Guru
Joined: 13/10/2014
Location: AustraliaPosts: 1845
| Posted: 12:49am 31 Dec 2022 |
First off Mike, thank you for taking the time to reply whilst on holidays, I know that feeling of missing the project calling you back to the work bench.
Actually, I found your build logical and easy to follow, I could see exactly where you were coming from, same as Poida, but it's like the last few pages of two detective Novels are missing - I'm looking for the Happy ever after ending
I look forward to your investigation into that dam Choke, it appears to be one of the last areas that has a "luck factor" about it. The brilliant in depth testing by yourself and Poida into failure modes is something I am passionate about, as are a few others now. Leaving aside the build quality and mistakes made in layout and construction, to me as least, it's looking like that area of interaction between the construction and electrical "stiffness" of the transformer [tiring to think of the correct term for that] and the Choke in the complex interplay of restive, inductive and capacitance effects in Home AC wiring, and everything between the transformer secondary and the various connected idle and running loads on each build, is the Gray area that makes or breaks the same inverter builds, why some work perfectly and others go bang?
I was also planning to test a choke using 4 of the E64 cores (gaped). I'm just building a few Jigs for testing FET voltage, current, heat and RDS limits and then matching as close as I can from the 80 odd FETS that I have here, and a Jig for the Choke inductance.
I came across Gaspos little board that plugged into the OzInverter board and apparently solved the shoot through issue, I was going to see if any were available, however I will go over your posts again (all downloaded in PDF) and look at the inverse opto drive. I would love to take up your offer and purchase the blank boards for the nano from you. I was just going to order a couple of Nanos to play with, so dam impressed with Poidas code and yours, Madness and Warpspeeds input on that - love the group think posts, apology if I left anyone out.
I'm a bit like Murphy's friend also love the construction side as well, so I'm not trying to make something compact. This will be a heavy unit, and I normally do a Cabinet on wheels to allow full access like Murphy's friend did.
I've always found that the easier something is to fix, the less likely it is to fail.
I have the OzInverter boards and will likely use the power board, but change the screw caps to the bottom of the board, use bigger heat sinks and a different connection setup to the control board. I can modify almost anything without electrically compromising the operation of the inverter, even under the highest load conditions (including noise - RF and induced).
I fully appreciate why the OzInverter board was made this way, I think it is brilliant for what it set out to accomplish. I, on the other hand, would like to have four separate power boards as per Murphy's last build like the WarpInverter construction, but like I said, getting boards made at the moment is difficult for me.
This is the reason I want to get the Transformer right, then construct the inverter in a such a fashion that allows me to easily change the driver and power board layout if needed.
Everything will be tested in stages then run up with NO caps, current limited supply and lots of Oscilloscope (digital and analogue) testing. That's also why I'm trying to convert the 24V inverter to 48v that I posted above. Get the transformer and choke tested and adjusted running at high power and blow it up first (hopefully not).
Mike.