I think you have that a bit backwards. What I intend to do is have the stored tune in flash, the values being used in ram (copied up at boot time). The tuning tool will send changes in tune to the device which will store them in memory immediately. If/when the user chooses to save them to flash, they will send a command down the serial line saying "write <this part of> your tune to flash now please" at which point it will become non volatile and flash will match ram again.Delta wrote:Most live tuned EMS's interrogate the serial port for the current values required. These values are then put into memory (if the value does not come from the serial port then the internal memory values are used, normally you can then tell it to flash the tune, which sends the whole tune to the cpu which writes it to flash.
Polling the PC for data constantly seems like a very bad idea, I don't think MS does it that way and I doubt other manufacturers do either. The ECU should be reactive to commands, not proactive. Listen for requests, not ask for information etc.
Ours will stumble too. I forget the figures, but long enough to not know where the engine is at and not be able to fire ignition for a revolution or two.This generally results in several stumbles as the cpu is writing to flash.
That's a nice idea. I'll keep that in mind.Since i believe we're going to have a permanent connection to battery with FreeEMS?? I'm assuming you could have a question when the key is turned off sent via serial to ask if the tune should be flashed....thus the car isn't running and hence no stumbles.
Fred.