Official OpenFlash Tablet FA20 Tuning Tech Discussion Thread

[ztan]

Originally Posted by Wayno

Yeah apart from the box itself and the ecu, there's nill hardware differences, so has to be something software related.

I'll switch to the MT MAF comp table and test it today.

I should also mention the same difference in behavior is at the same intake temp for both AT and MT.

I thought EL comp table was only used for closed loop operation. Data that Wayno is looking at is at WOT, therefore should be open loop, would be surprised if that is the factor here.

Wonder if this is caused by higher loads on the engine running through the torque converter at the start - you're getting erratic peak loads of >1.4 on the AT and 1.1-1.2 on the MT at the start of the pull -> ECU will be delivering fuel to hit that 1.4 target.

What are your engine load limiters A and B set at? My stock A01G limits below:

[Wayno]

Originally Posted by ztan

I thought EL comp table was only used for closed loop operation. Data that Wayno is looking at is at WOT, therefore should be open loop, would be surprised if that is the factor here.

Wonder if this is caused by higher loads on the engine running through the torque converter at the start - you're getting erratic peak loads of >1.4 on the AT and 1.1-1.2 on the MT at the start of the pull -> ECU will be delivering fuel to hit that 1.4 target.

What are your engine load limiters A and B set at? My stock A01G limits below:

There's only a few tables defined that are different between my map and MT now, none of which seem relevant.

- Base Timing Idle Below Speed Threshold
- Overrun Fueling Cut Counter RPM Threshold
- Misfire Count MAP Threshold B
- Catalyst Warm Up Idle

Mine you can get here, sans my MAF scale, with @Td-d defs:
http://www.ft86club.com/forums/attac...3&d=1433678088

The AT would never let you pile that much load below 4000 rpm in auto mode and you'd have to be an imbecile to do it on MT too. So it's sort of operating outside it's parameters which would explain being rich between 3000 and 4000.

@Td-d 's definition files don't have memory addresses for those Engine Load Limiter tables, it only has the description for the table in the base section.

My A01G def doesn't have addresses for those tables either. How'd you get them?

This is where I got the definitions.
https://github.com/TD-D/SubaruDefs/t.../RomRaider/ecu

Here's what load/AFR looks like on 98 - rich but completely normal - which could possibly be due to not hitting the load limiter.

http://datazap.me/u/wayne/stg2-uel-9...om=16328-17164


There is the global load limit of 4g/s - raising that will likely make the top end not go super rich. Doesn't account for the weird graph shape, I suspect I'd need the A and B tables for that.

[Wayno]

Ok, I've taken @Kodename47's 4 addresses for those tables, copied the hex from the A01G rom and the exact same hex appears in the A01I rom at the exact same location, so got extremely lucky there it seems.

Now the only question is - what values to use?

	<table name="Engine Load Limiter A" storageaddress="1047CC">
	 <table type="Y Axis" storageaddress="1047A8" />
	</table>
	<table name="Engine Load Limiter B" storageaddress="104804">
	 <table type="Y Axis" storageaddress="1047E0" />
	</table>

	<table type="2D" name="Engine Load Limiter A" category="Test" storagetype="uint16" endian="big" sizey="9" userlevel="3">
      <scaling units="Engine Speed (RPM)" expression="x*0.0000610351563" to_byte="x/0.0000610351563" format="0.00" fineincrement="0.1" coarseincrement="1" />
      <table type="Y Axis" name="Engine Load" storagetype="float" endian="little" logparam="P2">
        <scaling units="" expression="x" to_byte="x" format="#" fineincrement="100" coarseincrement="500" />
      </table>
      <description>Engine load limit against RPM.</description>
    </table>
	<table type="2D" name="Engine Load Limiter B" category="Test" storagetype="uint16" endian="big" sizey="9" userlevel="3">
      <scaling units="Engine Speed (RPM)" expression="x*0.0000610351563" to_byte="x/0.0000610351563" format="0.00" fineincrement="0.1" coarseincrement="1" />
      <table type="Y Axis" name="Engine Load" storagetype="float" endian="little" logparam="P2">
        <scaling units="" expression="x" to_byte="x" format="#" fineincrement="100" coarseincrement="500" />
      </table>
      <description>Engine load limit against RPM.</description>
    </table>

1047A8-1047CB
L:24(36)
44 48 00 00 44 7A 00 00 44 96 00 00 44 FA 00 00 45 09 80 00 45 16 00 00 45 1C 40 00 45 3B 80 00 45 48 00 00

1047CC-1047DF
L:14(20)
38 52 39 9A 38 F6 3A E1 3A E1 44 7B 47 0A 46 66 53 33 00 00

1047E0-104803
L:24(36)
44 48 00 00 44 7A 00 00 44 96 00 00 44 FA 00 00 45 09 80 00 45 16 00 00 45 1C 40 00 45 3B 80 00 45 48 00 00

104804-104817
L:14(20)
38 52 39 9A 38 F6 3A E1 3A E1 44 7B 47 0A 46 66 53 33 00 00

1047A8-104817
L:70(112)
44 48 00 00 44 7A 00 00 44 96 00 00 44 FA 00 00 45 09 80 00 45 16 00 00 45 1C 40 00 45 3B 80 00 45 48 00 00 38 52 39 9A 38 F6 3A E1 3A E1 44 7B 47 0A 46 66 53 33 00 00 44 48 00 00 44 7A 00 00 44 96 00 00 44 FA 00 00 45 09 80 00 45 16 00 00 45 1C 40 00 45 3B 80 00 45 48 00 00 38 52 39 9A 38 F6 3A E1 3A E1 44 7B 47 0A 46 66 53 33 00 00

[ztan]

Was just about to post those A01G defs for you. I had pulled them from the OFT .xdf tuner pro defs that Shiv released orignially (A00C).

