Page 2 of 2

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 15 Dec 2019, 12:48
by stumagoo
New angle on this post

I have been running my wideband now for nearly 6 months - I have not been dyno tuned as I still have some major things to sort - but when my tune goes in it starts off hitting close or spot on to all my intended specs on my fueling map. after an hour or so's driving I am running lean not a lot but just enough that high loads and crossover from closed to open loop results in detonation

My targets at this point are currently low 11's in the AFR and a fresh tune hits that nicely - over the course of the next few days (total driving normally of an hour or more at 30 minute drives) my closed loop goes from swinging 14.1 or so to 15.2 to 14 .5 to 15.5 and open loop afrs all go up .5 to.7 and for my 2.6 ton 4x4 running 8psi (remember the Pajero already has a decompressed 3.5 with around factory 9:1 comp ratio)

I have no issues at my target afr's with detonation but after the change its pinging at the worst time - mainly with the camper trailer at 100kph --- its set to drop out of closed loop at that load and it does but instead of the afr running around 12.5:1 which is hungry but no issues it sits around 13.2 which is killing it ---- understand this is running at around 2psi of boost at that point so its around 100% load ---- we are talking some heavy weights and without the charger I would be at full throttle and relying on a tail wind for acceleration (3.5 tom or more on big 4x4 mud tires)
standard 100kph cruising is perfectly fine and sits with the same engine load as normal with no charger -


I wonder if its long term trims leaning out the fuel mixtures or if I am missing a map somewhere that is pushing me back to stock fueling levels

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 15 Dec 2019, 22:16
by WytWun
stumagoo wrote:
15 Dec 2019, 12:48
I wonder if its long term trims leaning out the fuel mixtures
Long term trims definitely affect open loop mixtures! I'd be trying to get injector scaling/latency so that all 3 trims are within +/-5%. Note however that some ROMs only really use the low and mid trims (e.g. Evo 7/8) but the Magnas use all 3 - I'd expect that the V6 Paj would use all 3 too but it might not...

I'd also be not relying too much on the exact AFR numbers in the fuel map and adjust them to what is required to get your wideband reading where you want it - if the map needs 12.0 to get 12.6 on the WB that's what I'd enter (assuming your injector scaling is good!). With good injector scaling your AFR map values should be reasonably close though. You may strike some anomalies with the MAF scaling too with the turbo in the intake tract - the map named "MAF Compensation - Other" in my Magna defs is I think the place Mitsu intended these wrinkles to be smoothed out. But see how close you can get with just the fuel map first.

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 15 Dec 2019, 23:14
by stumagoo
funny you mention some only using mid and low because it seems the Pajero maps only use those 2

As for tuning to suit the wideband readings thats my point the mid range long term just keeps pulling more fuel the more I add to the map - thye current settings and current scaling (combined with the adjustable fuel reg) if I reset the ecu (battery terminals off and reset trims) both long terms are at 0 and the wide band reads pretty much what the targets are.

The issue is as I drive and the ecu learns its trims it settles on normally around + 2-3%on low and the last reading I had was - 8% on mid range which according to evoscan is the one it using when it starts to ping.

Worst part is today I topped up with fuel from a servo I now suspect to be iffy - my fuel pressure at idle has dropped 5psi and the pressure while on boost has drooped nearly 10 --- that is it is not increaseing as I open the throttle by stuff all - goes up and down about 2 psi from idle pressure.... I saw this start and get worse as I drove after filling up

I suspect I will need to swap out my fuel filter after emptying this tank and will ahve to do that before I can do any thing with the ecu

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 17 Dec 2019, 21:07
by WytWun
stumagoo wrote:
15 Dec 2019, 23:14
funny you mention some only using mid and low because it seems the Pajero maps only use those 2
Interesting, and useful to know.
stumagoo wrote:
15 Dec 2019, 23:14
As for tuning to suit the wideband readings thats my point the mid range long term just keeps pulling more fuel the more I add to the map - thye current settings and current scaling (combined with the adjustable fuel reg) if I reset the ecu (battery terminals off and reset trims) both long terms are at 0 and the wide band reads pretty much what the targets are.

