high power two step inverter design

Status
Not open for further replies.
M

muddasirwaheedmalik

Junior Member level 2
Joined
Jul 3, 2012
Messages
23
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Visit site
Activity points
1,535
Him I intend to design a two step (square wave ) at the moment, inverter. power target 4KVA, input 48v DC, switching module fuji BJT 500A/2.5 KW in pairs or EUPEC 11.kW IGBT.

But the transformer and current requirements are very steep for such a load, so i intend to make an intermediate stage inverter at say 50KHZ using a much smaller EI core ferrite transformer converting 48V dc to 110 or 220V . rectify it using high power high speed diode modules some filter caps and then then switch it with power module from 110v/220v dc to 220v ac.

Reason. its more complex, more components and costly, BUT, would i be saving on the 4kva transformer which would now be requiring lesser guage wires 20 AMPS on both primary and secondary at 4KW as compared to 100Amps on primary side if i were to switch from 48v Dc ????? any suggestions????
 

Dear muddasirwaheedmalik
Hi
There are some important things that i have to mention :
1-The only equipment that you can use for that power is IGBT ( because it is more reasonable for your aim ) ( high value of current ) .
2- you'd better to use some stages in parallel with master slave feedback . thus your cores will be smaller and of course cheaper . and my suggestion is H bridge !
Best Wishes
Goldsmith
 
Hi muddasirwaheedmalik,
For 220 VAC you will require 310VDC not 220VDC. if you find this tip usefull click helped me button below.
regards ani
 
Let me just clarify what you're suggesting. You want to convert 48V to >310V (as picgak points out) with an isolated high frequency SMPS. Then you will invert the 310VDC into 220VAC with some sort of PWM. Correct?

Then what you're describing is already common practice for inverter design. Your design will still have a 4KVA transformer, but it will operate at a high frequency and thus be much smaller than an equivalent iron transformer.

And for this power range, MOSFETs aren't out of the question, especially if you need very high efficiency. You should definitely use MOSFETs for the DC-DC converter part, but if you already have some IGBT modules then you can use those for the inverter.
 
Hi dear Mtwieg
efficiency. You should definitely use MOSFETs for the DC-DC converter part,
Click to expand...

When current rating is high , i don't think mosfet is a good choice , isn't it ? because it has Ron and it's dissipation can be higher than IGBT . so i think IGBT would be better .
Best Regards
Goldsmith
 

Can u please give me some schematic/idea. as i want to proceed is to make 48vdc to 310V DC, i have ferrite EI core 4KW transformer, for 50KHz switching speed and 100A ultra fast rectiifer 100A. I intend to make a pull push type inverter at this stage using IGBT.

Now when i have 310 v dc, what would be design to convert it to 220V ac without transformer. ???
 

I intend to make a pull push type inverter at this stage using IGBT.
Click to expand...
Hi again
I think a single stage push pull , converter isn't safe for this aim . why not an H bridge ?
You will have two stage first one to increase voltage rating to 311 volts and the other stage without transformer and with just an LPF in out put , and without PWM ! with SPWM . ok ?
Best Wishes
Goldsmith
 
I think you're referring to the DC-DC converter stage, which isn't an "inverter." For the DC-DC stage, a push pull could work, but isn't really optimal. A half or full bridge stage would perform better and require a slightly smaller transformer (but they are somewhat more complicated). Again, I recommend using MOSFETs for this stage.
Now when i have 310 v dc, what would be design to convert it to 220V ac without transformer. ???
Click to expand...
Typically you use a PWM full bridge to drive a lowpass filter directly.
 
H bridge would be complicated, i have PCB/kits modules for square wave generators/controllers, now i just want it to oscillate at higher frequency and using MOSFETS and using significantly lower size ferrrite core i can switch it to 1:8 step up converter to get around 320V, rectified and then use the "requested" filter design to generate modified sine wave without output transformer.????

So couple of things at hand and convenience wants me to drive to this approach instead of H bridge or half full type for which i donot hiave requisite smal signal circuitary. regards
 

Yes, push pull for the DC-DC stage will be simpler circuitry than a half bridge or full bridge, but push pull will require power components with higher ratings. Either way can work, it's just a tradeoff between complexity/performance/cost.

