Reducing Boost/Load

[mituc]

So (correct me if i'm wrong) I believe i'll need to lower the target loads to slightly below 2.2 to remove that overboost event and I should see a more consistent load value up until the turbo runs out.

Well, to be honest it really depends on what you're trying to do or what you want to achieve. The idea of that tune is to make the car give all it's got, but it will not be able to perform well at 100% everything. Basically that is where the parts limit performance... You can definitely have short bursts at full whatever the tune is targeting, but pretty soon heat and exhaust restriction (of the small turbo) will hit back.
By example you may get full performance in a 3rd gear pull form 4000-4500rpm to redline with the car only warm, not heated up. But if you try to do the same in 4th gear, because the engine is under mechanical and thermal stress for longer, the intake, exhaust, turbo, and the engine itself will get heatsoaked in a manner that you will not be able to reproduce in shorter pulls.
In these cases you may see significant performance differences with the same tune at the same ambient temps depending on the gear you're in, depending on how much you've stressed the car in the last few seconds/minutes and if it had time to cool down, and so on.
Think of a person trying to sprint: well trained amateurs may show sprint times for 0-20 metres similar to trained athletes battling for world records. After 30 metres things start to become clearer, while at the end of the 100m run you will see a 2+ seconds difference. Now repeat for a 200m sprint It's the same for cars... they also need oxygen + fuel and then throw those bi-products away.

Sorry yes I actually removed coolant temp on that run to free up a parameter to log as it was pretty much consistent around 80 degrees C.

The BL ECU have a pretty good logging rate, so you do not gain that much more logging resolution...
But does it also go (much) below 80C? Because if it does go in the mid/high 60s then you may get some enrichment from that, and may also mean you have a problem with your thermostat.

Interesting observation. I've looked back over all my other logs and I see the same. I just thought the intercooler is doing its job well. I'm assuming it has the stock sensors, but i'll look into it.

In the first post you said you have a TMIC... With any intercooler (TMIC or FMIC or water-to-air) is pretty hard to have the temperature almost constant during a pull with that boost simply because by the nature of things (gas law) the temperature will increase proportionally with the pressure and then the temp will drop as an effect of the thermal transfer in the TMIC.
But even if the TMIC does a good job lowering the boosted air temps while the speed increases, I'd still expect those temps to vary a bit more than just 4-5C (so at least go a bit lower).
The TMIC on a gen2/BL is pretty efficient, but in this case I'd expect also significantly lower BATs in certain situations if the peaks are only below 40, so significantly more variation than just 4-5C. So something is weird with this...

[quickdraw]

Well, to be honest it really depends on what you're trying to do or what you want to achieve. The idea of that tune is to make the car give all it's got, but it will not be able to perform well at 100% everything. Basically that is where the parts limit performance... You can definitely have short bursts at full whatever the tune is targeting, but pretty soon heat and exhaust restriction (of the small turbo) will hit back.
By example you may get full performance in a 3rd gear pull form 4000-4500rpm to redline with the car only warm, not heated up. But if you try to do the same in 4th gear, because the engine is under mechanical and thermal stress for longer, the intake, exhaust, turbo, and the engine itself will get heatsoaked in a manner that you will not be able to reproduce in shorter pulls.
In these cases you may see significant performance differences with the same tune at the same ambient temps depending on the gear you're in, depending on how much you've stressed the car in the last few seconds/minutes and if it had time to cool down, and so on.
Think of a person trying to sprint: well trained amateurs may show sprint times for 0-20 metres similar to trained athletes battling for world records. After 30 metres things start to become clearer, while at the end of the 100m run you will see a 2+ seconds difference. Now repeat for a 200m sprint It's the same for cars... they also need oxygen + fuel and then throw those bi-products away.

I actually bought the car for my son to learn to drive with and have a little bit of fun for both of us to tinker with. Car is essentially a daily driver so I would be leaning more on the 'fun but safer' side of a tune. I might be better off flashing the previous OTS tune - (Base 93/98 Octane High Load + SRI/CAI) and working up from there. I know that tune was running 14.7:1 in closed loop and Boost maxed around 18psi.

The BL ECU have a pretty good logging rate, so you do not gain that much more logging resolution...
But does it also go (much) below 80C? Because if it does go in the mid/high 60s then you may get some enrichment from that, and may also mean you have a problem with your thermostat.

I meant I was using all available parameters in the Versatuner dashboard --- I thought I was. I just had a look and didn't realise the dashboard can be expanded!

In the first post you said you have a TMIC... With any intercooler (TMIC or FMIC or water-to-air) is pretty hard to have the temperature almost constant during a pull with that boost simply because by the nature of things (gas law) the temperature will increase proportionally with the pressure and then the temp will drop as an effect of the thermal transfer in the TMIC.
But even if the TMIC does a good job lowering the boosted air temps while the speed increases, I'd still expect those temps to vary a bit more than just 4-5C (so at least go a bit lower).
The TMIC on a gen2/BL is pretty efficient, but in this case I'd expect also significantly lower BATs in certain situations if the peaks are only below 40, so significantly more variation than just 4-5C. So something is weird with this...

