Guru
Joined: 21/06/2018
Location: AustraliaPosts: 1172
| Posted: 01:17am 06 Jun 2024 |
Have had a total rethink of the startup stragedy oops strategy.
The main problem was caused by the slow shutdown introduced by the main capacitor bank and the regulators behaviour as the capacitors slowly drain & even recover briefly with the bad effects to both Nanos.
So the proposed on/off 48V startup circuit is a 2 pronged approach that I believe will solve the problems with minimal work (card mods) involved. The new scheme stops the main capacitor bank from continuing to feed power to the control card when the on/off switch is set to off. Further the switch being set to "off" crowbars the 12V & 5V supplies to ground (be nice through a 1R-4R7 resistor) so the 5 & 12V supplies get a whoa camel moment (instantly stopped) so the control card and LCD 5V are instantly stopped.
The card mods required is to leave off Q10 R27 R28 and C28 and on the card put a link where R27 & R28 were to short the previous "Stop" card connection to ground. Of course the off position of the switch could also be connected to a convenient ground nearby, it does not have to connect to the card ground but on the card is my recommendation.
1) Change the precharge resistor/diode & Contactor-feed wiring as per start up schematic.
2) Remove (or dont install) Q10, R27, R28 & C28.
3) fit 3 diodes and 1 resistor to card - easy using breadboard area (1 cut track). Edit: Maybe rethink the DB shorting the 12V to ground is not really needed & probably overkill.
4) instead of a link fit a connector and wires to J11 for a new run stop switch.
To start the inverter it wont matter which position the run/stop switch is in.
To stop the inverter though the run/stop switch should be set to stop first and then when the green led is not flashing now turn the on/off switch to off.
If the inverter is on with the run/stop switch set to run, the auto shutdown on low battery and auto restart when battery voltage rises again is all unaffected. if the batteries get to ~7V with the on/off switch still set to on I think you have greater problems than Nano code corruption.
Control Card Mods - only 1 track to cut (Note: the 48V is now switched to ground when on/off is set to off)
Start up wiring Mods.
Auto Start Mod.pdf
Sorry Mike this should have been posted I think in my thread so I will duplicate it there
Edited 2024-06-06 11:48 by wiseguy