, and we will make sure the inverters are programmed properly before sending them out to you, and that they are properly labelled as master and slave inverters so you will know which one goes where.)

[guydewdney]

Would tieing -ve of the bridge rectumfrier to earth solve the problem?

[Alan]

No

Imagine a Windy / Sunny Boy inverter. Each one has got a transformer inside
to isolate the alternator / inverter electronics from the grid connection.


You can have as many of this type of inverter connected to one turbine  / water wheel as you like.

The down side is each inverter would have 2 – 3 %  extra losses due to the transformer.

If you have a ground fault on either the positive or negative inverter input it would still work.

The Aurora has not got a transformer. It is not galvonically isolated between its input and its out put.

That is why the software is in place to detect / make sure the D.C. Input of the inverter is isolated above ground.


Ref : wish list.
( Still got software issues due to multiplexing, project ongoing. )

Regards

Alan

[w0067814]

IGBT devices in inverters are typically used in their saturated region - ie fully ON or fully OFF. They are pulse width modulated to control the average current / voltage requirements.

One method of checking for over current fault situation is to monitor the voltage across the transistor when switched ON. This should be pretty low - ~2v for a high current IGBT. If the voltage rises when turned hard on, this is a sign of over current and can be used to trip a hardware fail safe.

[Edit: Qualify the above statement - this is a technique used by the inverter tens of thousands of times per second. Not a test engineer!]

Perhaps you have a short circuit, or perhaps you have an issue with the drive circuit - ie the controller believes that the transistor is ON and therefore expects to see a low voltage across the device, but instead it is off, so the full DC supply rail is across the device (circa 380 volts I would imagine).

If poking around inside be very careful of the voltages, and don't assume anything to be referenced to ground - trust me - I've zapped myself a couple of times on UPS.

-Tim

[guydewdney]

Well - I installed the 4KVA isolation transformer today.


yeah! no earth trips!

But no power out - at all! Inverter reports that 'all is OK' input voltage 480V DC, grid 241.4V 50.01Hz (or whatever it happes to be at that moment....)

Power curve (programmable on these) set to do 500 watts at 500 volts, reducing linearly to zero at zero. So it should be making 480 odd watts. But its just sat there.

If I alter the curve, the inverter re-boots. It does this nicely (all communication is fine) - I can hear the relays kicking in and out (normal). Yet nothing is being 'made'

odd. very odd. Phoned power-one - they seemed confused. Theres no error messages.

[Alan]

Quote “ Phoned power-one - they seemed confused. Theres no error messages. “

If the V grid nominal setting is not close to the actual grid voltage the wind curve data is not followed and the out put power is very low even though the D.C. voltage is well up the software curve.
The problem gets worse when you have more than one inverter and you need to use a separate transformer.

To change it you need the Installer Code under the Advanced section. In the wind curve data section.

Being remote / village fed from a small transformer this voltage can vary quite a bit. And fall outside the statuary + / - grid voltage limitations.

Regards

Alan

[guydewdney]

Dragging this back up - I fixed it by re-installing the power curve. Must have needed a sort of re-boot.