98/99 Ecu tuning problems

[knochgoon24]

I'm having quite a perplexing issue right now.

I just switched from the Evo 8 ecu over to a 98/99 ecu. I've reverted all the wiring back, put the 2g maf back on, and set it all back up for tuning on the 98/99 ecu.

I'm using CeddyMods 1.9.


So right now, the only thing that I think should be affecting the tune is that I have PTE 680cc injectors.

So I used:
Injector scaling - 680cc
Injector latency -
3.825
2.67
1.725
1.26
1.05
0.735
0.48

With that, the car was so lean that it maxed out all the fuel trims...

So I changed the injector scaling to 631cc. Still maxed out all the fuel trims.

Next, I changed the Injector latency to -
3.825
2.13
1.44
1.035
0.690
0.57
0.48

And with this, I'm still maxing out my fuel trims.


So I turned to adjusting my MAF scaling. Added 10% to every value.
This finally started making some progress.

All my LTFT have once again maxed out (I did reset them), but my STFT has dropped down near 0, finally...

So my question is, why have I had to add so much extra fuel? This doesn't seem right. The car runs great otherwise.

-I know the maf is still good because my friend just used it.
-It's not a boost leak or else I'd be pulling fuel, not adding.
-I'm now hitting 180 load at only 10.5 psi. Is this normal for the stock maf? It's been so long since I used one.


Something just doesn't seem right.

 

[Absit]

When you went back to the H8 did you use the MAF scaling that came with Ceddymods or your old scaling? Also remember DSMInjectorScaling and InjectorScaling are different formulas.

 

[knochgoon24]

I did use the MAF scaling that came with Ceddy mods. It's the same as the stock scaling.

It's using InjectorScalingDSM. So 680 should have been about right. I set it to 634 because it was about 7% smaller. I was hoping that would be enough for me to start seeing a difference and it wasn't.

I didn't touch anything with the fuel system, and it all worked fine before the swap.

I have another MAF here. Maybe I'll try that and see if it makes any difference.

 

[Absit]

Searching around I found this for the PTE680s:

Scale
650
Latency
3.312
2.184
1.416
0.960
0.696
0.480
0.360

AND

Scale
622
Latency
3.624
1.992
1.344
1.008
0.768
0.600
0.456

But Ceddy actually suggests 640 as the scaling for those, can you post a log? Not saying you'd miss something simple, but more sets of eyes is always good.

 

[knochgoon24]

Thanks for those scalings/latencies.

I just realized that I adjusted the latencies the wrong way. I went to a smaller latency when I should have gone bigger. I'll fix that again.

I could get a log, but I'm not sure how much that would really help. I'll try and get one though.

-The o2 sensor are cycling properly.
-Coolant and air temps are both reading correctly.
-Battery voltage is good.
-MAF is giving me consistent readings.


Now that it's light out, I'm going to double check that I don't have a post-MAF / pre-turbo intake leak. That's the only place I could have a leak that's causing these problems.

It ran fine with the Evo 8 ecu just 2 days ago.



Just double checking, but with Ceddymod, non-SD, the IAT reading needs to be from where? Right now, it's from the MAF.


I'd also like to point out, the table Open Loop 2 is the load where the engine goes into open loop. Open Loop 1 is the load where the engine goes back into closed loop. At least from what I've been seeing. I'd take a guess the the Open Loop 1&2 tables are the same way.


----------------------------------

EDIT/UPDATE:

No leaks in the intake pre-turbo.
I'm finally making some progress, but I have to use an injector size of 597!?! This lets me use injector scalings in the range that I've seen for other people with these injectors.

UPDATE #2:
Still using an injector scaling of 597.
Got the fuel trims within a normal range using
3.735
2.34
1.545
1.2
0.945
0.705
0.495

When I used
3.735
2.34
1.545
1.14
0.915
0.675
0.465
it made it still a hair too rich.

I bet
3.735
2.34
1.545
1.18
0.93
0.69
0.485
would be pretty close, but the laptop needs to recharge first before I can try some more.

However, I think that I may need to drop the injector scaling even further and readjust the injector scaling, because I'm now starting to go leaner under boost. It's almost looking like I need to go with the suggested Evo scaling of 552 and then bump the injector latencies back down.


I don't know if any of you have read Merlins EcuFlash Tuing Guide (focuses more on Evos) but I'm going to try his "Injector Tuning Method #2" if my trims are still off.

 

[Ceddy]

PTE 680s only flow 640 at 43.5 fuel pressure.

• PTE is famous for overrating there injectors.

Some injector numbers here -> http://ceddy.us/tuning/injectors/

I would start with 634 Inj Scaling. Then try 621.

