Paul and Sabrina's Cheap 3 Phase Inverter (AC Controller) with Field Oriented Control

[P-hack]

inductor value is going to be critical especially if zvs is still on the table.

Interesting construction hints from that site, looking at the 80uh 300a jobbie. maybe 16 turns, and 10 strands (insulated), no twist, and with the size of the hole remaining I'm guessing it was done by hand. Any guess on the gauge? or the toroid size? Wire looks a little thick for 20khz to me. I would think you would want at least 1/2 twist (full twist I mean) over the length of the wire bundle to even out the lengths.

[cts_casemod]

They are produced by spec so we cant say for sure. They provide the toroid size on this table:



The traftor sales man called this afternoon and said the request was forwarded to their new project/prototype engineer for technical review.

I also received a sic diode and replaced the hyper fast diode on my PFC converters with it. Surprisingly the losses increased by 4W with equiparable heatsink temperature. Both are rated to 8A with the Si having a worst case recovery time of 30ns.

This was tested at 120KHz and a current of 2.90A @ 415V.
Perhaps only worth it above 200KHz??

[P-hack]

I like the construction of the prius inductor, but this paper gets into powdered cores in a similiar appplication (200v-500v) though the power levels are a little lower than our target (70a@510v). I've never seen an edge wound toroid or even a c-core (I suppose the prius kinda qualifies as a C core). I was hoping to see a round one somewhere.
http://www.jeet.or.kr/ltkpsweb/pub/p...aspx?ppseq=394

[P-hack]

also, not sure if I posted this, but mhw has some stock in csc cores, i.e. they have 26 of these:
CS1625026E - Sendust - Powder Cores | MH&W Direct , the csc cores seem to be what that excel spreadsheet is targeting. They have a search tool so cross referencing the csc catalog is helpful once you understand the part number system.

I got some CS740125s from them after much back and fourth (wish I would have gotten CS740060), but will probably be suitable for a few kw charger, just deciding if I want multiple 16 or 26 gauge magnet wire "in hand". Price per lb is about the same, and 26 will conform easier, though I'm sure it has other compromises.

they have other cores, only looked at csc so far: http://www.mhw-intl.com/wp-content/u...talog-2011.pdf

edit, looks like emw has their 70 amp pfc inductor back in stock, air flow needed (a shroud to force the air next to the windings inside and out could be useful). http://emotorwerks.com/products/onli...ated_product-7 This looks like a comparable inductor from csc (X2): http://direct.mhw-intl.com/ch1016060e.html , so not a bad deal from emw there, not sure if the windings are right, but can play with it in the spreadsheet for comparison.


Edit, didn't mean to interrupt a switching discussion, that needs to happen.

[e*clipse]

Originally Posted by cts_casemod

I also received a sic diode and replaced the hyper fast diode on my PFC converters with it. Surprisingly the losses increased by 4W with equiparable heatsink temperature. Both are rated to 8A with the Si having a worst case recovery time of 30ns.

This was tested at 120KHz and a current of 2.90A @ 415V.
Perhaps only worth it above 200KHz??

That's very interesting. Would you mind sharing the part numbers for the two diodes? This is mostly due to increased conduction losses?

I'm in the middle of designing the power stage of this inverter for using SiC switches (both mosfets & diodes), and this looks like some good "sanity check" type experience.

Thanks a bunch,

E*clipse

[MPaulHolmes]

Did anyone's hot air soldering tool from China last longer than 8 hours of use? That's how long my lasted. haha. And when it held temperature, the whole house lights flickered. I think it had a power factor of 0.

[e*clipse]

Mine is still going ok. The soldering Iron that came with it isn't so wonderful...

My only gripe about mine is that I wish I could slow the airflow down a bit, otherwise the temperature stability is really good and it works great.

- E*clipse

[Astro]

Paul, mine is still going. I use it heaps.
I do wish the cables were slightly longer at times. But that is because i work far to one side of the unit rather than in front of it.

Is it just the hot air part that has stopped or is it the whole unit?

Did you get the one with the air from the main unit or the air from a fan in the heat wand?

Mine came with what i believe to be a replacement hot air ceramic element. When i saw that i thought oh oh, if they supply a spare one then it must fail easily but no problems so far.

The soldering iron works well, heats quickly and holds its temperature quite well. I haven't soldered much really heavy stuff, just wires and SMD stuff.

The heat gun is usually set to 300 C and the soldering iron 250 C. Heat gun air flow at 50%.

I haven't tried either of them at full temperature.

I haven't noticed any flickering of the lights in the house. My house has all LED lighting so maybe they cope better. I did notice that the display on my LED monitor for my computer blanks for a second when i initially plug in the reflow station. They are plugged into the same wall outlet. I assume the reflow station has some sort of uncontrolled inrush to charge the capacitor(s). So the equivalent of short circuit for a fraction of a second. After that there is no issues.

[cts_casemod]

Originally Posted by MPaulHolmes

Did anyone's hot air soldering tool from China last longer than 8 hours of use? That's how long my lasted. haha. And when it held temperature, the whole house lights flickered. I think it had a power factor of 0.

Probably thyristor regulation. These guys don't care anything about line noise. Mine has some problem with the air pump, but I never bothered fixing it.

Originally Posted by e*clipse

That's very interesting. Would you mind sharing the part numbers for the two diodes? This is mostly due to increased conduction losses?

I'm in the middle of designing the power stage of this inverter for using SiC switches (both mosfets & diodes), and this looks like some good "sanity check" type experience.

Thanks a bunch,

E*clipse

Yes. I also tested on a high voltage buck converter.

The sic diode is a CREE SCD1060. Si diode's RHRP0860 & 1560. I believe these are more suitable for very high frequency power supplies in the MHz range, or where no load efficiency is of importance.

Edit:

I haven't tried with SiC mosfets as well, but I did try with an IGBT without freewheeling diode for the buck converter, to take the body diode losses out of the equation.

[On The Run]

Hello there,

finally managed to catch up on the thread.

Today I'll stick my head out of the lurkers trenches for two reasons:

First I'd like to pay my respect to the fantastic work going on here.

Second: The solution to the voltage issue without putting tons of batteries in series and having to build a seriously beefy buck booster or whatever could be here:

w w w.asd-sunstorage.com/series-connection/

Due to the rules, you'll just have to eliminate the spaces between the w's

I was already working on a translation for you guys, but then I found the company already had an english press release which saved me the pain.

Looking forward to what you make of it.


Regards,

On the Run.