The issue is as I drive and the ecu learns its trims it settles on normally around + 2-3%on low and the last reading I had was - 8% on mid range which according to evoscan is the one it using when it starts to ping.
This happens when your injector scaling is off, which could be for a variety of reasons (age, crud build-up, fuel pressure not quite right etc). With it pulling fuel like that, you need to reduce the size scaling to get the mid trim right then you might have to fiddle the latency map a bit to get the idle trim where it needs to be. Merlin's Evo Tuning Guide goes through the procedure.
stumagoo wrote:
15 Dec 2019, 23:14
Worst part is today I topped up with fuel from a servo I now suspect to be iffy - my fuel pressure at idle has dropped 5psi and the pressure while on boost has drooped nearly 10 --- that is it is not increaseing as I open the throttle by stuff all - goes up and down about 2 psi from idle pressure.... I saw this start and get worse as I drove after filling up

I suspect I will need to swap out my fuel filter after emptying this tank and will ahve to do that before I can do any thing with the ecu
I'd also think about checking your fuel pressure reg and the fuel pump (including the power supply to the pump). Agree that you need the fuel system working as it should before ECU adjustment.

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 19 Dec 2019, 20:07
by stumagoo
awesome info -I have spent a lot of time going through Merlins tuning guide - what a wealth of informations -

As for scaling -- it is removing ruel in midrange not adding it.... thats my issue if I shrink the scaling size down its going to want to remove more fuel isn't it.

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 19 Dec 2019, 22:44
by WytWun
stumagoo wrote:
19 Dec 2019, 20:07
it is removing ruel in midrange not adding it....
When I wrote "pulling fuel" I meant removing fuel as in "pulling it back" rather than releasing it into the port...
stumagoo wrote:
19 Dec 2019, 20:07
thats my issue if I shrink the scaling size down its going to want to remove more fuel isn't it.
No. It can take a bit to understand how this works because the effects are all the inverse of intuitive behaviour but reducing the fuel injector size forces the ECU to open the injector longer to achieve the same nominal amount of fuel delivered. Getting good results depends on the engine running in closed loop mode as much as possible to get solid LT fuel trims.

It's easiest to simply try it and see the results, as trying to write a short simple comprehensible explanation is beyond me at the moment :oops:

Given your mid trim is rising to 8%, I'd start by reducing injector the size by that amount. Once you've got a good size scaling (one that results in the mid LT trim staying fairly constant over a long period of driving), you then start looking at the injector latency to bring the idle LT trim close to 0%. You shouldn't need to reset the ECU as the LT trims will be adjusted fairly quickly if the engine is warm, but doing so will speed up the checking.

Keep in mind that fuel quality can make quite a difference to the trims! Given a variety of different fuels you may have tune the injector size & latency to average out the effect of fuel variation on the LT trims.

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 20 Dec 2019, 20:06
by stumagoo
And yes fuel qualtiy is huge - BP 98 consistantly reads rich the moment I fill up on it and it pulls fuel out and causes detonation

Shell 98 seems to be the best and caltex is next but is also not consistant

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 22 Dec 2019, 22:19
by WytWun
A further thought about your lean-out issue - if it's primarily under acceleration.

There is a fair number of 2D tables associated with acceleration enrichment and deceleration enleanment - the definitive discussion I'm aware of is mrfred's EvoM thread. I've found most of these tables in the Magna ROMs and they're included in my latest definitions. There's 1 or 2 I haven't found - they just don't appear to be in the code from comparing disassemblies of both Evo 8/9 and Magna ROMs.

Given that a turbo changes the accel/decel airflow dynamics compared to NA, I suspect that some tweaking of the accel enrichment related tables may be required. However you need good injector size and latency parameters to work from so get those sorted first, keeping in mind that any problems with acceleration enrichment (and/or deceleration enleanment) could be skewing the trims a bit.

