Xede as Boost Control

[Crusher_13]

Thanks guys.

I need to drop by my tuner and will get the dyno sheet from him by the end of the week.

[LittleredMPS]

Your AFR's are very interesting, never seen anything like it.

You are running quite lean up to the point where the boost creeps up, then it leans right out in that area, then riches right up after it.

Are you sure the guys you are using having tuned your car right, that seems very odd.

Have you added any mods since getting the car tuned?

[Crusher_13]

Your AFR's are very interesting, never seen anything like it.

You are running quite lean up to the point where the boost creeps up, then it leans right out in that area, then riches right up after it.

Are you sure the guys you are using having tuned your car right, that seems very odd.

Have you added any mods since getting the car tuned?

Thats due to the continuous issues with boost.

The actual AFR never hits 12. it peaks at something like 11.8.

The tuners have said it is all down to the boost. Are you saying I should be looking at someone else to tune my car? Leah, your opinion on this?

Then the season differences between Canberra and sydney would cause issues. The coldest day in Sydney over winter is the warmest day here...

[LittleredMPS]

I had mine tuned by Chiptorque then drove it back to Lithgow (1000m alt).

Never had any of those issues.

My AFR's were more like a sudden drop from around 12.5ish at low revs to more like 11's around where the boost came up.

I was also only running 17psi boost, but the initial spike around 2500-3500 was into the 19's.

I also noticed that my AFR ran down closer to 10.5 as I reached high revs, presumably to reduce detonation with the higher boost.

I never had that lean out that you do around the 4k rpm point though.

[Crusher_13]

I had mine tuned by Chiptorque then drove it back to Lithgow (1000m alt).

Never had any of those issues.

My AFR's were more like a sudden drop from around 12.5ish at low revs to more like 11's around where the boost came up.

I was also only running 17psi boost, but the initial spike around 2500-3500 was into the 19's.

I also noticed that my AFR ran down closer to 10.5 as I reached high revs, presumably to reduce detonation with the higher boost.

I never had that lean out that you do around the 4k rpm point though.

As much as I love the feedback I hate it!

I can just see the money flowing away!

[Crusher_13]

A further question. Our ECU's are learning ECU's.

Would a ECU reset have caused these issues?

i.e. it would have reset to its factory specific settings so the Xede would have worked on these old settings and not the learnt parameters?

Yes? No?

[2XS]

A further question. Our ECU's are learning ECU's.

Would a ECU reset have caused these issues?

i.e. it would have reset to its factory specific settings so the Xede would have worked on these old settings and not the learnt parameters?

Yes? No?

Yes and no.

The XEDE only intercepts what it sees, It has no memory of what the ECU has done previously.

I don't know which is "open" and which is "closed" loop, but, here goes.

The ECU learns in "open" loop only, i.e. when you use anything under say 80% throttle (not 100% sure on exact % of throttle), the XEDE intercepts these parameters. In "closed" loop the ECU reverts to the factory settings. With the flash tune setup, the ECU still learns as such and still reverts to the new factory tune. The XEDE will intercept all tunes.

Set the DashHawk up to read Boost, AFR, fuel pressure and LTFT (long term fuel trims) and watch when the signals change. TheLTFT will revert to 0.00 when the ECU changes from "open" to "closed" loop or vice-versa. You should be able to see what changes.

If it is boosting into 19psi (on the DashHawk)for more than a few seconds, the ECU cuts power. I have a boost gauge and DashHawk that show 4psi difference. The DashHawk will show what the ECU sees, the boost gauge shows exactly whats happening. The fuel cut show other changes.

[Crusher_13]

Xede intercepts the ECU's signal and then lies to the ECU about what is going on. Correct?

If this is done via equations based on the original map, would the equations not be wrong? Thus lean AFR's and excess boost?

when changing your map its just like unhooking the battery for the ecu (this is in relation to reflashing a car with a different AP Map). it has to start all over. i dont think you need to stay out of WOT though. thats how the car learns. the car has several levels fof fuel trims that are based on throttle position. if you dont go WOT it wont learn its fuel trims for it.

the LT trims show what the ECU has learned already. if they stay steady that means your ecu is doing a good job ! its when the short terms trims move all over the place that you have issues .

These are taken from this short thread at msprotege.com

I also found something from you 2XS. You said the car learns in "closed".

They do, yes, but, from a visual point, I think the ECU only learns to a certain degree.

I've noticed on the DashHawk that when I go past a certain part of the throttle, the signals/readings that are on the DH then change. It goes from closed loop (learning) to open loop (what has been flash tuned) and all the learning that has been done, is bypassed while the ECU only uses the base tune.

If my car went from engine mount being fit and battery disconnected to the dyno it would not of had the chance to learn how much of everything is available. Thus the ECU would have gone directly to what Mazda last flashed the car with. If Xede relies on what it receives from the ECU it would have been receiving false information. It would have recieved the message that at 100% WOT (a) air is available when infact (b) was available. Thus causing issues.

Does this work?

Leah? 2XS?

[2XS]

Xede intercepts the ECU's signal and then lies to the ECU about what is going on. Correct?

If this is done via equations based on the original map, would the equations not be wrong? Thus lean AFR's and excess boost?

True

These are taken from this short thread at msprotege.com

I also found something from you 2XS. You said the car learns in "closed".

I may have been wrong in this instance, I don't remember which way it is for sure.

If my car went from engine mount being fit and battery disconnected to the dyno it would not of had the chance to learn how much of everything is available. Thus the ECU would have gone directly to what Mazda last flashed the car with. If Xede relies on what it receives from the ECU it would have been receiving false information. It would have recieved the message that at 100% WOT (a) air is available when infact (b) was available. Thus causing issues.

Does this work?

Leah? 2XS?

The ECU always uses the base map from Mazda (or the new flash tune) to get it's own trims etc. It learns from this base tune to alter it's own "variation" into the tune. The XEDE only intercepts these signals.

The XEDE is like a person sitting on a wall. The ECU is on 1 side, the engine/signals on the other.

The person sees the engine/signals "Figure A", intercepts them and then relays them to the ECU, which inturn produces "figure A1". The same signal only slightly different.

Whether the car has had time to learn the new tune before tuning shouldn't matter one bit. It always runs from the ECU base tune.

You could try removing mods until you find the culprit. It may be a combination of 1 or 2 mods not working together. Start with the easiest mods first. If in doubt, rip it out.

---------- Post added at 05:38 AM ---------- Previous post was at 05:34 AM ----------

Maybe get your tuner to give Chiptorque a call.

[Crusher_13]

The reason I thought the issue could be the ECU reset is simple. It is the only thing that changed between 225kw and no boost issues to 211kw and boost issues.

Hmmm. I am sure it can get sorted.

Any issues with the flash and boost pressures 2XS?