Building an Inverter

[BrianEllul]

Hi

I would like to build a 240v sine-wave inverter. It will run from a 240v dc battery bank! Any circuits around?

Regards

Brian

[thunderhead]

There are a few circuits about, but they normally convert from 12v DC.  Easy circuits for home construction tend not to be very efficient.  This is because of something called "core losses" in the transformers that are used.  If you rewind a 2kW transformer from a microwave oven, for example, you will probably need something like 200W of power on the primary before you draw a single watt from the secondary.

Commercial designs for invertors normally convert the 12v into +350v and -350v and then use a switched mode amplifier and filter to convert this into 240v AC.  You understand that a RMS voltage of 240v sinewave actually reaches 340v at the peaks (and -340v at the troughs).  There is still some "core losses" in the output filter and in the circuit to turn 12v into +/-350v, but they are normally less than at mains frequency, since the frequencies are higher and so the magnetic cores are smaller.

Why 240v? If you were aiming to produce a simple squarewave output, +240v and -240v might make sense - a total battery voltage of 480v in all.  The circuit would be much simpler, but many things will not run correctly on squarewave input.

There are two good reasons homebrew circuits convert directly through a single transformer: it is much easier to get the circuit to work, with cheap components; and it is harder to electrocute yourself, since the output voltage is not stored in capacitors!

Fooling with homebrew sinewave invertors (and switching transformers according to load) is on my to-do list: but I am well qualified and experienced in electronics, and I'm planning to do it with single-conversion transformers, not with two 350v rails.  I value my life and health too much to want to fool with that much DC, and (since I'll publish circuits) I don't want to be responsible for the death and injury that would surely follow a DC design.

If you want more than sketched-out, untested schematics, you'll have to wait, though: my to-do list is fairly full for the rest of this year. :-(

[Phil Timmons]

Yeah, I am with Thunderhead on this.

I am pretty sure we can design a reasonable DIY inverter -- but there are some potentially deadly downsides to a boo-boo.  That is part of engineering ethics -- make your designs safe for the folks that use them.

If you all want, I will help along on a generic DIY inverter design, but this should have a lot of "internal" safety reviews between us so that anyone following along on the path does not get hurt or harm others.  

I suppose a good start would be a set of generic "first draft" specifications?

e.g.  Inverter Shall:

e.g.  Be grid feedback compatiable.

e.g.  Ground Fault protected.

e.g.  Operate on an intake voltage(s) of ___

e.g.  Produce ____ volts AC, ___ cycle ~ single or poly-phase?

Maybe build it in modules so the current capacity (amps) can be built up in parallel?  This would let folks build up or down according to the their needs?

[River Goat]

Hi Brian, I just saw your post and was wondering why you have 240 vdc and need 240 vac and at what power levels? 25 years ago I had a 12 vdc system that I lived with for several years and found that between the losses in conductors and especially connectors I was losing close to half of my power. I wanted to change to a higher voltage to reduce the losses and decided on 120 vdc because standard lights, motors and a few other devices will work on 120 vdc with little or no modification. And the losses are cut by a factor of 100!!! The 120 vdc system have served me well for over 15 years. DC is harder to switch than ac so I use heavy duty switches and 480 vac circuit breakers. Most of the newer electronic devices work directly on 120 vdc like, compact fluorescent lights, tv, vcr, sat receiver, and almost anything that says 100-240 vac will work on 120-300 vdc. My point is there may be a better way to get where you want to go.

I like high voltage dc and think its no more dangerous than ac of the same voltage.

Good Luck

Jerry

[hbelanger]

Just a little curious how do you charge 120vdc battery bank ?

[wdyasq]

I'm wondering why he didn't do a search and find the one that has been built, proven and schematics are "open source" - there is one out there designed, built and tested by a major university.

Ron

[Phil Timmons]

Looks like he has only been around here a week . . .

I suppose you could post the link for him?

[River Goat]

I have 36-4 volt pv panels in series in one array and 12 strings of 3-48 volt panels in another. My wind genny is wound for 150 vdc and a 120vdc pm motor belt driven by a 5hp engine for backup power; they all feed direct to the main power buss with a home-brew shunt regulator.

Jerry

[wdyasq]

Looks like he has only been around here a week . . .

I suppose you could post the link for him?

--------

You could do it Phil.....

I'd have to do a search to get it myself.  Not that I'm lazy, just if he wants the information, it is there - it is probably on the fieldlines site - google on fieldlines seems to work well.

One should not the original posted has not replied - is he/she/it stupid, lost their way or just doesn't care? Should one spoon-feed new folks and change thier diapers too?

Ron

[BrianEllul]

If it's too much trouble helping 'lazy' people why do you take the trouble to answer back and reply with useless comments! If questions seem stupid to you, just skip them, don't waste your time answering them! there are other people who WILL answer back in a polite manner!

And I do find my my around! Don't you think that you're the only smart guy who can navigate around this forum!

