Issue tuning compensation tables

Hello guys,

I've been wrenching on my shitbox for 2 years now. Lots of learning, mechanical improvements and fixes.

The setup is a very common setup; 97 DSM, ECMLink v3, speed density, FP Red turbo. The intake piping has all been replaced, and I have a seriously oversized intercooler.

The car is used as a "winter beater" in Canada. We often get in the -15C degrees, and sometimes even lower.

I have 2 problems.

1. I have a hard time getting low load/low RPM area to run properly, and the fuel trims are all over the place. For example, I will start in the morning, drive to work and log the whole thing. In many cases, the fuel trims will go all the way to like -15%. I would then apply corrections and save. Leave from work to go back home, and then the corrections applied would be wrong and I would have to basically put the values I had in the morning back in there to have the correct fuel map. This changes every day.

2. During cold start and warm up, I see that the commanded AFR and the result from the wideband are very far off and it's difficult to tune properly.

My suspicion is that the problem would be in the compensation tables. Unfortunately, this ECU is fairly limited and the only real compensation tables I can change are these.

This one below also affect the commanded AFR.

At this point, I am unsure what strategy to adopt to be able to nail the compensation tables and the speed density table.

For reference, here's the speed density table, and I am highlighting the area that is specially problematic.

Any guidance is welcome.

Thanks for your time!

Hi Francis, unfortunately I haven't got direct experience on that platform so it's hard to give accurate advice. It's interesting that it provides weighting between the IAT and ECT - This is a more advanced way of dealing with charge temperature and also quite difficult to optimise the table. What I'd probably try is setting the entire table to 1.00 so the air density is calculated solely off IAT since it sounds like that is probably your biggest factor given ambient temps. The warm up enrichment on the other hand should be pretty straightforward.

Hey Andre,

Thanks for the answer. Unfortunately, the car isn't really functional without that feature. We do have quite significant temperature swings here and the fuel trims will basically max out in one or the other direction, and open-loop cold start flat out doesn't work.

I'll keep experimenting. If anyone else has tricks, I'd love to hear them!

Have a great day!

I just re-read your original post. You say you are applying the correction -- to what table? If the Main SD Fueling table is correct, you should only be changing the correction tables (so mostly the coolant compensation). Are you sure the system is running on Speed Density at that point and not the MAF? I see reference to a MAF compensation -- perhaps that's the table that needs editing...

Or perhaps the SD Temperature weighting table -- what are your air temp and coolant temp when the 15% trim is being applied -- are you sure this is in steady state condition, no transients?

Also, what is the load axis on SD Fueling table?

BTW - Andre wasn't suggesting leaving the weighting table set that way, only using it to get the SD dialed in while the engine was warming up -- during that time the weighting is most certainly almost all air temp. Once you get your fueling under control on cold days, then you can worry about the heat-soak issues the weighting table addresses.

Hello David,

Lots of questions here, I'll break down your post.

"You say you are applying the correction -- to what table?"

Both the SD and the compensation, but not at the same time. I try to make a change, drive for a couple days, see the difference and iterate. I've been working on this for 2 years at this point.

"If the Main SD Fueling table is correct, you should only be changing the correction tables (so mostly the coolant compensation)"

I didn't have a working SD fueling table when I started. I have also made quite a significant amount of mechanical changes in the last 2 years, so I am trying to dial in the SD table. The section above 3000 RPM and above 18.4 PSIA is pretty well dialed in at this point. My issue is idle (or close to idle) at low loads.

"Are you sure the system is running on Speed Density at that point and not the MAF?"

Yes, I am 100% positive about this. I took the engine out and had it rebuilt, then reviewed every sensor and replaced many (most) of them already. The reason why there are many MAF references in this ECU is because it's a chip on the factory unit, which normally operate using the MAF sensor. We can even log MAF frequency. I suspect this is just the chip translating SD table to MAF frequency then feed that to the stock ECU, which then does the fueling.

"Or perhaps the SD Temperature weighting table -- what are your air temp and coolant temp when the 15% trim is being applied -- are you sure this is in steady state condition, no transients?"

IAT temp varies between -4 and 15 depending on the driving condition (ambiant would usually be closer to -5/-10). Yes, I ignore all the transition areas here, only steady state driving, like highway for 10 minutes. Some days, that box would say -15% trim, then the next day it would be +8 trim, same box, maybe a couple degrees of ambiant difference.

"Also, what is the load axis on SD Fueling table?"

That is PSI absolute.

"BTW - Andre wasn't suggesting leaving the weighting table set that way, only using it to get the SD dialed in while the engine was warming up -- during that time the weighting is most certainly almost all air temp."

Well, ok, that makes more sense, it felt like he said to just turn it off entirely.

"Once you get your fueling under control on cold days, then you can worry about the heat-soak issues the weighting table addresses."

The problem is, this engine warms up rather quickly. Like the first 40 degrees happen within a minute which makes tuning this area very difficult. I agree this sounds like the correct way to do this, and maybe I can bypass the thermostat and run a large basket of cold water, but not sure how good of an idea that is.

I'm not trying to refuse solutions here, it's just that I have tried many of those things already :-(

I suspect your speed density temperature weight table is certainly too biased toward coolant temp at idle RPMs. At say 1000 RPM and 90c, you are going to be at about 50% weighting. So if the air temp is 10c, that means the weighted average is going to be 50c. It is very unlikely the air temp rises that much. It's hard to believe the IAT is only 15c after a turbo, but you did say "oversized intercooler" -- do you block some of the airflow in the winter?

Perhaps your air temp sensor isn't measuring the air temp correctly -- is it calibrated correctly and is it located post intercooler and before the throttle, or is it inside the manifold?

If you are trying to measure air temp before the turbo that could why you are struggling, and feel the speed density temperature weight table is so critical - it's trying to compensate for the temp sensor location. Many SD installations don't have this table and run just fine (effectively 1.0 in the table all the time), so this table shouldn't be that critical to running.

Hello David,

I have tried all sorts of values and it seems to be better when making it even more biased toward coolant. I don't know why, but I always assumed like you. For the first time, it seems I am actually making progress.

[img="blob:https://www.hpacademy.com/d5783f09-15bb-4fe0-b97c-6b55d2415655" alt="" alt="" ]

About the oversized cooler, well, it's similar to this guy. It's around 22x13x3.5. Considering I'm making maybe 350hp at the crank and this thing is rated at 1k, I'd say it's very big.

When I got the car, the IAT was located very close to the intercooler exit and I was reading close to ambiant no matter how much boost I was making as long as I was moving. I recently relocated the IAT at 6 inch from the throttle body and I am reading 10-15 degrees hotter, especially if I am sitting in traffic.

Today, I have also retarded intake cam by 4 degrees. The car seems to run a lot better, so maybe this was related.