DIY tuning

[RealM135i]

Hi guys

With regards to the P-I-D tables, if my Boost Target (Eg. 23psi) - Boost Actual (Eg. 21psi) = 2psi boost deviation, would it be a positive or negative deviation?

And then with this deviation, would i be looking at the negative or positive axis in the PID tables?

 

[wheela]

Hi guys

With regards to the P-I-D tables, if my Boost Target (Eg. 23psi) - Boost Actual (Eg. 21psi) = 2psi boost deviation, would it be a positive or negative deviation?

And then with this deviation, would i be looking at the negative or positive axis in the PID tables?

View attachment 59397

Let me start by saying that I don't know! But as an engineer, I would assume that your scenario would be considered a negative deviation because your're adding duty cycle to get to up to target, but still negative with respect to target (your deviation to target is negative). In sure others can confirm/deny. I'm still running an OTS tune, but am starting to poke around in my stock bin and work up the courage to start self tuning.

For me, this thread along with Anjuna's over on the other forum have been phenomenally inspiring, and I'm grateful to all who've contributed!

 

[P33P33]

The boost error is calculated as "boost target-Boost measured in chargepipe". Pldr_soll being boost target and Pvdkds boost actual in chargepipe.

 

[KRussell]

Would anyone confirm for me MHD handles the checksum? (Siemens MSD80) I assume it does unless you are all using an external check sum corrector?

 

[houtan]

Anyone here self tune and run mafless? One of my next projects is a hot side short intake, and running without the maf will make the design much easier. I have tried to disable the maf and my logs look great, but I still get maf related CEL. Anyone know what tables/changes to make to run mafless? I am 98G0B if that helps.

 

[Phack79]

Hi Members!

Does anybody has idea why got additional 5-6% WGDC in 6th gear compared 4th and 5th gear. it generates huge overboost.

the car is M135ix EWG with hybrid turbo.

datazap.me | Phack79 | MHD 1-es slot 1.8bar_v8_7

datazap.me

 

[JPuehl]

Hello fellow DIYer's

I've come across an issue where it seems I have a 28.5# boost limit and I'm wondering how to get past it. I have a 2011 135i with N55E, 98G0B DME. I recently upgraded from a PS2 to a PS750 and am trying to tune to 28# but of course sometimes hit 28.5# and when that happens the DME freaks out, drops load actual a lot which causes the clutches to slip because a much lower torque is sent to the trans. The throttle doesn't close because my boost ceiling is set to 31#. I noticed that in post 252 Steven was able to hold 29# boost without the drama. Does anyone have any idea what could be going on? I know that @mwpa71 (Mark) and @houtan is having the same issue so I know it's not just me, lol.

One thing that was suggested to me was to try raising the max voltage the MAP sensor is seeing from 4.8486 to 5.1, which didn't seem to help.

This is an example log of my issue: https://www.spoolstreet.com/graphs/log-02-27-2024-18-17-11.34176/

Here's a log from @houtan showing the issue: https://www.spoolstreet.com/graphs/log-03-01-2024-21-27-01.34237/

Here's a log from Mark (@mwpa71): https://www.spoolstreet.com/graphs/mark-p750-v2-1-mafless-4th-gear.34232/

Update 3/2/2024:

I seen in a log that the magic number is ~28.5, when boost goes over that number the load actual, MAF actual, etc. drops, P-factor goes positive even tho boost is above target. Reported torque drops because load actual drops causing the clutches in the DCT to slip. AFR also goes lean. I am currently MAFless but I hacked together a MAF'd intake using the BMS intake I used on the PS2 before going MAFless to test with and have the same results. When boost drops below 28.5 things start returning to normal. The drama happens at 28.5 regardless of RPM, I've had it happen during spool or during post-shift over-boost. I do my own tuning and am helping Mark so we both basically rock the same tune (BIN). Houtan does his on tunes but has a different BIN his tune is based on, I also transferred all my changes to a "new" STOCK BIN and had the same issue eliminating the cause being a corrupt BIN. My estimation based on looking at logs is the issue is >=28.5# but it could actually be 28.4 or 28.8.

