GDI: 2018 Cadillac ATS "The last gamble"
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
GDI: 2018 Cadillac ATS "The last gamble"
This clean title low miles Copart special is a replacement for the disastrous Malibu https://rusefi.com/forum/viewtopic.php?f=2&t=2835
https://github.com/rusefi/rusefi/wiki/GM#e80
https://github.com/rusefi/rusefi/wiki/GM#e80
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: 2018 Cadillac ATS "The last gamble"
What the hell, 2018 tire did what?! Was that caused by the accident?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Re: 2018 Cadillac ATS "The last gamble"
Probably. I the car from the NE? In the hot sunny states dryrot is very common, some places tires over 5 years don't pass inspection i read so if from a place like that maybe just age. Here in PA I just learned most real tire places will no longer mount used tires of any age.....so winter/summer you kind of need separate wheels now.
Hopefully it's a winner, they are nice cars.
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: 2018 Cadillac ATS "The last gamble"
Wrong color great condition fender and some less amazing parts, all for $150? And the fun of junk yarding during a light snowfall? Sold!
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: 2018 Cadillac ATS "The last gamble"
https://en.wikipedia.org/wiki/General_Motors_Alpha_platform means engine bay architecture matches that of https://en.wikipedia.org/wiki/Chevrolet_Camaro_(sixth_generation)
Problem: ECU is under the fuse box making it uncomfortable for me to do my things.
Proposed solution: take fuse box apart, install cables between parts allowing tilting of larger square piece. Mine is 2018 with eight plastic latches, on early models these pieces are bolted together.
Problem: ECU is under the fuse box making it uncomfortable for me to do my things.
Proposed solution: take fuse box apart, install cables between parts allowing tilting of larger square piece. Mine is 2018 with eight plastic latches, on early models these pieces are bolted together.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: 2018 Cadillac ATS "The last gamble"
WIP
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: 2018 Cadillac ATS "The last gamble"
End result
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: 2018 Cadillac ATS "The last gamble"
https://rusefi.com/docs/pinouts/GM-E87-TCU/ a step forward 

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: GDI: 2018 Cadillac ATS "The last gamble"
Start/stop adds second relay:
https://www.cadillacforums.com/threads/enhanced-starter-motor-operation-in-engine-stop-start-systems.1042850/It is interesting that the the ECM controls the starter motor and pinion separately:
In order to smoothly restart the engine as quickly as possible while managing the greater number of engine starts, the Stop/Start system uses an enhanced starter motor that operates differently from a conventional starter motor. (Fig. 6) It has a high performance electric motor and a stronger pinion engagement mechanism than a conventional starter. It also has independent control of the pinion and motor.
"When the vehicle is coming to a stop, just before the engine stops rotating (at approximately 50 RPM) during stop/start operation, the ECM energizes the Starter Pinion Solenoid Actuator Relay to easily push the pinion gear into the flywheel gear without gear clash. (Fig. 8) When the engine stops rotating during Stop/Start operation (Auto Stop mode), the starter pinion gear is fully engaged, ready for the starter motor to become energized to quickly start the engine again."
A normal starter cannot engage the pinion until the engine is completely stopped. The separate control enables a restart even before the engine is completely stopped if the driver attempts to accelerate before that complete stop.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: GDI: 2018 Cadillac ATS "The last gamble"
Fun fact: TS already embeds tunes into log files but not sure if there is an official toolset to properly extract?
- Attachments
-
- 2025-02-02_16.12.27 angle 40.zip
- (2.02 MiB) Downloaded 9 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Re: GDI: 2018 Cadillac ATS "The last gamble"
Looking at the logs,
1. I don't see "GDI: HPFP duration" go very high. This makes me think the integral term is too conservative? It might be that you need to increase the P or I term to get things going, then once you get a running car you can tune the HPFP tables, then decrease P&I so the controller doesn't overcompensate.
