Magnuson supercharger + smoothboost with factory E92 as PWM boost controller

[pisces_0]

What year/model/os are you running?
Maybe certian vehicles came with some of the behind the scene parameters already populated.

Also, can you confirm/Experiment that the P012B error is enabled in the DTC list?
Without that error specifically enabled, the weird dips in Supercharger Inlet Vaccum dont appear.
All you should need to do is start and idle the vehicle for the error condition to run

2016 Suburban
OS 12672612

P012B, P0101 and P1101 are all enabled to set on 2nd error. None are triggering.

[will_974]

2016 Suburban
OS 12672612

P012B, P0101 and P1101 are all enabled to set on 2nd error. None are triggering.

This confirms that these parameters are indeed different across different variants of Gen V vehicles
Were your SCIAP linear and offset also set from the factory?
If i recall correctly, I read your thread and everything worked well for the swap in your case.
If you dont mind, can you PM me your (preferably stock )Tune file. I promise you I am not a professional tuner, I am only going to use it to see what the deltas are in the Binary, And if i can determine what the source is doing between the two OS's

Sounds like your pressure sensors may not be configured properly though... Sensor measurement isn't considered trustworthy until it passes necessary diagnostics, P012B in your case. And based on your description, it sounds like you don't have a baro sensor installed...so I'm curious what the baro variable is doing. A data log would tell a lot.

The barometer variable in the logger is reading from the MAP inside the MAF sensor.
I confirmed this at one point by breaking out the map from the MAF and wiring it to a standalone sensor.
What would you like to see in a data log? I am happy to provide

[smokeshow]

The barometer variable in the logger is reading from the MAP inside the MAF sensor.
I confirmed this at one point by breaking out the map from the MAF and wiring it to a standalone sensor.
What would you like to see in a data log? I am happy to provide

Just how baro, SCIAP, MAP and the throttle are behaving during various drive conditions would tell a lot.

[will_974]

I dont believe there is an issue with the barometer. It stays at 99-100kpa throughout the entire log.
There are no engine codes disabled, and no engine codes present.
Is there a way to tell from the logs if a sensor is 'untrusted'?
I will gather up some logs that show this, but have not had any cycles through the holidays

With the fix to P012B working, The car does not go into a MAF failed state. However, The car will no longer open the boost solenoid.

[will_974]

Update Here,

There is a lot of behind the scenes things that work together to keep the MAF out of a failed state.
I fought this piece for a long time and i think i have finally worked out the kinks.

The code I posted earlier with the screenshot is irrelevant and not used under boosted conditions.

However what is used is all the different DTC enables and checks.
These I have worked tirelessly trying to isolate down to exactly which switches to throw.
If anyone wants to go down this path, I can help. However I am still very much working out the details.
currently I have it working about 90% of the time. the other 10% it decides not to allow boost to build.


On a different note, and hijacking my own thread.
I successfully upgraded my Instrument cluster to have the ZL1 Blower Boost gauge.
The gauge works flawlessly, and the Instrument panel is very happy with the ZL1 update in the SS

I posted some details here: Link

20231212_155331.jpg

[jmardlin]

I wonder how much of this would apply to turbochargers in a truck with an E92. I would love to have boost control from the ecm.

[GPGTP]

Update Here,

There is a lot of behind the scenes things that work together to keep the MAF out of a failed state.
I fought this piece for a long time and i think i have finally worked out the kinks.

The code I posted earlier with the screenshot is irrelevant and not used under boosted conditions.

However what is used is all the different DTC enables and checks.
These I have worked tirelessly trying to isolate down to exactly which switches to throw.
If anyone wants to go down this path, I can help. However I am still very much working out the details.
currently I have it working about 90% of the time. the other 10% it decides not to allow boost to build.


On a different note, and hijacking my own thread.
I successfully upgraded my Instrument cluster to have the ZL1 Blower Boost gauge.
The gauge works flawlessly, and the Instrument panel is very happy with the ZL1 update in the SS

I posted some details here: Link

20231212_155331.jpg

Sir...you are my hero lol. I was just wondering about that myself. I was told a complete ZL1 OS swap would be necessary to enable that gauge. I'll send you an email.

