What Is the Best DC to AC Converter for a 500V, 50A Generator?

[yonnie]

I have a 500v, 50A dc generator (~25kw) and need to get 240v ac (rms) for operating typical household and shop appliances including those sensitive to quality of the ac. (this 240 would be run through another transformer for 120v appliances)

I've found an abundance of information on various designs but absolutely nothing on power capabilities of said designs, nor on cost of materials, how reliable, how practical, how efficient, etc...

Hopefully someone can assist in making good choices in potential designs or point me in right direction.

 

[Bob S]

There are devices called IGBTs (insulated gate bipolar transistors) that will work for both the current and voltage you are talking about, but they are expensive. You will need two, in an H-bridge configuration, to get the push-pull power you need. For 60 Hz output, design the frequency control circuit for 120 Hz, and then use a JK flipflop at the output to guarantee a 50-50 duty cycle, or else you may have some DC in your transformer primary.

 

[vk6kro]

The obvious answer is that you need a 500 volt DC to 250 volt AC inverter.
But you probably can't buy one and it will cost heaps to get one made up for you.

One possibility is to see if the generator is really an alternator with diodes.

That would mean you could get inside it and take off some AC before the diodes and then run this into a transformer.
This would do nothing to change the frequency so anything that was frequency dependent may have a problem. Certainly if the frequency was too low, anything with a transformer should not be connected to it.

 

[yonnie]

No it's not an alternator.

Can you suggest some igbt's big enough to handle the job?

 

[Bob S]

Here is a full H-bridge ckt, 69A @ 600 volt. $55.00 ea. See http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=VKI75-06P1-ND and click on "datasheet"
I have usually used the Powerex brand, but there are other good brands. You can easily go to 150 amps at 1200 volts, but they become pricey. Be sure to provide the recommended drive ckt, and emergency shut-off signal (all gates open). If you do not have a line signal frequency reference, use a 120 Hz NE555 followed by a JK flip flop to guarantee a 50% duty cycle square wave.

Here is a higher rated Powerex half-bridge. You will need two. click on "datasheet."
http://www.pwrx.com/Result.aspx?q=CM100DY-24A

 

[yonnie]

Is the bg2b powerex bd the recommended drive ckt and emergency shut-off you are referring too? These are some pretty slick gadgets, had no idea stuff like this was available.

thanks!

 

[Bob S]

The IGBT driver circuit I was referring to is the VLA504-1 recommended by Powerex. We lost a pair of IGBTs when the AC power to our equipment was accidentally shut off in a way that left two of the gates ON**, so we used the optocoupler input to open all 4 gates as soon as a fault condition was detected anywhere in the input circuit.

**You will need possibly as much as 500 MF at 600V on the top of the H-bridge to prevent mid-cycle droop. If you lose gate drive control before this capacitor is discharged through a discharge resistor, the gates may not toggle and the current in the IGBTs could exceed their max ratings. I don't know how you are going to shape the output voltage. Voltage square waves may not be ideal for household and shop equipment.

 

[yonnie]

Oh, ok, I saved and filed that but wasn't sure if the vla504-1 was what you meant. thanks for the pointer on 'smoked' parts.

Wave-shaping, how'd you guess that was my next topic? Do you think a saturable-core transformer might work?

 

[vk6kro]

"I've found an abundance of information on various designs but absolutely nothing on power capabilities of said designs, nor on cost of materials, how reliable, how practical, how efficient, etc..."

Just looking at your original post, it seems like you already have some information.
Would it be worth contacting some of the people involved and getting them to justify why you should be interested in their product?
I'm surprised you found anything that started with 500 volts DC as this would be a very unusual voltage.

I do wonder why you have 500 V DC anyway?

Could you take off power from the shaft that drives this generator and drive a 220v / 120 v alternator via an extra pulley and belt?

You probably need a unit that is ready to run and fully guaranteed, but I doubt if that is going to happen. Nobody will guarantee a rough design on paper without building it and anybody that designs and builds a one-off prototype for you is going to charge about $5000 for it. This normally gets spread over the cost of the production units, but in your case, you would have to pay the real cost up front.

There are very nice portable generators available. Because there is a very big market for these for camping, they are cheap and efficient. Just put gasoline in the tank and away they go.

 

[yonnie]

I found an abundance of designs with no specs, that's why I came to this forum. I'm using or going to be, this generator for main power. I don't get to have a constant-speed drive shaft either, and there is no motor. I had thought of putting squirrels in a cage, but haven't figured out how to keep them running 24/7. Just kidding.

Seriously though, if you have an engineer in mind or someone who would like to quote have them pm me. I'm just helping out with a project and am a little out of my normal field. It does need the fail-safes Bob S suggested and it also needs to be able to sync-up with AC power utility for re-selling back to the grid. Oh, and it needs to be a good substitute for commercial AC power. (nice clean 60hz sine wave)(120 and 240v rms) and that's just the beginning. There's other features needed later.