• Using - InjectorScalingDSM

Latency is usually +0.288 - 0.300

+0.300 is:

• 3.825
• 2.115
• 1.425
• 1.095
• 0.885
• 0.735
• 0.615

With the H8s large injectors do tend to be scaled a little smaller, but you don't really see this until you get to 800s or 1000s.

If you need to rescale the MAF use the MAF Smoothing table, and only adjust up to around 150 Hz. If you rescale high Hz values you are fighting with Injector Scaling.


I've had a stock factory MAF that needed a lot added on the MAF Smoothing.
Around +24% at low Hz gradually tapering back to the stock values around 150 Hz.

 

[knochgoon24]

So what to adjustment to the MAF smoothing do?

 

[Ceddy]

So what to adjustment to the MAF smoothing do?

MAF Smoothing adds or subtracts fuel at different Hz levels.

It has a nice % scaling.

So 110% = 10% more fuel.



I prefer to use MAF Smoothing to adjust fueling.

And use MAF Scaling for swapping to a Evo or other MAF.

 

[knochgoon24]

Ah. Ok. Makes sense. Thanks.

I just put a new tank of gas in the car and wired up my DC-AC power inverter. I'm all set for a several more hours of tuning.


EDIT: 3 hours of tuning has me thinking that something is really messed up.

Using the injector latencies you listed above, my injector scaling needs to be really low (like 465) for my AFRs to line up with my target AFR. This isn't right.

This is where I'm at now. It just doesn't make sense.
I need to pull 20% fuel at lower Hz values and add 20% at higher ones.

You must be logged in to view this image or video.

That is finally starting to bring everything back in line.
-AFR matches the map.
-Low/Mid/High LTFT are reasonable. (Using the trim control points in Ceddymods)

I didn't have this trouble at all with SD on the Evo 8 ecu. Actually, I had to scale my injectors a little larger than most of the Evo guys had to.

And yes, I'm using both the HEX and the XML files from the Ceddymods folder.

 

[knochgoon24]

Well the final MAF Smoothing table I ened up with isn't quite as severe at the top one, but it's the same general shape. I'm pulling 10% from 0-200Hz, then it decreases to 0% at 400Hz, but then I have to gradually add fuel up 1200Hz, where I'm adding 20%.

Do I need a smaller injector scaling still? And less injector latency?

Anyway, I'm able to run 20psi right now and all the AFR are in check. Low and Mid LTFT are +/- 2%. High LTFT needs a little more playing with. Sits at +8 to +11%.

I'm at 220 load now at 20psi. I need to move up to the Big Maps so I can continue upping the boost with the same map resolution.

 

[Ceddy]

Its strange that you need so much MAF compensation. Usually you just need to enter Injector Scaling and Latency and you are good to go, MAF Smoothing is more just for fine tuning.


http://ceddy.us/tuning/injectors/
Is back up with pics. But I need to update some things, mostly the Fuel Trim Control Points.

Stock 2G DSM ROMs, only have Low and Mid Trims. And Mid would effect WOT fueling.

CeddyMods has Evo8/9 style fuel trims(Low, Mid, & High) And High has no effect on fueling. This gives consistent WOT fueling.



If your ROM is giving you consistent results, I would stick with it. The only suggestions I could give for the future is the Trims have enough time to adjust(16 mins). Reset ecu between changes. And if trims are giving you a run around you could try Full Time Open Loop, (Perphery0, bit 4).


Is it possible that your MAF could have a issue? Like K&N Filter Oil contamination. Or a really weird intake pipe?

 

[knochgoon24]

I drove for 20-45 min before switching up the tunes. I looked at the logs and I could see that the trims had adjusted and settled down.

The MAS is on a stock 2g intake pipe with an Apexi filter and adapter (old skool).

Cold start today was almost dead on. I was really surprised.

 

[knochgoon24]

I think I know why my MAF smoothing was so far off. I completely forgot that I was still running the vented catch can. I was fine doing this with the SD setup I had.

So under vacuum, I was sucking in extra unmetered air. I wondered why my cruise fuel trim bounced around so much. And under decel with very very light throttle, it would go so rich that it wouldn't get back to stoic with the fuel trims.

It all makes sense now.

 

[knochgoon24]

Well now I've started trying to tune the SD stuff. I don't have a single cell above 100% VE. I'm guessing that's because my MAF was all messed up.

Should I just tune SD without using the MAF? I've done it before. It might not be as easy, but it's pretty straight forward.

I just want to avoid some of the cold start issues I was having with my SD tune last time.

 

[EVLGSX]

This is a good one