[Blownlt1]

Hello guys, is it possible to use the regular sensor for IAT2 and not use the LT4 map?, also what pins does it go to in the ecu?.

[will_974]

yes,
Follow the pinout instructions in the first post of this thread

[will_974]

Sorta Good news.

I figured out where i went wrong.
The gm supercharger bypass valve from the ecm perspective is a enable signal not an disable signal.
This means that I have had my logic wrong the entire time, and as far as the ecm was concerned, boost was allowed.

what this means for the smoothboost aspect of the project is that the signal that comes from the ecm would need to be pwm reversed to properly actuate.
smoothboost expects 100% duty cycle as fuly disabled, while the ECM will give 0% duty cycle for fully disabled.

This would also explain fully why when we encountered an error condition, the boost solenoid would actuate, since the ecm thought it was disabling the signal. and smoothboost was like "this is my time to shine". Very confusing, but tracks with all my data points.

[will_974]

Working to put together a class/module/learning session on some of the lower level things that i have been learning.

I have made a forum for interest from the community. Looking for feedback and to gauge how much interest there would be
This would be a low level class on deciphering a binary calibration file, which would have alot of expected prereqs around low level computer science.

https://minigeist.net/blog/?p=215

[GHuggins]

When you say low level computer science?

[RobZL1]

he means all the nuts and bolts, and you need a CS degree to understand it.... LOL

[Alvin]

Working to put together a class/module/learning session on some of the lower level things that i have been learning.

I have made a forum for interest from the community. Looking for feedback and to gauge how much interest there would be
This would be a low level class on deciphering a binary calibration file, which would have alot of expected prereqs around low level computer science.

https://minigeist.net/blog/?p=215

This is really good for the community. The more people doing this the more break though we might have. Thank you.

[will_974]

When you say low level computer science?

For deeper digging you will need a good foothold in the C language
The decompiler does not make things easy to follow and sort through.

However, for a lot of the cross calibration table locating, you really only need a basic understanding of C data types and the ability to use the search feature of the decompiler
https://en.wikipedia.org/wiki/C_data_types

the majority of the values i have seen are either float or short.
The short values for the most part use a multiplier to increase precision and limit range.


I wouldn't say its required to have a C background, but it does help.

[GHuggins]

For deeper digging you will need a good foothold in the C language
The decompiler does not make things easy to follow and sort through.

However, for a lot of the cross calibration table locating, you really only need a basic understanding of C data types and the ability to use the search feature of the decompiler
https://en.wikipedia.org/wiki/C_data_types

the majority of the values i have seen are either float or short.
The short values for the most part use a multiplier to increase precision and limit range.


I wouldn't say its required to have a C background, but it does help.

Soooo, how soon is the class going to be and what would you recommend for learning C language or is it even possible to learn quickly? I knew it was going to be something like this. Already had one person walk me through this and I was lost in the binary part in about 10 secs

[will_974]

Soooo, how soon is the class going to be and what would you recommend for learning C language or is it even possible to learn quickly? I knew it was going to be something like this. Already had one person walk me through this and I was lost in the binary part in about 10 secs

I wanted to gather interest first.
And for the official class i was going to build a syllabus and powerpoints and try to break it down as much as possible.

My goal was to host it once i had atleast 10 people that were interested in the class.
But if we dont get there, I may do it on a one on one basis.

Im using the forum as the list of people who are interested in the class. So if you are interested, please fill out the forum so i know who to expect and whom to email when i get it all set up.

For learning C,
It would be difficult to ramp up very quick. because this will be very deep in the weeds.
But if i were to suggest a reading subject, I would read up on C functions.
Given function signatures, and their respective argument types. pass by value, pass by reference.
Also read up on variables, and how they are switched on and how they are set.

https://www.geeksforgeeks.org/c-functions/

This will not be enough to jump in running, but hopefully will be enough to keep your head above water.

[Alvin]

C is really not that bad to learn.

I learned a lot doing arduino projects. It reads ok IMHO. Obviously syntax is important and knowing how to use variables so that they don't over flow.. or when you need a integer verses a decimal etc..

[GPGTP]

I'm interested. Filled out the form at the link