Bob S, got me started and a square wave will run ac motors which is good enough for some testing.

 

[calceus]

I must stress: that it is INVERSTION, not CONVERSION, you are wanting to do.

vk6kro mentioned this.

search: dc ac inverters

 

[calceus]

I'm surprised you found anything that started with 500 volts DC as this would be a very unusual voltage.

I do wonder why you have 500 V DC anyway?

Probally from a home-base solar panel system. Since sunlight is intermittent, most all solar panel systems for houses require a "helper" source of DC electricty as well.

This can be done with wind mills, and other types of dc generators.

Of course, the entire system would be "feeding" batteries of some sort.


DC is the ONLY way to "green" livin' existence.

If Thomas Edison had won this battle, OOOOOMMYYY how the world be cleaner today...

http://en.wikipedia.org/wiki/War_of_Currents

All computers, phones, and electronics are DC already, with an AC convertor some where and have been for years !

The "waste" of natural reasources from AC to DC converstions and inversions has to be eliminated.

We need many working DC appliances like washing machines, and start a graduall elimination of AC electricty all together.

DC- based systems would allow for more localized efficient production methods, like solar panels, that already exist !

The biggest hurdle to DC is heat production for colder climates.

 

[calceus]

Only in America..

Edison carried out a campaign to discourage the use of alternating current, including spreading information on fatal AC accidents, publicly killing animals, and lobbying against the use of AC in state legislatures.

Edison opposed capital punishment, but his desire to disparage the system of alternating current led to the invention of the electric chair.

Harold P. Brown, who was at this time being secretly paid by Edison, constructed the very first electric chair for the state of New York in order to promote the idea that alternating current was deadlier than DC.



Instead, the State Government of New York said: "Now theres a great idea !"


And I thought it is was "tuff" getting my posts understood on BBs...

I am a native-born USAer, but DAAMNNNNNN people!

 

[Averagesupernova]

If Thomas Edison had won this battle, OOOOOMMYYY how the world be cleaner today...

Although this is the only part of your post that I quoted, the rest of it makes as little sense as the above.

 

[yonnie]

I was not consulted in the choice of the size and capacity nor the type. 500vdc@50A was decided upon for some historic reference to Edison and that those early designs are now public domain. I'm merely assisting in adapting the power output to match to what a typical house requires as they need help with their short staff and next to no funding. (so much for the BS stimulus plan).

My guess is Averagesupernova has never seen photos of New York City circa 1900. In places, you couldn't even see the sky because of the rat's nest of copper. It's takes more than twice the amount of copper to transport an equivalent amount of power using dc than ac (a waste of limited resources). For a dc power-grid you need re-generation (small power-plant, pollution source) every mile or so. DC is just as deadly as AC, the electric chair was built to be gruesome so it would grab public attention. However, two small needles attached to a "D" cell would be enough to do the job. I wouldn't want to touch a 100vdc live wire any more than a 100vac live wire. If Edison had won the ac/dc battle, IMHO, this world would be a very much more polluted place.

 

[calceus]

If Edison had won the ac/dc battle, IMHO, this world would be a very much more polluted place.

My opinion is that in the 140 years since he "set" up the DC system, each "house" could now be independant.

The transmission issues of DC are eliminated by every "house" having its own- generated power source, elimating "power plants" all together.

Of course, that is IMHO as well.

 

[calceus]

I was not consulted in the choice of the size and capacity nor the type. 500vdc@50A was decided upon for some historic reference to Edison and that those early designs are now public domain.

I assume the "primary" source of "power" for the unit is AC current based, assuming it was originally built as an auxillery component to a wind or solar driven system.

That would mean it begins as AC at your location, converts to DC; and you are wanting to invert it back to AC.

And also then stating the "primary" state of the original generation was DC.

So, it went from DC to AC at the plant. Then AC to DC at your location. Then you are wanting to make an additional change back to AC.

 

[yonnie]

My opinion is that in the 140 years since he "set" up the DC system, each "house" could now be independant.

The transmission issues of DC are eliminated by every "house" having its own- generated power source, elimating "power plants" all together.

Of course, that is IMHO as well.

In Edison's world that would not have been possible, and we would have vast amounts of pollution as a result. In Edison's time each generator was powered by coal or hydro. If we continue with the large power corps. running our world we will soon run out of resources. IMHO we must get off our dependency of oil and coal before it's too late. Ditch the old tech and go with the new.

 

[vk6kro]

I was not consulted in the choice of the size and capacity nor the type. 500vdc@50A was decided upon for some historic reference to Edison and that those early designs are now public domain. I'm merely assisting in adapting the power output to match to what a typical house requires as they need help with their short staff and next to no funding. (so much for the BS stimulus plan).