Seems that bit of code is not used in AT if it is showing 1.4g/rev load (unless the logged valued used is before it gets cut by the engine load limiter). Which parameter address are you using to log load?

I have logs from my A01G ROM showing 1.3 g/rev (and lean) under boost!!! My engine load address = 0xFFF89148

[Wayno]

Originally Posted by ztan

Was just about to post those A01G defs for you. I had pulled them from the OFT .xdf tuner pro defs that Shiv released orignially (A00C).

Seems that bit of code is not used in AT if it is showing 1.4g/rev load (unless the logged valued used is before it gets cut by the engine load limiter). Which parameter address are you using to log load?

I have logs from my A01G ROM showing 1.3 g/rev (and lean) under boost!!! My engine load address = 0xFFF89148

I'm using OFT to log so the addresses are nicely hidden away. @steve99 would know which address on tactrix it matches, but I would guess it's the same as what you would use normally whenever you wanted to view load.

[Wayno]

If anyone doesn't have any advice, I'll try this I guess, along with returning to OTS MAF scale.

Unfortunately never logged MAF g/s so don't know what to set the MAF limit to.

[ztan]

Originally Posted by Wayno

I'm using OFT to log so the addresses are nicely hidden away. @steve99 would know which address on tactrix it matches, but I would guess it's the same as what you would use normally whenever you wanted to view load.

OFT feature request: control of logging addresses by RAM value.

There are multiple load values residing in the logic, you can pull it just after it has been calculated and before compensations, you can pull its' final value that gets used in the 3D table pulls (0xFFF89148 is used for multiple table pulls), or anywhere inbetween. Only way to know is to have control of the logging addresses and have a look at the disassembly.

[Kodename47]

Sorry I'm a little late to the party on this one. It would seem that engine loads could be the issue if you're hitting loads over the ones in the load limit tables.

What I will say is that in ECUtek the only tables that I can find that seem to differ between auto/manual:
Open Loop Throttle Thresholds. Auto has high/low altitude tables, manual only has one - these are both defined for both I and G which suggests different table locations.
Open Loop Per Gear RPM Thresholds
Ignition Timing Idle (In Gear) Above Threshold

There's nothing that should affect open loop fueling.

Have you also checked the open loop fueling additive? If you're hitting different loads then this could also be coming into play.

[brn12345]

Guys check out my post from last year around this time. I was very sure that I was hitting a load limit but although I said in that post I would change the load limit from 1.3 to 1.4, I never got around to doing it and just started enjoying my car on track. Now that the issue is brought up again I think you might find some of the things I logged and looked into last year useful. Here is the link, please spend time to read it as I am sure it will be of help:


http://www.ft86club.com/forums/showt...t=65033&page=2

[Wayno]

Does anyone know the difference between the A and B tables?

[steve99]

Originally Posted by Wayno

Does anyone know the difference between the A and B tables?

guess would be its like base timing tables A safe mode B normal

[Wayno]

I feel like I'm getting close.

http://datazap.me/u/wayne/v49-12-dyn...4112-4137-4160

- Increasing the load limit to 1.4 over 2400rpm should make the 2400-3000rpm region not lean out. Just increasing the 3200+ helps in the real world but wasn't enough for unrealistic dyno pulls.
- I'll adjust 3000-4000rpm with -8% MAF scale at 3.2V, which will make that region go a lean if I'm ever unlucky enough to hit 3.2V in open loop while not at WOT.

http://datazap.me/u/wayne/v37-1?log=...-2895-2903-398

[Wayno]

Hey mate, I went around in circles for months changing that fueling additive table, until I read Td-d's description in the definition which states it's only active when IAM < 1.

Originally Posted by Kodename47

Sorry I'm a little late to the party on this one. It would seem that engine loads could be the issue if you're hitting loads over the ones in the load limit tables.

What I will say is that in ECUtek the only tables that I can find that seem to differ between auto/manual:
Open Loop Throttle Thresholds. Auto has high/low altitude tables, manual only has one - these are both defined for both I and G which suggests different table locations.
Open Loop Per Gear RPM Thresholds
Ignition Timing Idle (In Gear) Above Threshold

There's nothing that should affect open loop fueling.

Have you also checked the open loop fueling additive? If you're hitting different loads then this could also be coming into play.

Yep, I read that after steve and ztan mentioned the load limits, but the load the AT generates seems to be in a different ballpark than MT anyway. It's obvious vishnu never had an AT with the E85 tune on a dyno.

Originally Posted by brn12345

Guys check out my post from last year around this time. I was very sure that I was hitting a load limit but although I said in that post I would change the load limit from 1.3 to 1.4, I never got around to doing it and just started enjoying my car on track. Now that the issue is brought up again I think you might find some of the things I logged and looked into last year useful. Here is the link, please spend time to read it as I am sure it will be of help:


http://www.ft86club.com/forums/showt...t=65033&page=2

[Wayno]

Here's a different AT with OTS E85 tune. OTS 11.25 O2 sensor scale. Going lean where it's over the load limit as @ztan said.

http://datazap.me/u/praneetn7/stg-2-...mark=1409-1316