But for the inverter stage, you will need a full bridge.
 
As i mentioned i have the high power components for the pull push stage, EUPEC IGBT module which has a 1800A IC continous with 11.5KW Pd, at 25oC, now at 50oC i assume i wouldnot be loading the IGBTs for pull push duty for 4.8KVA , i.e. 4.8KVA/48 volt = 100Amps max , with switched the Pd will be even lower and donot assume them to run any hotter and very efficiennt, since i would be using around 50KHZ switching frequency the EI ferrite would be significantly smaller.

So far so good. ( i will just share the dimensions of EI ferrite core for windings advice) .What about full bridge design any adivce ????????? how to implement it from 310V DC ?????
 

I wouldn't be so sure about this... do you have the datasheet or part number for the IGBTs? IGBTs generally aren't too good for low voltage, medium current operation, due to their high saturation voltage. With IGBTs, the conduction loss alone will probably be around 7% of the input power. And very large IGBTs often don't work well at high switching frequencies. At 50KHz, I be the switching losses will be greater than the conduction losses. Properly selected MOSFETs would actually be much more efficient.
 

Ok , i can live with 500A, 2.5KW Misubishi Darlinton power BJT module, ID500A-030 , can i do the work for primary switching
 

Attachments

  • 500-030.pdf
    130.3 KB · Views: 257

Those have rise/fall times of 2/4us, which is pretty slow. They might work okay at 20KHz, but at 50KHz they will have very high switching losses (like hundreds of watts each).

Compare that with a small MOSFET module like this:
**broken link removed**
It would be much more efficient, and much cheaper.
 
Last edited:

Expected current draw at 48V will be about 90 amps. This will be a task.

Notice that you cannot have more than 1/2 ohm overall impedance in your initial power loop. With or without going through a transformer.

What if you can achieve 1/10 ohm overall resistance? That's a drop of 4.5 V. Power wasted as heat is 405 W. That is substantial.

Wires to carry 90 amps should be 2 or 3 or 4 awg. Several millimeters diameter.

Paralleled switching components will probably be needed, with leads thick enough to add up to 4 gauge wire.

Is the power coming from batteries? If so then is it possible to gather enough batteries, and string them in series to obtain 220 VDC from the start?

It will be an easier challenge to switch 20 A directly from that. Through an H-bridge most likely.
 

Hi muddasirwaheedmalik,
If you did not purchase the 48v battery already you can purchase low ah 12v batteries 25 nos cost wise it does not make much difference. you can avoid 48v to 310v conversion stage all together. The Dc so obtained can be given to the H Bridge (sine tabled)to directly drive the load which is very easy because you are dealing with only 12A. The battery charger is also of not much problem.
regards ani
 

I already have a 3KW twin power pack plus 50A, 48V ELTEK rectifier for charging, so irony is i have 48V bank, 3KVA charger, i just neeeeeeed inverter ???????? 4KVA but very high efficiency, Question, can i use high frequny pull push conversion for dc 320V and then use any industrial type motor controller single phase setting frequency to 60HZ and output to 220VAC, i assume they also accept DC input, i am asking becasue in a flea market here i have seen many 5KVA Variable speed/frequecny inverters at very cheap prices. just a thought ???
 

muddasirwaheedmalik said:
4KVA but very high efficiency, Question, can i use high frequny pull push conversion for dc 320V
Click to expand...
Yes, but for high frequency and high efficiency you need to use MOSFETs for the push pull.

Yes, so long as the VFD/inverter has similar ratings to what you want, then it will probably be usable. Or at the very least you can strip some useful components out of it. Personally I love tearing apart surplus motor drives in order to get the nice high power components.

To use the inverter with your DC/DC converter, you will probably need to modify the inverter quite a bit, though.
 

Hi muddasirwaheedmalik,
Ya that is a good idea of using industrial motor controller but can you set a fixed frequency in that, if possible then it is good. high freqency pull push conversion is possible efficiency is limited to your power device 2.5KW Misubishi Darlinton power BJT module, ID500A-030 how fast it can switch.
 

ok, so i decided for 2KVA inverter, with power mosfets switching at 50KHZ, is it any practicable now ?????? 48v bank input ?????? what modification i need to do with DC VFD ac motor drives????
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…