I sometimes get presented with someone's silly choice and a request to get them out of the mess they created. My first inclination is to suggest they stew in their own juices.

For whatever reason ( I don't think you've said why ) they have created an almost impossible task out of one that could have been quite promising.

Presumably, you have some power source other than squirrels. Maybe it is an old car engine.
Someone is generating 500 volts out of this when they could just as easily generate something useful like 110 volts or 220 volts AC.

 

[yonnie]

I can tell, you really think it's important on the why. I don't really know other than it had something to do with their initial funding which is mostly spent. (the squirrel-cage is an old joke, I was trying to inject a little humor) What they are trying to do is marry an Edison style generator to a wind turbine and they asked me to volunteer help with the ac portion. There is no way, you can regulate wind energy into a constant rpm for ac power. That's tough enough to do with an engine and even then I don't think that would be good enough to feed back into the grid, which may be a future desired option. Also, since the wind doesn't blow all the time, we'll need batteries. Again, an AC power source isn't very useful at charging batteries directly. So the selection of designing a DC charging source for batteries must have been part of the decision process. I'm not so convinced that the choice of 500v is so bad. Now that I think on it, it may be a good choice. Most solar panels put out either 24 or 48v, by design and charge up banks of batteries. Well, 10 banks of 48v battery groupings would be 480v, plus you need a little extra to get them to charge from a single source. So using this voltage may be a good marriage to other 'green' power sources.

A couple years ago, I recall reading that the new battery format for future cars, trucks and RVs is going to be 48v and 24v with 12v being obsoleted. Two main reasons being less copper required for winding motors and wiring the vehicles, and more efficient power storage in battery design. This would also make commercial batteries compatible with military uses as well as solar-banks.

Drop by a car museum and take a look at the 6v battery cables and then go look at a 12v car. The 6v car used double the wire gauge than the 12v car and the starter motor is nearly twice the weight. Same will be true for 24v systems on cars, lighter, smaller, equiv. power.

As 'green' tech starts to take over, we'll probably start seeing household appliances become available as either 24v or 48v dc as a choice rather than as a 'special' high-priced option. And we may start seeing homes with 24v or 48v dc outlets.

 

[vk6kro]

That makes some sense.

If you have 480 volts worth of batteries, could you get to the center point of them and use this as a neutral point?
If you switched to sample the +ve then the -ve voltages relative to this central point, you would at least have a crude square wave AC waveform. It would have a peak value of 240 volts and an RMS voltage of 240 volts too. This could be done with power FETs or IGBJTs.

You could turn this into a stepped waveform by selecting the 25% and 75% points as well.
This would give a 0V +120 +240 +120 0V -120 -240 -120 0V waveform.

It is starting to look a lot better.

 

[yonnie]

What a terrific idea. We were going to have a meeting today, I'm going to bring this up.

 

[mheslep]

That makes some sense.

If you have 480 volts worth of batteries, could you get to the center point of them and use this as a neutral point?
If you switched to sample the +ve then the -ve voltages relative to this central point, you would at least have a crude square wave AC waveform. It would have a peak value of 240 volts and an RMS voltage of 240 volts too. This could be done with power FETs or IGBJTs.

You could turn this into a stepped waveform by selecting the 25% and 75% points as well.
This would give a 0V +120 +240 +120 0V -120 -240 -120 0V waveform.

It is starting to look a lot better.

What a terrific idea. We were going to have a meeting today, I'm going to bring this up.

Tapping lower voltages from from multiple battery cells across the 480DC does provide a lower DC source voltage but it increases your complexity quite a bit, perhaps to the point of being impractical. First, if this is a one of a kind project then almost certainly you want to buy the DC-AC inverter(s), not build them, if you value your time at all; it will take substantial time to do this with all the appropriate safety measures.Second, multiple taps means multiple inverters, and then unless the loads are completely separate in both load and toleration for failure, you'll have to synchronize those inverters somehow. I'm not aware of commercial inverters that provide such a feature. Phase would have to be sync'd and active measures taken to enforce current sharing. And regards DC/AC transmission up thread: DC transmission is indeed more energy efficient than AC, but it did not become practical until recent technology enabled economic up/down conversion of high power and high voltage DC to DC. Back in the day, Tesla was right (about AC) and Edison wrong. Even now DC transmission is more expensive than AC because, for all but the longest runs, the DC/DC conversion is still more expensive than AC/AC transformer magnetics even given the greater energy loss in AC lines.

 

[vk6kro]

I think it is necessary to read the whole thread of this story.

For whatever reason we were stuck with a 500 volt DC supply. This could not be changed.

The problem was to somehow produce a workable 110 / 220 volt supply from this.
Commercial inverters to do this were unavailable and getting one made as a one-off was likely to be very expensive. If you know of any please let us know.

