 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Kent’s 10KW Inverter
  Page 3 of 4   |
|||||
| Author | Message | ||||
| BenandAmber Guru  Joined: 16/02/2019 Location: United StatesPosts: 961 |
Make it is exactly how you want it and then epoxy it and it might last you the rest of your life Warp speed will tell you exactly how to make it the perfect toroid I can only look dream and drool Edited 2019-08-14 12:07 by BenandAmber be warned i am good parrot but Dumber than a box of rocks |
||||
| tinyt Guru Joined: 12/11/2017 Location: United StatesPosts: 431 |
Looks like the 240v winding is only one layer (2 in hand). You might not have to un-wind it. It is easy to split it into two 120V windings by looking for the midpoint of the existing winding. Somewhere here is how I did mine (Posted: 03:55am 04 Jun 2018). Then you can just add to each of the two separated windings more turns to lower the flux density. Edited 2019-08-14 12:53 by tinyt |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Next step, measure the exact dimensions of the bare core now that you can actually see it, and count the turns that are already on there, and measure the wire diameter. Once we know exactly what is there, we can decide what to do. Cheers, Tony. |
||||
| kentfielddude Regular Member  Joined: 09/05/2019 Location: United StatesPosts: 89 |
OD - 280mm ID - 140mm Height -120mm Turns: 66 Wire size - Two 4.6mm diameter wire Edited 2019-08-18 08:45 by kentfielddude |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Thanks Kent, it was the 120mm height I was missing. O/k, going to our handy flux calculator: https://daycounter.com/Calculators/Max-Flux-Density-Calculator.phtml For the existing secondary we have now, enter the following: Voltage 240 Frequency 0.0006 Mhz Turns 66 Core area 108cm sq (90mm x 120mm) Flux = 1,264 gauss, (or 1.264 Teslas in metric) We really need to reduce the flux density to 1.0 Teslas or less, to reduce the idling current and the initial inrush current. Not surprisingly, the primary and secondary are wound from the same 4.6mm diameter wire, and we know that the original secondary has wire length enough for 25 turns. Now suppose we add 25 turns to our existing 66 turn secondary, making 91 turns. Voltage 240 Frequency 0.00006 Mhz Turns 91 Core 108cm sq Flux = 0.916 Teslas. That should work out pretty well, but we can test it first and see how well. Add 25 turns to the existing winding using any convenient thin plastic insulated wire that you have there. Connect that up to the 220v supply, and measure the idling current. We have reduced the flux density by 66/91 or to to about 73% of what it was. But the idling current will fall by a much greater amount than that, to probably less than 50% of the original 100 or so watts of idling power. Try it and see  Cheers, Tony. |
||||
| kentfielddude Regular Member Joined: 09/05/2019 Location: United StatesPosts: 89 |
Will this be suitable to use for testing? I don't have a 220v source readily available Edited 2019-08-18 10:23 by kentfielddude |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Yes, that will be fine. I actually have one of those transformes myself. Cheers, Tony. |
||||
| tinyt Guru Joined: 12/11/2017 Location: United StatesPosts: 431 |
Almost all residential wirings here have 220VAC (black wires) at the power service entrance. Between the two black wires is 220VAC. Inside the circuit breaker box, there are neutral wires and between these neutral wires and each of the black wires is 110VAC. The neutral is usually white and the bare wires are also at neutral. At the circuit breaker box, they connect the 220V loads (AC, Electric Range) to 220VAC. They split the regular 110VAC outlets and 110VAC lighting between the two 220VAC black wires and neutral. The 110VAC outlet will have one of its terminal connected to a white wire and the other to a black wire(live, if wired to code the shorter slot of the socket). Because of the 110VAC loading split, it is possible that one 110VAC outlet will be served by one of the black wire and another 110VAC outlet served by the other black wire. So if you have some insulated wires and a multimeter, you can look for this pair of outlets and have 220VAC for your testing, no need for a doubler transformer. I actually, installed a 220VAC outlet in one of our rooms using this method. Edited 2019-08-18 11:52 by tinyt |
||||
| kentfielddude Regular Member Joined: 09/05/2019 Location: United StatesPosts: 89 |
With the current 66 winding it uses 83 watts at 233V |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Now try adding various numbers of turns (of thin wire) to the existing 66 turns, and see how much that reduces the idling current. Cheers, Tony. |
||||
| kentfielddude Regular Member Joined: 09/05/2019 Location: United StatesPosts: 89 |
With an additional 21-22ish messy windings it draws 41 watts. |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Keep going with more turns in steps of about twenty. Messy does not matter, one pass through the hole equals one full turn, no matter where it is in the hole. Cheers, Tony. |
||||
| kentfielddude Regular Member Joined: 09/05/2019 Location: United StatesPosts: 89 |
