Convert 1 Phase AC Source to 3 Phase

Try a Google search. Use "three phase converter". I did and got 330,000 results in 0.28 seconds.

Of the many, many results, here is one link I found: http://www.metalwebnews.com/howto/ph-conv/ph-conv.html

Hope this helps - KJK

Great photo. It's appropriate that we salute all the people who gave their time, and lives, in service to our country.

Amen to that brother!!!!!!!!

It's a good idea to plan it based on what your needs are, but I can't read your mind as well as I used to be able to.

It is good idea but the voltage and current are not balanced,and I'm as experimenter i can not rely on such unbalanced 3 phase power source to conduct my experiments.

i need it for testing power supplies only

I have only single phase in my workshop and i need to test 3 phase power supplies

Ah. Can you tell us the 3-phase voltage and current for your supplies?

There are a number of workarounds for 3-phase power-supply testing, without 3-phase power, some of which can be quite useful for testing certain aspects of a power supply.

You can test the supply at limited power levels using single-phase AC. You may have to use a variac, etc., to obtain an AC voltage equal to one of the three phases, such as 208 volts. Many power supply failures are severe enough to manifest themselves under low load-power conditions, and after repair, the supply will work properly again at high load-power.

You can do low-load-current tests, up to 1/3 to 1/2 of the supply's rating (to avoid overheating whichever rectifier diodes are in use). You can do short-term full-power pulse tests on the supply, tests that are shorter than the discharge rate of the bulk caps (e.g. 3 ms max). If the power-supply load-step transient-response is fast enough, and if the supply is working properly, it should pass those tests. You can measure the input power and determine power loss. You can do bulk-capacitor discharge-rate tests by seeing how long a specific load current can be delivered after you remove the input power. This measures the bulk capacitance value. You can remove the load and then remove the power, wait a bit, watching the output voltage, and then reapply the load. This is to get a handle on the bulk-capacitor leakage currents. Both are relevant when looking for a bad capacitor.

If you have a detailed schematic of the power supply, and you thoroughly understand certain design details, you can use DC power, charging the bulk capacitor through the rectifier diodes. The supply must not have any 60Hz transformers, even small ones! You can do steady-state tests up to the rating of the your DC supply, provided, as with single-phase AC, that you stay below about 1/3 of the supply's rating to avoid overheating diodes. You can do all the usual tests mentioned above. You can also do turn-on in-rush-current measurements, equal to the worst-case when the AC line is nearly at its peak, because that's the value of the DC voltage you're using.

The bulk-capacitor discharge tests we talked about are easier with a DC source, because you can turn it off or disconnect it instantly, without worrying about synchronizing with the ac phase (don't make the mistake of adding its bulk- and output-capacitor storage to your supply's bulk-storage measurements).

One test that's hard to do without a proper 3-phase source is prolonged full-power operation, looking for overheating problems. But if your power supply passes all the low-power tests you would normally give it, and passes short full-power pulse tests, at least you're left with a limited number of failure modes.

Just to point out the new world of 3-phase power, e.g., VFD for driving motors, etc. First create a DC source (e.g., from single-phase, with a PFC converter, or from 3-phase), 360V or higher. Can be balanced, e.g., ±180V. Second, create sine waves using three half-bridge IGBT switches, using, Third, controller to generate IGBT chopper-drive pulses to establish 3-phases of desired AC voltage and frequency.

It is excellent idea but we need further details with some illustrating drawing with further explanation.

Thanks for the schematic,but by the way is IR company still existing or sold out ?

Oh no, they're doing fine - just explore other IR links from the app-note web page I gave you. Their stock price has been going up steadily for the past year, up by about 2.5x. (They did sell part of their power mosfet line to Vishay a few years ago, but apparently that was a minor deal.)

Is there any other similar chip for single phase inverter from IR ?

Yes, check out their website, they have a whole line of parts, especially ones that drive partial setups (e.g., one leg only).

A number of other companies also offer these things, which are well suited for non-motor uses, such as fluorescent bulbs, etc. For example, I was just looking at some NXP parts, with p/n in the UBA20xx family, before seeing your post.

Quite a fab houses have figured out the 600V semi IC scene. The growing market for these things has driven the prices down to well under a buck in quantity. I think all kinds of design opportunities await the curious engineer.

Is such inverter is safe to drive three phase transformer or motor ?

Yes indeed!

The chopping frequency is very high (e.g. 50 to 200kHz = 1 to 4 thousand chops per cycle) compared to the transformer or motor's 60Hz max normal operating frequency. Mains transformers have enough leakage inductance to filter most of the chopping signal, and motors have enough self-inductance and inertia to not respond.

One issue that remains is RFI, the potential RF interference created by these circuits. Most designs benefit from the relatively-slow response of IGBTs (200ns risetime), which are commonly used in these circuits because they are cheaper and have lower Ron than high-voltage mosfets. (The IGBT / mosfet high-voltage conduction-tradeoff voltage comes pretty low, at about 200V.) And the better units have also added serious internal RFI filtering. It's hoped that they are relatively benign.

What the maximum KVA could be obtained from this three phase inverter?

Is it possible to connect several of this inverter in parallel to increase the total KVA ?

Many of these designs use 600V parts, but a significant fraction use 1200V parts. Given typical safety margins, these allow for sine waves with say 1050V peak-peak amplitudes, or 370 Vrm. As for current capability, large IGBT modules are commonly used, and many of these are rated into the thousands of amps for continuous use. So a rating of 750kVA, for example, would not be unheard of, and higher powers are possible using the same approach.

The part pictured is a 3-transistor eupec fz2400r17kf6c, rated at 1.7kV and 2.4kA (9.6kA peak), which Mouser sells for $3,494.25 each. A similar-looking part is their fz1200r33kf2c, rated at 3.3kV and 1.2kA, you get the tradeoff idea.

I buy mine on eBay, got one for $99 (they usually go for more), and I enjoy puttering around and making little things out of them, 2MW pulse generators, things like that. I have a friend who pushes them beyond spec into the 20MW territory for his TMS work.

The designs do not readily lend themselves to parallel operation that I know of, or maybe more accurately, I should say I don't know how they would do it. Perhaps with transformers. Leakage inductance is your friend. I know there are higher-voltage and higher current parts. And I know that at the upper extreme, room-sized apparatus converts say 375kV DC to AC for DC transmission-grid ties to AC lines. These beasts are ultimately rated into the MV GW territory, e.g., see the thread "7.2 GW, 1.6 MV DC Pole-Pole 2,000 km Transmission Line" That system was made with 8.5kV thyristors with 6-inch diameters, "operating at a direct current of 4000 A," using design techniques that I don't understand (anybody knows, speak up!). I posted photos and links in that story, check them out. It's all the same basic idea: start with DC and generate the exact AC voltage, frequency and phase you want.

So the answer is, you can go from large parts like the above-mentioned 4000 kVA module's maximum rating (for each of its three transistors), down to a pair of small parts like a 3 kVA irg4rc10, in little TO-252 D-Pak smt packages (shown in a magnified view), for each leg of a 3-phase drive. The cost ranges from $3.5k to $1.

Three of the larger modules may be powering a locomotive engine, six of the smaller parts may be running your washing machine, smoothly accelerating and reversing its motion, back and forth, then spinning up for the drying phase.

please do you have a simple circuit

Simulate?

Just imagine an inverter.