BTW. I did search on this web site but the link to the inverter schematic was dead.

[BrianEllul]

Hi

Thanks for your comments! I missed out the fact that I'll need +/- 240dc to get 240ac (without using any transformers). I just happened to receive twenty 12v batteries (about 1 year old) and was thinking if better to connect them to my existing 24v DC bank or else try something different!

After reading all the comments.... I've changed my mind in playing with a high dc voltage inverter!... I'll stick with 12/24v designs!

Regards

Brian

[thunderhead]

If you all want, I will help along on a generic DIY inverter design, but this should have a lot of "internal" safety reviews between us so that anyone following along on the path does not get hurt or harm others.

The safety features I'm looking for are "no more than mains voltages" and "all voltages go away when you turn it off".  My objection to DC is that it doesn't go away when you turn it off.

If the constructor can't deal with mains voltages, they shouldn't be fooling with mains wiring - and the wiring that comes out of an invertor is mains!

I suppose a good start would be a set of generic "first draft"

specifications?

While we're dreaming, mine are:-

1. sinewave <6% distorted, maximum output of at least 240v/125A, 3-phase
   
2. modular design to make it efficient, with more modules connected as load increases, and the supply cut off if the load is too high for the number of modules.
   
3. AC rather than DC loading
   
4. less than ten watts in standby - with load detection using mains-voltage sinewave AC
   
5. 5ppm frequency accuracy measured over the year (less than five minutes)
   
6. as many junk/recycled components as possible, and cheap, readily available components for the rest
   
7. adjustable cutoff voltage so I can use NiFe batteries.
   
   

The ones of your list that I'd question are:-

e.g.  Be grid feedback compatiable.

There is no way a homebrew generator circuit will be grid feedback compatible.  It won't be allowed to be, since it will not be able to be tested.  Or rather, each one will have to be sent to the test house individually, which will be prohibitively expensive.

Besides, I'm starting to worry about grid-tie.  I think that, if there is too much grid-tie in your neighbourhood, the impedance effects might actually cause blackouts when faced with complex loads.

e.g.  Ground Fault protected.

No.  This is not something that belongs on an invertor - it belongs on the distribution board.  Putting the ground-fault circuitry in the wrong place can actually make things less safe.  The invertor should have the case tied to earth, and one side of the output (or the centre of the three-phase) should also be tied to earth.  Neutral only appears the other side of the RCB on the distribution board, after the "protective earth" pole that is buried in the ground.

In short, a good invertor can be thought of as an electronic generator - and should be wired as such.

[wdyasq]

http://www.epanorama.net/counter.php?url=http://www.energychallenge.org/2001Reports/tamu.pdf

Ron

[ghurd]

Yikes!

If the plans didn't scare me enough (they did),

the $750 for parts for a 1500W unit sure did.

G-

[thunderhead]

There's no code for the DSP.  So there's not enough information here to build the complete unit.

[Nando]

Brian:

To use a 240 volts battery bank for a 115 / 230 VAC is quite common, I have designed at least a dozen of them, from low watts to high watts levels.

The procedure is to take the 240 Volts using a switcher to generate the Negative 240 Volts.

Having the two voltages then one can generate two 115 VAC out of phase to have 115 / 230 volts output.

Go to GOOGLE and type TAMU.PDF find it and download the article & read the section that makes the sine waves to see an example, also the conversion from 48 volts to +/- 200 volts.

If you are using 24 or 48 volts, still you will see the positive and the Negative voltages each (200 to 240 Volts DC).

Good luck

Nando

[weirdo557]

couldnt you just use an H-bridge with some high power mosfets or darlingtons and switch them with some sine wave input from a small ac motor rotating at a set speed?

[oztules]

I think you could use the h bridge to get square wave 240v rms, but to get sine wave from this will only produce about 160 or so volts rms in sine wave form.... ie the peak of the sine wave only will be approx 240v.... rms value will be about 160vac.

[commanda]

Not to mention the efficiency. Running your output devices in the linear mode means they're dissipating a fair bit of power. I know I should be able to rattle off the figures off the top of my head, but it's been far too many years since I did high powered audio design.

In real rough figures, if you're putting out 1000 watts sine wave, you're going to be dissipating about 400 watts.

Amanda

[dinges]

Commanda,

I was recently reading a story about driving 3phase motors not by VFDs (variable frequency drives). The report was going in detail into the extra burden this puts on motors.

But, how about simulating a sine wave by means of PWM; this means the switches won't operate in linear mode (thus, high efficiency & little heat developed).

Check out this link, page 7, fig. 1.1 :

http://www.motordoc.net/ebooks/RepairBook.pdf

The question is, whether any type of electronics would work of such a simulated sine wave.

[commanda]

But, how about simulating a sine wave by means of PWM; this means the switches won't operate in linear mode (thus, high efficiency & little heat developed).

google "class d amplifier".

Amanda