 

[Stokes]

Pinging others for help...

@Jake@MHD
@V8bait,
@carabuser,
@Sbrach

 

[wheela]

Hello fellow DIYer's

I've come across an issue where it seems I have a 28.5# boost limit and I'm wondering how to get past it. I have a 2011 135i with N55E, 98G0B DME. I recently upgraded from a PS2 to a PS750 and am trying to tune to 28# but of course sometimes hit 28.5# and when that happens the DME freaks out, drops load actual a lot which causes the clutches to slip because a much lower torque is sent to the trans. The throttle doesn't close because my boost ceiling is set to 31#. I noticed that in post 252 Steven was able to hold 29# boost without the drama. Does anyone have any idea what could be going on? I know that @mwpa71 (Mark) and @houtan is having the same issue so I know it's not just me, lol.

One thing that was suggested to me was to try raising the max voltage the MAP sensor is seeing from 4.8486 to 5.1, which didn't seem to help.

This is an example log of my issue: https://www.spoolstreet.com/graphs/log-02-27-2024-18-17-11.34176/

Here's a log from @houtan showing the issue: https://www.spoolstreet.com/graphs/log-03-01-2024-21-27-01.34237/

Here's a log from Mark (@mwpa71): https://www.spoolstreet.com/graphs/mark-p750-v2-1-mafless-4th-gear.34232/

When you hit 28.5# is it load actual or load request that's dropping? I'm confused how load actual can drop unless either the throttle plate closes or boost drops?. Perhaps your valvetronic could be closing down to reduce load actual, independant of your throttle plate staying open?

If the clutches start slipping, it almost seems like the DME starts reporting an incorrect load actual, so the torque reported by the model ends up much less than reality?

Edit: Okay, I just looked at a log. That's weird, it's like whatever part of the DME that converts airflow to load actual falls off a cliff and reverts to to some small value

 

[JPuehl]

When you hit 28.5# is it load actual or load request that's dropping? I'm confused how load actual can drop unless either the throttle plate closes or boost drops?. Perhaps your valvetronic could be closing down to reduce load actual, independant of your throttle plate staying open?

Check out the logs, it's load actual and a bunch of other values causing lots of drama, lol.

 

[wheela]

Yeah, I should have looked at the log first, lol. I just edited my post above

 

[wheela]

Are you running with the MAF or MAF'less?

Thinking back to the side intake thread, I think all 3 of you guys are MAFless, of I'm remembering correctly?

 

[houtan]

We tried maf and mafless, same result

 

[houtan]

Pinging others for help...

@Jake@MHD
@V8bait,
@carabuser,
@Sbrach

@all4bspinnin

 

[wheela]

Yeah, your airflow/load calculation is definately going off at those points. MAF, load actual, and torque all follow eachother down, while AFR spikes up super high, and stft pegs at +30% trying to get AFR back on target (super lean because it's injecting based on incorrect airflow/load).

 

[JPuehl]

I've added more details to my original post and will update as I think of more.

 

[wheela]

The P value spiking may also make sense since even though the boost pressure is okay the air flow component is being mis-reported low. So compressor power is probably being undercalculated (DME thinks exhaust mass flow is lower than reality), so P spikes up to compensate since it's based on compressor power.

 

[JPuehl]

What's interesting is that 28.5# is 1.965004 bar and 2 bar is 29.007578#. Maybe there's a 2 bar limit in the code even with the N20 TMAP sensor?

I just sent MHD an email describing the problem so hopefully they have some insight to the issue.

 

[wheela]

What's interesting is that 28.5# is 1.965004 bar and 2 bar is 29.007578#. Maybe there's a 2 bar limit in the code even with the N20 TMAP sensor?

I just sent MHD an email describing the problem so hopefully they have some insight to the issue.

I'm excited to hear what MHD says about this

 

[JPuehl]

I'm excited to hear what MHD says about this

We ALL are, lol.