![]() |
Forum Index : Electronics : Recovering from disaster by using Wiseguy's mosfet driver design
Author | Message | ||||
tinyt Guru ![]() Joined: 12/11/2017 Location: United StatesPosts: 444 |
I made 3 of these using modified chinese inverter power boards. ![]() ![]() Two are currently working at each of my two son's house, powering half of their household loads. During install, we tested succesfully start-up of their HVAC compressors. But it was not part of the household load. Last week, one of my sons connected the HVAC compressor to the inverter and BOOM! - I have now joined the elite club of mosfet destroyers. Edit: Picture was taken after loosening the mosfet mounting screws. ![]() In addition to adding a soft-start modification to the compressor wiring, I am now attempting to adapt Wiseguy's robust mosfet driver design to the chinese inverter power board. Opto Inverter.pdf I hope I will not make mistakes. Now waiting for ordered parts while I figure out how to install and connect them. Edited 2025-08-29 01:02 by tinyt |
||||
analog8484 Senior Member ![]() Joined: 11/11/2021 Location: United StatesPosts: 156 |
Sorry to see the blow up but interested to see how you modify the board. Not sure if you can do it but it would be great if you could get some scope captures of the drive wave forms before and after the mods to see if Miller shoot through is at play. One thing that has stuck in the back of my mind is the Miller current induced false turn-on's found in these egs002 power boards that I saw in a poida post a while back. I agree with him that it's likely a main culprit that degrades FET's over time and behind many of the "sudden" blow up's. I don't have proof but I suspect negative drive bias voltage on the upper and lower FET's would be more helpful than opto interlocking. |
||||
tinyt Guru ![]() Joined: 12/11/2017 Location: United StatesPosts: 444 |
I was also wondering why the upper mosfets don't have negative bias. Maybe, it will also help to add it. Edit: Here is schematic for 3 bias'ed drivers. Caution - not sanity checked. Opto Inverter 3-Bias.pdf Edited 2025-08-29 06:13 by tinyt |
||||
Godoh Guru ![]() Joined: 26/09/2020 Location: AustraliaPosts: 553 |
Just wondering TinyT, did you modify the 8010 boards on your inverters to prevent some of the problems that were built into them. Years ago Oztules showed how to modify the boards, because they had a habit of over protecting and causing blowups like your board has had. I have a few of those inverter boards running at my place. So far the only inverters that have died on me from Mosfet explosions were PowerJack inverters. After a few blowups on them I rewound the transformers and used the aliexpress boards you have used and have not had a blowup since. My 4kw shed inverter has been running now for about 3 years from memory and works great. The PowerJacks did not like voltage dips caused by starting inductive loads. good luck pete |
||||
tinyt Guru ![]() Joined: 12/11/2017 Location: United StatesPosts: 444 |
That inverter board came with an undocumented spwm board. ![]() Claimed by the seller to be compatible with EGS002 (8010 chip). So I replaced it with EGS002(modified per oztules plus some more modifications). ![]() I have 3 of those inverter boards which I have reversed. ![]() Then I modified to suite my requirements. ![]() Two of them have been running for about 6 months until one blew up. Prior to installing at each of my two son's house, they were tested with air compressor, incandescent flood lights, and momentarily with the same HVAC that it blew up with. The start up load when it failed was a 5-ton HVAC with a compressor rated at 230vac, LRA of 148A (RLA is 24A). We waited 10 minutes of no compressor power before connecting it to the inverter. The toroid is a re-wound 5KVA but together with the inverter power board was tested to be able to deliver up to 8KVA for a few minutes with forced air cooling. With a load LRA of 148A it probably experienced a load of 148 x 230 = 34 kVA. I think it is too much even for this toroidal inverter. That is why I think I also must add a soft-start module to the compressor. Edited 2025-08-29 14:11 by tinyt |
||||
Godoh Guru ![]() Joined: 26/09/2020 Location: AustraliaPosts: 553 |
It sounds like you have it under control. That starting load you are talking about is huge, I put the blowups on the Powerjack inverters down to starting loads. They failed when using power tools, especially they did not like my 1500 watt power saw. I don't know if it would help your situation but to help with dips due to starting inductive loads I added a 500 farad capacitor bank in parallel with my batteries. The capacitor bank is capable of supplying 1900 amps, (they are used for starting diesel trains.) It may be that my inverters are better than the powerjacks or it may be the capacitor banks assistance but I haven't blown an inverter up since fitting them. I am crossing my fingers. I don't have any loads like air conditioners though. I use normal shed power tools on my inverter. air compressor, power tools, wood splitter, MIG welder. So far so good. I hope the new system you are using works fine. Good luck Pete |
||||
phil99![]() Guru ![]() Joined: 11/02/2018 Location: AustraliaPosts: 2675 |
A conventional single phase induction motor / compressor combination really requires a variable frequency drive for a soft start. A conventional soft start circuit may prevent the motor developing enough torque to turn over the compressor, making the situation worse. The other option is a very beefy solid-state relay or triac to switch it on at the peak of the AC cycle. With the first half cycle reduced to a quarter cycle the chance of pushing the stator into saturation is greatly reduced. Saturation briefly takes the current well above the LRA. The normal inrush current will follow but your inverter should be able to cope with that. |
||||
![]() |
![]() |
The Back Shed's forum code is written, and hosted, in Australia. | © JAQ Software 2025 |