PDA

View Full Version : Disassembly: Stock Boost Gauge



Greg E
03-26-2012, 10:35 PM
I've had this already mapped out for Chrome V1, but in V2 I'm going to give you guys the ability to edit the stock Boost Gauge.

The stock Boost Gauge signal is PWM based signal sent across a single wire from the ECU. At 100% the gauge points to the top, 0% it's at the bottom. The ECU calculates the Boost Gauge Duty Cycle (BGDC) from a single 2D table in the ECU.

http://i.imgur.com/hlTqf.jpg

The axis is just engine load which is calculated from air flow, then compensated for temperature and then is stored in a location in memory (this is called "Load"). The look up values you see in the table are where the needle physically points on the gauge itself (-8 = bottom, 14 = top).

There's really no rhyme or reason to how mitsu setup the gauge. One would think that 100 engine load = 0psi but the engineers probably knew better. After flash tuning a couple cars with the EVOX map sensor, I've been noticing 0psi is actually closer to 70-75 Load which almost matches the table...

You guys can edit both sides of the table to read whatever you'd like!

donniekak
03-27-2012, 12:49 AM
Any plans on moving the knock sum to read on the stock boost gauge?

Greg E
03-27-2012, 12:51 AM
Any plans on moving the knock sum to read on the stock boost gauge?

No. I have no way to test that.

Think I figured out how to make the CE light blink under knock though. Further testing is needed.

ChargerX3
03-27-2012, 02:29 AM
No. I have no way to test that.

Think I figured out how to make the CE light blink under knock though. Further testing is needed.

How about with the Low fuel light, or the door chime? The chime would be cool...if possible.

Greg E
03-27-2012, 09:03 AM
none of those are tied thru the ecu.

A low fuel chime would drive you nuts. :p

ChargerX3
03-27-2012, 11:56 AM
none of those are tied thru the ecu.

A low fuel chime would drive you nuts. :p

Call me crazy, but I would be tempted to run a jumper from the CEL to the chime. That way its audible knock. ;)

Greg E
03-27-2012, 12:32 PM
A knock alarm would be sweet!

BigTyla
03-27-2012, 04:43 PM
Think I figured out how to make the CE light blink under knock though. Further testing is needed.

ECMLink has this feature and you can parameterize the knock count at which it is activated. Sure would be nice to have for the 3S community!

AdamVR4
03-27-2012, 05:23 PM
http://i.imgur.com/hlTqf.jpg

The axis is just engine load which is calculated from air flow, then compensated for temperature and then is stored in a location in memory (this is called "Load"). The look up values you see in the table are where the needle physically points on the gauge itself (-8 = bottom, 14 = top).

There's really no rhyme or reason to how mitsu setup the gauge. One would think that 100 engine load = 0psi but the engineers probably knew better. After flash tuning a couple cars with the EVOX map sensor, I've been noticing 0psi is actually closer to 70-75 Load which almost matches the table...

You guys can edit both sides of the table to read whatever you'd like!

Confused: Why would the highest load value be 14 kPa? That's 2 psi. Are those the factory values? If so, no wonder the stock gauge is so horribly worthless.

Awesome work by the way. Thank you!

Greg E
03-27-2012, 05:48 PM
LOL!

The numbers on the right side of the chart correspond to where the gauge needle physically points. So when the gauge shows 14, the engine load is about 140.

lawdogg
03-27-2012, 08:53 PM
Greg, all this stuff you're doing it mind-bogglingly awesome. I remember the first time you told me you were a computer engineer / programmer type. Holy hell you weren't lying!!!! :bigthumb:

Anyway it's interesting that airflow "LOAD" would be in kPa, must use a volume flow rate conversion based on the cross sectional area of the MAS and the pressure of the air inside it ... sheds some light on the whole MAS-webbing gutting thing.

donniekak
03-27-2012, 09:02 PM
I have hex code to scale the dsm boost gauge to knock sum, if that could give you an idea. I just have never tore into any of the 3s code at all.

Greg E
03-27-2012, 09:39 PM
Anyway it's interesting that airflow "LOAD" would be in kPa, must use a volume flow rate conversion based on the cross sectional area of the MAS and the pressure of the air inside it ... sheds some light on the whole MAS-webbing gutting thing.

I need to change that (carried in from Ceddy's original disassembly). That's not REAL kPa. Load is just a Mitsubishi designated number for corrected airflow. Changing the axis to g/sec gives you air mass based on flow. Then you can calculate exactly how much fuel you need to maintain your desired AFR.

The MAS is nothing more than a super accurate devise to measure air mass. When you gut it or install a piggyback after it, you're removing the one thing about it that makes it superior to any other MAF on the market.

Greg E
03-27-2012, 09:42 PM
I have hex code to scale the dsm boost gauge to knock sum, if that could give you an idea. I just have never tore into any of the 3s code at all.

I have a very good idea how to scale the code, but I don't want to damage my engine testing it.


...then again, I could change the knock raw value to a fixed value in the code very easily.... :chin:

lawdogg
03-27-2012, 10:16 PM
I need to change that (carried in from Ceddy's original disassembly). That's not REAL kPa. Load is just a Mitsubishi designated number for corrected airflow. Changing the axis to g/sec gives you air mass based on flow. Then you can calculate exactly how much fuel you need to maintain your desired AFR.

The MAS is nothing more than a super accurate devise to measure air mass. When you gut it or install a piggyback after it, you're removing the one thing about it that makes it superior to any other MAF on the market.

Ooooh ... Gotcha!

BigTyla
03-28-2012, 11:54 AM
The MAS is nothing more than a super accurate devise to measure air mass. When you gut it or install a piggyback after it, you're removing the one thing about it that makes it superior to any other MAF on the market.

I only wish more people realized this.

jba3
03-28-2012, 11:52 PM
I have a very good idea how to scale the code, but I don't want to damage my engine testing it.


...then again, I could change the knock raw value to a fixed value in the code very easily.... :chin:

Leave the engine off and gently tap the knock sensor area with a rubber mallet? I think that could give you some phantom knock to test with, but you'd need a second person to either tap or to watch the gauge.