So, tapping into the battery bank was an adequate solution.

It would have certain problems. For instance, the new neutral point at the center of the battery bank would need to be the only Earth point on the battery bank and this may be a problem if the generator was earthed on one side of its output. I expect it would be.

Failing that, it would be possible to select a larger number of points on the battery bank and get a sufficiently cleaned up waveform to run it through an isolation transformer.

Actually producing this supply would not be difficult.
From a signal source, which could just as easily be crystal controlled or synchronized to the mains supply, you turn on, in succession, a series of optocouplers which each turn on a suitable switching semiconductor. This is yet to be decided, but possibly an Insulated Gate Bipolar transistor or bank of them.
This would need researching because the requirements are rigorous.
The outputs of these switching devices would simply be joined together to produce an AC output, relative to the center neutral point on the battery bank.

Another problem I see is that the loading of the batteries would be uneven. Hopefully enough other current would be drawn from this supply to make this unimportant.

Multiple inverters would not be used, and time appears to be available gratis.

This is not a trivial project, but at least we have a way forward.

Safety would be a vital factor too, with a power source like this one. 480 volts worth of lead acid cells is a seriously lethal supply.

 

[mheslep]

I think it is necessary to read the whole thread of this story.

For whatever reason we were stuck with a 500 volt DC supply. This could not be changed.

The problem was to somehow produce a workable 110 / 220 volt supply from this.
Commercial inverters to do this were unavailable and getting one made as a one-off was likely to be very expensive. If you know of any please let us know...

<shrug>
Xantrex GT4 or GT5, 240VAC out.

GT4.0N 240-480VDC 4000W
GT5.0 240-550VDC 5000W
http://www.solar-earth.ca/products/inverters.htm

I would expect costs of $0.7 to $1.0 per Watt.

Edit: On reread I see the OP wants a 50A system - 25KW - we don't often see that questions about that kind of power here so it hit my blind spot . So this really is industrial sized equipment. However there's no mention of whether the load is monolithic or distributed. Maybe five of these would work on five separate loads, all fed by the same generator. The next step up with Xantrex is three phase - as expected - to 100kw.
http://www.xantrex.com/web/id/254/p/2469/pt/23/product.asp

Or?:
http://www.brusa.biz/pricelist/e_pricelist.htm
DMC514
52kw, 120-480VDC

 

[mheslep]

...

So, tapping into the battery bank was an adequate solution.

It would have certain problems. For instance, the new neutral point at the center of the battery bank would need to be the only Earth point on the battery bank and this may be a problem if the generator was earthed on one side of its output. I expect it would be.

Failing that, it would be possible to select a larger number of points on the battery bank and get a sufficiently cleaned up waveform to run it through an isolation transformer.

Actually producing this supply would not be difficult.
From a signal source, which could just as easily be crystal controlled or synchronized to the mains supply, you turn on, in succession, a series of optocouplers which each turn on a suitable switching semiconductor. This is yet to be decided, but possibly an Insulated Gate Bipolar transistor or bank of them.
This would need researching because the requirements are rigorous.
The outputs of these switching devices would simply be joined together to produce an AC output, relative to the center neutral point on the battery bank.

Ok, I am still not sure I follow. So the idea is to simply run the IGBs at 60 Hz, and then throw the ~square wave into common power transformer? Given a common 60Hz power transformer, isn't that going to throw away all the harmonic power at 180Hz, 300 Hz, etc, and burn it up as heat in the windings/core?

 

[vk6kro]

Yes it would if it was just a square wave.

That is why I said this:

Failing that, it would be possible to select a larger number of points on the battery bank and get a sufficiently cleaned up waveform to run it through an isolation transformer.

 

[mheslep]

Yes it would if it was just a square wave.

That is why I said this:

Failing that, it would be possible to select a larger number of points on the battery bank and get a sufficiently cleaned up waveform to run it through an isolation transformer.

Yes, which then must draw the uneven currents as you mentioned, but maybe that's ok. I still think that would mean considerably plus rating the transformer up (20-30%?) from 25kva, and then still coupling more harmonic noise into the secondary than electronic appliances want to see.

 

[yonnie]

Well, that solar Earth place seems to have a 7kw unit and on the 2nd page they hint at an 18kw but couldn't find a part number. I did not find anywhere on their website a reference to the quality of the output voltage.

The meeting last week was a no-show, but I'm definitely keeping this good info for reference.

 

[mheslep]

Well, that solar Earth place seems to have a 7kw unit and on the 2nd page they hint at an 18kw but couldn't find a part number. I did not find anywhere on their website a reference to the quality of the output voltage.

The meeting last week was a no-show, but I'm definitely keeping this good info for reference.

3% THD, 0.99 PF is typical for most of that Xantrex gear.
http://www.xantrex.com/web/id/172/p/2604/pt/23/product.asp