Three Phase Power Supply

[vangjer]

I've been working on this project for a good one month and having only one EE mentor is quite troublesome when they're busy. I figured, the EE forum would be best to help me out with my project.

I need to design a power supply that will take in 3 phase, 460-480V, 35A and output a single phase variable 90-240V, 15A and a DC 124V.

I was thinking of doing a 3 phase AC/DC for 640V DC Bus, DC/AC for single phase output and a DC/DC for the DC output.

One problem I am running into is designing the AC/DC converter. I have selected my bridge rectifier module but am having problem with selecting a capacitor and improving the power factor. I simulate a 4000uF cap as the DC link and it gives me roughly a 3% ripple, but where can I find a 4000uF cap rated at 1200V? From my google research, choke filters improves the power factor and reduces the ripple voltage. How can I simulate a choke filter and size it?

Your help and guidance would be great!

 

[ScottyUK]

If you're struggling with the front end then I would stop trying to build one and buy commercial units instead. Pacific make some very nice AC sources.

You might also consider a variable speed drive as a source of the variable AC voltage.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

This borders to school work. And you are supposed to do that work on your own. Site rules and such.

But these are questions that quite a few engineers ask in daily work and I think it is OK to answer them. Even if it is school work.

First: It is not very often you can buy a capacitor that fits. If you look at VFDs, you find series connected capacitors of lesser voltage in VFDs with 1200 V DC link. And you also find resistors connected in parallel. Those resistors are there to make voltage distribution equal. Then, strings of capacitors are connected in parallel to get the desired capacitance. Remember that capacitor tolerances are in the two figure range. Not uncommon to see -10/+50 percent.

Six pulse rectifiers have a quite low ripple. Are you going to use a six pulse or three pulse rectifier? Remember that a six pulse rectifier cannot be referenced to ground if you do not use a transformer. A three pulse rectifier can, but gives more ripple. There are also rules how much DC you can take out of the mains (which is what you do when using three pulses) in some countries.

Power factor is bad if you do not use a mains reactor. And not very good if you use one either. There are so called active front ends that force the PSU to draw sine current from the grid. They are also known as PCF rectifiers and are mandatory in som countries and in certain power ranges.

Lots of things to consider. What will this power supply be used for?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Skogsgurra]

Scotty already condensed most of what I said.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[vangjer]

Thanks for your input,

This project is at work and there's only one EE in our facility. The cap itself, seems ridiculously huge and expensive. I found a manufacturer for a 1000uF rated at 1kV for $350 (min. purchase is 25) and I only needed 4.

I considered buying things off the shelf and put it together, but I can't seem to find one that works around my specification via the right input and output.

I guess this project might be a no go, our head EE in our area suggested that I look for a vendor to design this project because its beyond the scope of what I know and outside industrial projects.

 

[Skogsgurra]

I just noticed your thread240-304706
It seems you had better get a consultant involved.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.