merlin.oz

Have done lots of reading and searching, but have not seen anything relating to a MIVEC Temp Trim.

Have any of the dissassemblers out there spotted such a table?

It would be a nice thing, take some of the edgyness off when cold.

tephra

iTrader: (6)

as in cold start or cold weather?

merlin.oz

Just a simple temp trim, like 5 to 8 elements, with %application to the 3d map.
eg
20% at 20*C
40% at 40*C
60% at 50*C
80% at 70*C
100% at 80*C

Something like that.

mrfred

iTrader: (50)

There is such a table. Its MIVEC max vs coolant temp. I came across it a few months ago when I was looking through the MIVEC routines for a 2nd time. It adjusts the ceiling value, so for instance if you set the ceiling value to 5 deg MIVEC advance at say 15C coolant temp, then MIVEC advance will not go past 5 deg when the coolant temp is 15C. With that said, the factory setting for the table is that the ceiling value is 72 deg, so it does nothing with the factory values. I'll post up the table later today.

merlin.oz

Thanks mrfred, that should do the job nicely, if the code does in-fact look at the setting. As you have pointed out before, tables/maps exist but never get called.

I was thinking last night that a trim table would impose a fair amount of processing overhead, for little gain. This way would be better and simpler.
Provided you can tweek the temp setting.

On another note, I have not seen anyone make mention of injector end-angle timing. Do you know how Mitsi set the injector firing point?
Its a parameter I've played with on MoTeC and Autronic ECUs and was naturally curious.

mrfred

iTrader: (50)

Here is the table for an 88590015 ROM:

The table is used in the code, so it will do what the table name says.

Mattjin

88580015's map is at 6b26 for those interested. Yeah, the 72 deg values are quite odd given the limited mechanical movement.

mrfred

iTrader: (50)

Mitsu just maxed out the bit value to 255 because they never had a desire to use the table.

merlin.oz

Had a look at this today mrfred, the address you gave as 6fde for the coolant table values, is smack in the middle of the load scale on a AUDM IX.
Which corrupts the load scaling something chronic!

So, I used the "Engine Temp" scale, (at 71ca) which is 8 elements, just for safety, although I have no clue as to that being the right scaling to use for this function.

Mattjin

Try 70e6

merlin.oz

Thanks Mattjin, that works fine.
Has the same temp values as at address 71ca (the main water temp scale).
I was wondering if any other part of the code accessed 70e6, or just MIVEC.

Mattjin

Others might, it is just a temperature scale and could be used by idle control too. I just haven't spent much time pulling apart 8588, but you can maybe try Tephra.

You can use any of the addresses that have the same values, it doesnt really matter so long as you are adjusting the map correctly. Unless you plan to change the temperature conditions themselves.

GrocMax

This table doesn't do what y'all think it does, don't mess with it. It sets the zero for the cam position, its an offset in crankshaft degrees to locate 'zero'. Its 72 deg ATDC #1 when the cam sensor triggers, when cam is in zero (full retard) position.

The only time you'd want to mess with it is if you have had a BUNCH of head decking, block decking, changed to a 2.4L block, or other reason causing a difference in cam to crank timing.

Noticed ValveTiming and ValveTiming16 both use the 72 deg offset in their calcs.

mrfred

iTrader: (50)

It is really MIVEC max vs coolant temp. I think the stock 72 deg values are throwing you. The high stock values simply mean that this table will not have an effect on the MIVEC setting. If you set all the values to 0 deg, there will be no MIVEC advance.

GrocMax

Of course it won't you just told the ECU the cam '0' point is 72 crank deg further advanced than it actually is.