42 turns: 26 watts 60ish turns: 19 watts I don't have enough wire to really do any more turns. Edited 2019-08-24 14:57 by kentfielddude |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Great stuff there Kent that is enough. There is method in my madness  Next question, what will be the lowest dc input voltage your Warpverter will ever see? This is the sixty four dollar question, because it involves working out the exact turns ratios of the four transformers so that all the final very small steps will end up being exactly even. For 10Kw, I would suggest around a 100v nominal dc input voltage, as that will be around or just over 100 amps at 10Kw, and enable IGBTs to be used. If you wish to do it at 48v nominal, its going to be well over 200 amps, and the number of mosfets, and the size of the heatsinks will be pretty enormous. Its possible, but I would not choose to do it that way myself if given the choice. The actual minimum design input voltage will have a lot to do with the battery type, and chemistry, and the number of cells. So give all this some very serious consideration before going any further, and once you come up with a final decision as to the minimum design dc input voltage of your Warpverter, we can then look at the detail design of the big number #1 transformer for split phase operation as regards primary and secondary turns, exact turns ratio, and probable final idling current. Cheers, Tony. |
||||
mackoffgrid Guru Joined: 13/03/2017 Location: AustraliaPosts: 460 |
KFD, Good going there, I've not sure if any of the Ozinverter guys have done that test on the "3kW" cores. Tony, Further to what you've been voicing for quite some time. I've been wondering of late whether flux density, or Volts per turn relative to the core size, has been playing a much bigger role than people think in the unexplained smoke releases of SPWM inverters. If perhaps they were to decrease their flux density or Volts per turn by 20% or more (50% in KFD's case) they might see a significant change in reliability? I can't help thinking the benefits would flow onto choke design, and low load stability - where all the problems seem to lie. That core would make a great transformer for a WarpInverter Cheers Andrew |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Kent and I have been contacting each other via PM's. Kent is now thinking seriously in terms of building a Warpverter instead of SPWM. The 10Kw transformer design we end up with will work equally well for both, the only difference being the number of primary turns required. And that can be easily changed later if need be. So the same core and secondary combination will work with either a Warpverter or SPWM. Increasing the secondary turns will really pay off in reduced idling and inrush. But there is a limit to how many extra turns will fit, before the hole shrinks down too small for the primary. So it all needs to be looked at and thought through, and once the various factors are known, some kind of compromise design reached. As far as lowering the design flux density goes, there are two enormous advantages for inverter operation. The idling power is much reduced, and the often violent inrush current at initial turn on will also be greatly diminished. The advantages of a lower idling power are pretty obvious. But the start up inrush effect can create problems that are not entirely solved with electronic soft starting. If there is some momentary interruption to the drive, say a software crash, or momentary dc brownout, the soft start mechanism may fail to kick in as it should with a normal full cold startup, leading to a mosfet killing current surge. Oh the inverter design issue is a whole jungle of interrelated problems to sort through. There are so many possible causes of blow ups that fixing a few or many of them may not produce any apparent improvement in reliability. The only real solution to all this, is a complete rethink which is what I have attempted to do with the Warpverter. A great many of the high frequency SPWM problems just do not exist with low frequency switching. And the few remaining potential problems can be avoided by very careful design of the magnetics and the gate driver circuitry. Anyhow, about split phase operation. Our friends across the Pacific have a very strange and unique power distribution system, where there are two 110/115v phases and a neutral. The usual single phase Warpverter uses four transformers, each having a secondary voltage that goes up in voltage ratios of 1 : 3 : 9 : 27. For split phase operation, we still have the same four transformers, but each transformer will have two independent secondary winding of half the voltage. There will be two identical independent sine wave outputs produced, which can then be connected in series, with the centre point being the neutral. The secondary voltages could be 4v, 12v, 36v, and 108v, each transformer having a pair of secondaries of those voltages. When added together we get 160v peak, and with that we can create a sine wave of 113.136v rms. It would also be possible with this system to connect both 110/115v outputs in parallel instead of series. Cheers, Tony. |
||||
| mackoffgrid Guru Joined: 13/03/2017 Location: AustraliaPosts: 460 |
 I think I have my head around the split phase thing, complicates it a bit, but it's just wire Cheers Andrew |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Its just a case of splitting each secondary into two separate windings each with half the usual number of turns. Then connecting up two series strings of secondaries, instead of one series string. Cheers, Tony. |
||||
| kentfielddude Regular Member Joined: 09/05/2019 Location: United StatesPosts: 89 |
I think 100V is a good idea but I haven't been able to find any mppt charge controllers that support high voltage batteries besides the bluesky ones but I'm not a big fan of the Chinese low quality items. |
||||
| kentfielddude Regular Member Joined: 09/05/2019 Location: United StatesPosts: 89 |
Anyone know of any high voltage mppt charge controllers? |
||||
| Page 3 of 4 |
|||||
 Print this page |
