Logged in users > User Diaries
Score of a life time!
DanG:
That short circuit delay of 1-1/2 seconds would be weirding me out - that could be a lot of molten copper doing 1/10th lightspeed along strobing DC magnetic flux lines.
Can I tweak that to 1/4 or 1/3rd second, if the welder etc. inrush won't trigger it?
XeonPony:
Well any short of concern the fuse will deal with that. Don't want the controller trying to do it, as due to tiny tiny differences in switching speed it ends with one MOSFet trying to hold the whole current till it too switches off, usually by vaporization.
So using the Semiconductor fuse, it will be isolated in picoseconds and I'm out a 300 dollar fuse, but BMS intact and able to use full rated capacity and no magic smoke being released.
Fuse is sized so that no load will ever come close to its rating ever, other then a dead short.
Next step is install a DC circuit breaker panel, with each rated to the load driven, so that any fault will favour the CB, befor the fuse.
I really wish the BMS had a signal out that can be put on a shunt trip CB, or contactor for fault protection as that way it isn't the sacrificial component.
JW:
I hope you have scope meter or something to look at those live circuits, you can conduct without the Mosfet. I was working with a low volt 330amp mosfet array that I built.
XeonPony:
I saw some real world tests on you tube, and most had the same problem! In theory, with aggressive pull down circuit, you still have switching variances, so in the cascading shut down it can lead to only one or two carrying the full shut down flow of current.
The one unit was tested only one mosfet blew and unit remained functional.
The odds of a short is so low that fuse is the best fail safe followed by breakers to handle any normal system over load.
kitestrings:
Prety cool. Congrats.
Is the 'Dashboard' monitoring part of the package, or is that something that you developed, or adapted for another storage system?
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version