# How can I convert very high, fluctuating, AC voltage to 240VAC at 50Hz?

I'm about to finish a generator-alternator setup which will generate 9-phase AC electricity at (as yet unknown) very high voltage, most likely around 4800V. I'm not sure what the frequency will be, as I'm going by trial and error in areas, but I think much higher than the 50Hz required for UK mains electricity.

I need the final result to be a smooth-enough flow of AC electricity at 240V with a frequency of 50Hz. At the moment, I think the below would do it but I was wondering if there are any simpler and/or more energy-efficient ways of doing this:

1. The alternator produces (for example) 4800V at a frequency much higher than 50Hz (I'm estimating 400Hz based on the likely RPM of the alternator). It is very likely that both the voltage and the frequency will fluctuate significantly.
2. This is passed through a number of step-down iron-core transformers so that the voltage is now fluctuating around 240V AC.
3. A voltage regulator (most likely from a car's alternator) then smooths out the voltage to as close to 240V AC as possible.
4. A silicon-controlled rectifier (thyristor) converts the flow to DC, preferably at the same voltage.
5. A capacitor (or more than one if required) is connected to smooth out the DC signal.
6. A 6, 12 or 18-pulse inverter then converts the DC signal back to AC to be fed into the grid.
7. Another voltage regulator and/or transformer could be fitted just before the grid tie-in point, if the voltage were not yet stable enough at 240V.

I'm very much learning as I go and may have misunderstood a number of points, but from the research I've done, I'm left with the following questions:

1. I presume that by generating 9-phase AC electricity, I am still able to 'merge' the phases such that there will be 3 bigger phases? UK mains electricity is naturally 3-phase and, as such, that's what will need to be fed into the grid.
2. Is a voltage regulator (I've found a supplier of 250V ones) suited for what I'm intending to use it for?
3. How can I tell what the frequency of the AC current will be once it's inverted back from DC?
4. Are there any steps that could or should be altered to make the whole process more efficient?

Much thanks for any and all advice!

Cheers, Chris

• I'm not sure if the voltage regulator step is required if your thyristor and inverter stages are smart enough? Also, if you're going to feed the grid, building regs/Part P/CE marking may apply. Jan 19, 2015 at 22:12
• Also, I'd assume you'd merge the phases at the DC step. The capacitors required may be horribly expensive. Jan 19, 2015 at 22:14
• I had originally planned to merge at the AC step, if possible, to minimize energy loss? Would merging at the DC step be more efficient? Thanks Jan 20, 2015 at 11:47
• Thyristors are dumb, the intelligence would be in the trigger logic. I'm somewhat over my head here as well. But I suggest building a 1/100 voltage scale model (48V -> 2.4V) of the system which can be debugged without risking death. You don't get a lot of trial and error at 5kV. Jan 20, 2015 at 12:06
• (Since this old post has been dredged up again) Don't even think of connecting a grid-tied inverter to the mains until it's been certified to either G83 or G59 standards (whichever is applicable). A badly designed inverter could be dangerous for everybody else on the same supply. Jan 1, 2016 at 19:43

A voltage regulator (most likely from a car's alternator) then smooths out the voltage to as close to 240V AC as possible.

A 12 V alternator regulator will not work on 240 V AC. It will be destroyed.

A silicon-controlled rectifier (thyristor) converts the flow to DC, preferably at the same voltage.

Use diodes.

A capacitor (or more than one if required) is connected to smooth out the DC signal.

With poly-phase rectification no smoothing is required. One of the phases is always up.

A 6, 12 or 18-pulse inverter then converts the DC signal back to AC to be fed into the grid.

What is meant by this?

Another voltage regulator and/or transformer could be fitted just before the grid tie-in point, if the voltage were not yet stable enough at 240V.

No. You would rectify the low voltage and then convert to 3-phase 240 V AC using an inverter. This is standard on uninterruptible power supplies.

I presume that by generating 9-phase AC electricity, I am still able to 'merge' the phases such that there will be 3 bigger phases? UK mains electricity is naturally 3-phase and, as such, that's what will need to be fed into the grid.

No. You can't merge phases.

Is a voltage regulator (I've found a supplier of 250V ones) suited for what I'm intending to use it for?

No. Use an inverter.

How can I tell what the frequency of the AC current will be once it's inverted back from DC?

The inverter determines the frequency.

Are there any steps that could or should be altered to make the whole process more efficient?

Yes. Generate with three phases and, if possible, at the correct frequency and voltage.