TMIC is aftermarket - Corksport unit https://corksport.com/mazdaspeed-3-top- ... bb8ybHpAMj. If that makes much of a difference. I'll pull the MAP today and clean it. EDIT: Update, MAP sensor was pretty dirty. Lets see what happens now.

[mituc]

I might be better off flashing the previous OTS tune - (Base 93/98 Octane High Load + SRI/CAI) and working up from there. I know that tune was running 14.7:1 in closed loop and Boost maxed around 18psi.

All the OTS tunes will target 14.7:1 in closed loop, so there is something else that the ECU did not like.

I meant I was using all available parameters in the Versatuner dashboard --- I thought I was. I just had a look and didn't realise the dashboard can be expanded!

...if you used the dashboard preset I sent you...

EDIT: Update, MAP sensor was pretty dirty. Lets see what happens now.

I'm really curious as well. Some dirt should not make that temp sensor inside the MAP to read wrong, but it really depends I guess how coated it is in that oily dirt that we get in the intake manifold...

[quickdraw]

Hi Mituc,
I made some tune changes and did another pull to 5300 in 3rd gear.

Summary of Changes to the Tune
Copy of 93US/98EU Octane Stage 2 CPE CAI + Corksport Race Pipe
- Stock MAF Calibration
- Reduced Max Boost before throttle plate reduction to 240kpa (20psi)
- Load Table - Amended from 62.5% Throttle and 2500 Down by 12% - Targeting 2.0 Peak
- Gear Load Limits - Copied Bottom Row of Load Table to each
- Revised WOT AFR target to 0.75 / 11.0
- Reduced Boost Level Base by approx 8%

Some of my observations

Boosted Temps

No real change to before.

Knock

I'm still getting knock. Up to 3 degrees (expected as I've changed nothing to reduce it except slightly lower peak boost). What is also interesting is the commanded AFR drops (richer) when knock is present. But I checked the tune and the WOT AFR With Knock actually has leaner values. Any ideas if there is another table its reading from?

Shall I decrease timing to see if it removes knock? Could it be false knock perhaps.


Boost

Boost is about where I want it (20psi). I'll adjust the base table to be closer to what it was adjusting to.

Cruise AFRs
I did a Fault Code wipe over the weekend and since then the Cruise AFRs are back to 14.7:1. So that is all good now.


Thanks

[mituc]

For anything above 19PSI I recommend an aftermarket BPV. I'm not sure if you have one, I don't think it was mentioned. For the boost you're pushing the MAF readings are a bit low most likely because the factory BPV will slightly open (by design) to prevent overboosting. So the turbo will work extra time to basically recirculate some of the air back into the manifold.

The ignition retard you're seeing may have multiple causes. I can't really pinpoint a certain cause. May even be traction related. I do not think it's too much timing, I only think it's related to the ECU not being able to hit the load targets while the boost target is hit or close to hit (after correcting down from like 200 to 216), or may also be traction related.

Exceeding the boost limit can result in ignition retard triggered by the ECU, and that is displayed in these logs as knock retard. The ignition maps of the OTS tunes follow pretty much the factory tune which has very conservative timing values, so I personally do not think it's real knock.

[quickdraw]

I do have an aftermarket BOV. Turbosmart Kompact Unit.
I could check for boost leaks

I'll lower the boost slightly and see if I still get ignition retard.

Thanks again Mituc.

[mituc]

Please try this MAF calibration and see how it works: https://www.versatune.net/forum/viewtop ... 239#p17239

Increasing the airflow values for the corresponding MAF voltage will also increase the calculated load, which means that with this MAF cal - assuming it will work as expected, the fueling may not need that much correction from the ECU. Normally in such circumstances the car would become slower (we're basically just telling the ECU we have more air compared to previous calibration, but we don't actually have it), but it may make it faster by making it work properly and the ECU not to freak out and pull timing because of various fueling corrections, overboost, and so on.
This MAF calibration is currently in use on some cars using an intake similar to yours, so because it's pretty much the same on all of them makes me believe that they're all sane, with no leaks, no dirty MAFs, and so on (the other possibility would be to be all broken in the same way, and they're just too many of them for such a large scale coincidence ).

[CrazyCanadian]

In addition I've noticed at cruising closed loop conditions the commanded and measured afrs are 14.3.1 . Any idea why this is not 14.7? The afr tables are all set to 1.0lambda at these loads.

Thanks

Something that doesn't get talked about or looked at much is a Rear O2 sensor trimming in order to maintain/hit Calculated Cat temps for emissions.. We don't have enough access to tune around that, but it'll get added in to the commanded AFRs in the back ground.

[mituc]

Are you sure this was done for that intake? This intake has a stock MAF Housing size and Cork Sport mention no re-calibration of the MAF is required.

Did you try this calibration and it was not right? Considering that this MAF calibration works on several cars that I tuned I have no reason to believe there's anything wrong with it.
Also, the differences are not that significant, please make sure you pasted it right and the cells did not shift.

[quickdraw]

Hi Mituc,

I compared the first few cells and saw up to 2000% differences so thought something was wrong. After looking at it closer I can see most are 14-15% greater across the board. But big differences in the first few. Do you know why?

i've attached an excel screenshot comparing them