Initial version for OL zero tps patch.
Different version for manual and auto.
Forced Open loop with external switch on pin B23. ground needs to be applied on this pin through light diode and resistor, for visual confirmation.
Also MAP and MPH check to enable.
Initial version so use at your own risk.
Feedback needed.
awesome, i'll test it (except the auto version, i can't test)Originally Posted by kur4o
i tested your manual OL IDLE patch today; with the patch the car refused to go closed loop at all
That`s good to know. Most likely something jumps not as expected. Bad news is, I don`t have access to running vehicle now and debugging will be very hard.
I`ve been thinking to take new approach. Cell 16 is zero tps, zero mph, cell 17 is zero tps, very low map, cell 18 is zero mph, tps>0.
Do you want open loop only in cell 16 and keep closed loop in other cells. I want to try to duplicate conditions for OL with the conditions of entering this cells.
I`ve switched the settings for forced open loop commanded by PIN b23. (in open state I thought it was 0, but actually it`s 1)
Now is fixed I hope the other part works as expected.
i know that story... im afraid to experiment too much with my car since its my only car! but i'll try again soon
that would be perfect...Do you want open loop only in cell 16 and keep closed loop in other cells.
Added a new patch. It duplicates conditions for entering cell 16. If all conditions met it goes to open loop.
This patch depends on CCP functionality in the part of selecting blm cells.
If, BLM min ccpDC% enable byte_2626 is changed it might not work as expected.
didn't realize but mode3 can return multiple memory bytes in a single request just by stringing them together:
since aldl request length is limited to the single length byte (0xFF - 0x52 = dec173) minus the checksum/mode/length/device bytes.Code:F4 5C 03 01 EE 01 ED 01 00 CF > F4 59 03 00 00 87 29
so 169 message bytes, and since they're 16 bit addresses, that's 84 bytes of memory in a single mode4 request.
it's almost as good as a mode1 request, but the request itself is way longer which slows stuff down to about 3-4ms per byte requested.
i have tested this by grabbing the entire BLM table memory from both banks in a single request, and it works.
one gigantic request but still a magnitude faster than multiple mode3 requests...
Code:f4a203001c001d001e001f0020002100220023002400250026002700280029002a002b002c002d002e002f0030003100320033003400350036003700380039003a003b003c003d003e003f0040004180 > f47c0380808080808080808080808080808080808080808080808080808080808080808080808080808d
... mode 2 dumps 64 bytes of data doesn't it?
Subd
Posting Permissions
Reply With Quote
Bookmarks