2. Why doesn't the motor start? It sure seems you have enough fuel pressure to get something in the cylinders. If you remove a spark plug do you smell fuel? (as I type from a safe distance).
3. I do see a change in the amount the controller is working as the angle changes.
ANGLE=10, GDI: HPFP Duration around 50-55, Fuel Pressure _High around 40-50 bar
ANGLE=40, GDI: HPFP Duration around 70-75, Fuel Pressure _High around 25-40 bar
ANGLE=55, GDI: HPFP Duration around 90+, Fuel Pressure _High eventually around 50 bar
FWIW it's probably "more reliable" at low power levels to go with a higher angle (30-40). If ANGLE=10 is too low, and the HPFP Duration goes to 120 degrees (rough duration of half a HPFP lobe), then you might catch the downstroke of the pump. Since the activation angle is usually relatively short (say, 10-15 degrees), then the solenoid might turn off before the HPFP piston moves up, meaning you'll get -zero- fuel instead of all-the-fuel.
Though given the pump duration is around 50-55 degrees at ANGLE=10, I would guess the correct angle is even lower. HOWEVER, this is not what's stopping you from starting the car or getting a decent high pressure reading. I would focus on getting the car started first using ANGLE=10-20, then worry about stabilizing the HPFP system (which you should be able to do at idle pretty with some trial and error).
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: GDI: 2018 Cadillac ATS "The last gamble"
Oh, this 4 cylinder totally stars and idles. Just that all these logs were created within 15 minutes of first start ever, it idles around 350 RPM - hope to improve the tune once I am back from vacation!
The point was to grab _any_ logs with different angle parameter.
The point was to grab _any_ logs with different angle parameter.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Re: GDI: 2018 Cadillac ATS "The last gamble"
ah ok. Well think of this way: You need to specify where the top of the lobe is. The control algorithm will fire off the solenoid somewhere from 0-120 degrees BEFORE that point (assuming 3-lobe setup, and nothing odd like 110/130 degree down/up duration on lobe).
If you set the angle 40 degrees too late, then the control from 0-40 degrees won't do anything (it will map to the HPFP piston downstroke). So the controller will work in the 40-120 degree range, but that really only maps to 0-80 before TDC, meaning you won't be able to get the full delivery of the HPFP. It will still drive, at least at part throttle, but if you ask for too much fuel, the pressure in the rail will fall and you'll lean out the motor.
If you set the angle 40 degrees too early, then the control from 0-80 degrees will be fine. As long as the controller sticks in this range, you'll be fine, and in fact you'll be able to get the full use of the fuel pump. The pressure control may not be very stable, because even a small request from the HPFP will result in a lot of fuel (so the pressure rail may look like a sawtooth pattern, assuming your logging is fast enough - I'm not convinced all the variables in your log files are being updated fast enough). HOWEVER, and this is a big exception, if the controller ever exceeds 80 degrees, and operates in the 80-120 range, your pressure will drop like a rock (basically to the same value as the low pressure rail). This is because you'll be operating 120-160 degrees before the actual TDC, meaning you're on the downstroke of the HPFP piston. The solenoid will turn off before the piston moves up, thus the pump won't latch and no fuel will be added.
So late is safe but won't deliver full fuel. The controller will compensate. This is what your log files showed - as you increased the angle, the HPFP Duration increased to match. The car still seems like it's fine but that's because the controller/PID/etc were masking the problem. This is the answer to your question "do we even see a difference." Yes, yes we do - in the GDI: HPFP Duration.
I'm guessing the actual TDC is in the -10 to 20 degree range, but that's a mostly random guess. You can determine where this point is by falling off the wrong end then backing up 120 degrees. You limit the operational range of the HPFP pump to, say, 20 degrees, set the no-op deadzone to 10 degrees (I forget what it's called, but it's the range where if the controller were to ask for < X degrees the solenoid is not operated at all). Then you back up the angle of TDC until the rail pressure falls to 5 bar. Then add 100-110 degrees back to the TDC value and reset the operational range to 120 degrees. Just be prepared for wild high pressure rail swings when doing this. I think it's safe, I think I did this before, I think.....