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 23 Dec 2019, 10:25
by stumagoo
nope its defininetly related to mid LTFT - I am wondering if its because I dont have the afr on my map set to 14.7 in the whole cruise (closed loop) area ---- I have changed the fueling map and added fueling because the super charger is pushing the map into cells it does not see under load normally while NA eg I can see 80 to 90% load at 1000rpm so easy now its not funny.... so its very easy to go from cruise at 2500rpm on an open flat road to being under load with the same sort of air and still accelerating (like a hill with a camper trailer) I have lowered my TPS and load triggers for open loop to make sure this is not an issue and I have also made the fueling a bit richer for when this triggers

I am wondering if the fact the map is targeting say 13.9 instead of 14.7 when in closed loop that the ecu starts feeding fuel at 13.9 the o2 pulls the fuel out to 14.7 and is altering the trims to suit....

There is no specific comment on this but it does say in Merlins how to that all areas of the fuel map that come under closed loop cruise ranges should read 14.7 (pretty sure I read that at the start of my journey as well) my problem is 14.7 is not always enough when I am under load but I expect that is the enrichment tables you have mentioned. (and mentioned before I will add however my earlier issue was the mapped limits of the MAF not allowing for boost and the ECU disreguarding readings over 2 psi)

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 23 Dec 2019, 23:43
by WytWun
stumagoo wrote:
23 Dec 2019, 10:25
I am wondering if the fact the map is targeting say 13.9 instead of 14.7 when in closed loop that the ecu starts feeding fuel at 13.9 the o2 pulls the fuel out to 14.7 and is altering the trims to suit....
This...
stumagoo wrote:
23 Dec 2019, 10:25
There is no specific comment on this but it does say in Merlins how to that all areas of the fuel map that come under closed loop cruise ranges should read 14.7 (pretty sure I read that at the start of my journey as well)
... is why Merlin advises that.
stumagoo wrote:
23 Dec 2019, 10:25
my problem is 14.7 is not always enough when I am under load but I expect that is the enrichment tables you have mentioned. (and mentioned before I will add however my earlier issue was the mapped limits of the MAF not allowing for boost and the ECU disreguarding readings over 2 psi)
In addition to the accel/decel fuel tables, you probably need to adjust the open loop load & RPM threshold tables to better control which parts of the fuel map are open/closed loop (especially at lower revs I would think). In the NA ROMs, 80% load is the maximum expected from the factory setup and in the Magna ROMs at least those open loop load threshold tables have values of 95% (i.e. always closed loop when warmed up). Check your TPS for range too - mine ranges from about 12.5% at idle to about 85% at WOT (from memory!) and in the TCU code >80% TPS is always taken as WOT but I don't know for certain that ECU code is the same.

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 24 Dec 2019, 00:23
by stumagoo
I have done the tps range and thresholds and I have those triggering nicely now - I can cruise on 110 and stay in closed loop at 10 inch mercury or what ever it is and even get up to 0 just by being real gentle with the throttle but the slightest stab or push me into boost and it goes into open loop no issues

I also have a 4" LS1 TB on the engine to deal with the extra air flow required into the charger - its oversized for what I really need (could probably ahve got away with 90mm) and I have a 4" snorkel feeding it

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 24 Dec 2019, 18:44
by WytWun
stumagoo wrote:
24 Dec 2019, 00:23
I have done the tps range and thresholds and I have those triggering nicely now - I can cruise on 110 and stay in closed loop at 10 inch mercury or what ever it is and even get up to 0 just by being real gentle with the throttle but the slightest stab or push me into boost and it goes into open loop no issues
Excellent!
stumagoo wrote:
24 Dec 2019, 00:23
I also have a 4" LS1 TB on the engine to deal with the extra air flow required into the charger - its oversized for what I really need (could probably ahve got away with 90mm) and I have a 4" snorkel feeding it
:lol: 8-)

Re: H8 NL Pajero tuning - Findings . likely matches with 3.5 Magna ecu

Posted: 29 Dec 2019, 13:57
by stumagoo
yep its been a real learning curve.... but super interesting