? about parallel operation of inverters

[brutus1955]

what i want to try to do, is to take two ups
inverters, tap the first at the logic drive to the FET's
driver, and jumper this logic over to a second inverter
which has had its logic disconnected.

the goal is to be able to then parallel connect the outputs of both inverters for more power

is this possible?

bob g

 

[brutus1955]

having spent a huge chunk of time reverse engineering the koyo/ge/automation direct d305 series plc so that i could insert my own choosing of micro processors, in a master and multiple slave configuration this project seems like it should be a bit easier to plot out.

i caught alot of flack over that project as well, and finally got it to working very well, and in the end learned
a lot in the process.

the more i dig into this the better the odds of it working appear to me, but we shall see

know where i can get a good flak jacket?



its been a while since i release the magic smoke or made an molten pile of scrap.

i don't have a movie cam, but have a digital camera
so if it all goes boom i can document the aftermath
and present it here for you guys entertainment.

thanks for the help

bob g

 

[OperaHouse]

Now that I see you are using really old technology with line frequency transformers, I give you a better chance of success.

 

[Noway2]

From what I can tell reading about the topology being proposed, it sounds like it should work. In effect, the goal is to parallel the FETs/IGBTs, presumably across the same DC link, and use one controller to drive both simultaneously. This is considerably less complicated than attempting to apply a closed loop control system to co-ordinate the operation of two separate inverters, coupled on the AC side.

The biggest potential problem that I see is current hogging by one set of transistors. It will be necessary to either use a type of transistor that naturally shares current when operated in parallel, such as a MOSFET, or some form of feedback will be required, such as a small resistor in the emitter path (assuming this is a problem with parallel IGBTs).

 

[brutus1955]

the oem topology already uses current limiting resistors
between the pnp driver emitter and each of the FET's on each board, so they ought to current share as designed.

doing a bit more reverse engineering it appears the 2222a
and 2907a are setup in a pushpull arrangement, so that the fet's can be driven on and driven off, but i gotta look a bit deeper. my eyes are not as good as they once were.

the base of the 2222a is driven at approx 3.4 v at no load
and increases to 4.8 v at near full load, i am assuming that
what i am seeing is a change in the pwm average voltage?

soon as i am feeling a bit better, i will drag out the scope and take a look and see if this is the case.

the unknown then becomes is the pwm drive sufficient to drive another set of 2222a's

if so it looks like i am golden in that regard, and maybe i won't smoke the logic drive.

the unit seems fairly well protected from overload, if it gets anywhere close to an overload or a short it shuts down instantly, without damage and will restart when the overload or fault is removed.

so maybe it can protect itself from my frankenstein approach

interesting project so far anyway.

so are my odds of success improving?



bob g

 

[brutus1955]

after tracking down a bunch more traces i found one that is obviously a feedback loop trace from the output of the FET bank back up to an IC that is near the microcontroller.

so..

now i am thinking rather than connect the secondaries of the output stepup transformers (secondaries) i should connect the primary side so the feedback circuit can see both outputs in tandem rather than seeing one direct and the other from the second board via induction back from its secondary.

plotting along

bob g

 

[Skogsgurra]

This thread - and Dilbert- is my favourite morning reading nowadays.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[brutus1955]

i guess i can add to "Red Green's"

"if the women don't find you handsome they oughta find you handy"

"if the engineering forum don't find your project interesting, they oughta find it entertaining"



bob g

 

[brutus1955]

update:

well i learned a bit today

you don't want to get in a hurry and messup connecting wires

i managed to heat up one bank of the final drive, likely damaging it, but it seemed to work fine after i let it cool down and corrected my boneheaded mistake.

got it to drive both output transformers no problem
got the logic jumpered across to the other driver, and that seemed to work just fine, but i had no output voltage coming out of the recepticals because the 2nd board was not powered up to close the output relays and connect the secondaries to the recepticals,, but it was working

then i decided to do something nearly as boneheaded as the misconnection in the earlier step.
i decided to removed the connecting jumper from the primaries of the two transformers.

now either that is a no/no, or i had damaged the final drive in the earlier short/overheat, but the result was
not very dramatic or interesting.

several pops and that acrid smell of magic smoke!
one bank FET's of the primary inverter took a dump, it was the one i had overheated badly. so i am not sure if the earlier short had damaged it or my disconnection of the two board outputs broke the feedback or what?

so i think i am close, i just need to go find another 10 dollar inverter, and take what i have learned and try again
this time not making the earlier mistakes.

lessons learned?
1.one logic driver apparently runs two inverter sections without blowing up.
2. don't make bad connections that result in shorts of course.
3. don't disconnect the jumpers between the output transformers? (this might have caused the failure, but not sure, however all was well till i did that)
4. a video camera would have been a waste of time, all the
fireworks are contained inside the case, so you can't see the carnage.
5. got another transformer i have been needing for another project, that has a much higher chance of success, along with a pile of other useful stuff.
6. this is too much trouble!

conclusion:
i got a better understanding of how inverters work, particularly the types used in small ups systems.
i have learned a lot about what more i need to learn
and i guess that is worth the price of admission.

thanks guys

bob g

 

[MacGyverS2000]

Ever disconnected a car's alternator from the battery while the engine is running? It does bad things to unprotected circuits...

Dan - Owner

http://www.Hi-TecDesigns.com