If you set the angle 40 degrees too late, then the control from 0-40 degrees won't do anything (it will map to the HPFP piston downstroke). So the controller will work in the 40-120 degree range, but that really only maps to 0-80 before TDC, meaning you won't be able to get the full delivery of the HPFP. It will still drive, at least at part throttle, but if you ask for too much fuel, the pressure in the rail will fall and you'll lean out the motor.
If you set the angle 40 degrees too early, then the control from 0-80 degrees will be fine. As long as the controller sticks in this range, you'll be fine, and in fact you'll be able to get the full use of the fuel pump. The pressure control may not be very stable, because even a small request from the HPFP will result in a lot of fuel (so the pressure rail may look like a sawtooth pattern, assuming your logging is fast enough - I'm not convinced all the variables in your log files are being updated fast enough). HOWEVER, and this is a big exception, if the controller ever exceeds 80 degrees, and operates in the 80-120 range, your pressure will drop like a rock (basically to the same value as the low pressure rail). This is because you'll be operating 120-160 degrees before the actual TDC, meaning you're on the downstroke of the HPFP piston. The solenoid will turn off before the piston moves up, thus the pump won't latch and no fuel will be added.
So late is safe but won't deliver full fuel. The controller will compensate. This is what your log files showed - as you increased the angle, the HPFP Duration increased to match. The car still seems like it's fine but that's because the controller/PID/etc were masking the problem. This is the answer to your question "do we even see a difference." Yes, yes we do - in the GDI: HPFP Duration.
I'm guessing the actual TDC is in the -10 to 20 degree range, but that's a mostly random guess. You can determine where this point is by falling off the wrong end then backing up 120 degrees. You limit the operational range of the HPFP pump to, say, 20 degrees, set the no-op deadzone to 10 degrees (I forget what it's called, but it's the range where if the controller were to ask for < X degrees the solenoid is not operated at all). Then you back up the angle of TDC until the rail pressure falls to 5 bar. Then add 100-110 degrees back to the TDC value and reset the operational range to 120 degrees. Just be prepared for wild high pressure rail swings when doing this. I think it's safe, I think I did this before, I think.....
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: GDI: 2018 Cadillac ATS "The last gamble"
Fun fact: 4 cylinder Global A GM vehicles seem to have four lobes while single-pump Global A GM v8 has three lobes. No idea about v6
So 90 per lobe here but same idea.
So 90 per lobe here but same idea.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Re: GDI: 2018 Cadillac ATS "The last gamble"
A bit of a segue. Are you sure about this? I took a look at the info field on one of these and found:
"rusEFI master.2025.01.26.alphax-8chan.2832297523: rusEFI v20250126@1431655765"\n"Capture Date: Tue Jan 28 13:52:56 EST 2025, File author: TunerStudio MS Dev version 3.2.05"\n\n\x00
I used this document: https://www.efianalytics.com/TunerStudio/docs/MLG_Binary_LogFormat_2.0.pdf
to extract the info field. The doc says its:
Info Data is optional ASCII text that would normally be found in the header of a delimited log. It is unstructured
informational data that may include the firmware version, date captured or any other informational text.
FWIW I wrote a decoder for a graphing program I'm working on (oriented more toward track use as there's a support for laps, using distance on the X axis, and aligning with video).
https://github.com/racer-coder/TrackDataAnalysis/blob/main/data/megalog.py
- AndreyB
- Site Admin
- Posts: 14611
- Joined: Wed Aug 28, 2013 1:28 am
- Location: Jersey City
- Github Username: rusefillc
- Slack: Andrey B
Re: 2018 Cadillac ATS "The last gamble"
but v8 Camaro harness is routed very differentlyAndreyB wrote: ↑Fri Jan 10, 2025 4:42 amhttps://en.wikipedia.org/wiki/Chevrolet_Camaro_(sixth_generation)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute