EcoBoost 2.7 compatibility - VVT, dome control, etc.

[Laminar]

Hi all,

I currently have a 2.7 EcoBoost bolted into my RX-8 and my previous ECU plan appears to be falling through, so I'm exploring my options. I know Link has good CANBus support for the RX-8 and the G4X XtremeX appears to have the features I need, but I still have a couple of questions.

To head off the first question, I'm not running DI, I have a 2018 engine so I'll just be running port injection on E85. 

I believe the Link can handle my cam/crank/VVT setup, as I see in this thread:

another user has it working on a 3.7 V6, which shares the same cam tooth pattern. From his tune, it looks like he's running the "Ford Duratec" wheel setting, which I'm not 100% sure on. "Duratec" referred to both the 2.5/3.0/3.5/3.7 V6 engines, and then later to the 2.0/2.3 I4 engines which are totally unrelated to the V6. The cam has four teeth on it, arranged like a 6 tooth cam but with two teeth missing, maybe like a 3+1. 

 

The thread about the 3.7 didn't mention any VVT issues so I'm assuming it will work. The only difference is that the 2.7 came with a 60-2 crank wheel instead of the 3.7's 36-1. I have both wheels in hand but I figured I should use the 60-2 if it's supported. Thoughts?

Secondly, poking around the 3.7's tune it looks like VVT control is 1D, where cam position is tied only to RPM. The factory control is...maybe...4D? Based on driving conditions the ECU picks a "mode" (max power, economy, warmup, etc.), then based on that mode it uses load and RPM to select a cam position. I dug through all of those tables and came up with a single load vs. rpm for each cam, selecting values from economy or power based on what I want in that region, at least as a starting point until I can get it on a dyno:

For a turbocharged steet car, VVT can definitely affect spool and what's right at 2000rpm for cruise may not be what I want for 2000rpm and WOT. Can Link do 2D VVT control?

Third is boost control. The 2.7 uses its own style of dome control. The driver's side intake cam drives a vacuum pump that's piped into the boost control solenoid. When the solenoid opens, it generates a vacuum within the wastegate control tubing, opening the wastegates. Closing the boost solenoid bleeds pressure back into the wastegate circuit, closing the wastegates. The boost solenoid is PWM controlled at 250Hz and the circuit has its own pressure sensor within the wastegate circuit. The optimal control system would have a bias table with a desired wastegate system pressure per load and RPM, and then PID control to dial in the boost target from there. I see Link has a normal 2-solenoid dome pressure control system, but would it be able to handle a single-solenoid system like this?

 

And fourth, fuel pressure control. I have a new DW pump for the stock RX-8 in-tank assembly and had planned to control it via a Ford FPDM that takes a PWM signal and uses that to drive the pump. A PWM input of 15% to 50% varies the driver output 0-100%, and a PWM input of 75% shuts off the pump. I figured this would help me run big injectors but maintain idle quality instead of just running the pump at the RX-8's 60psi bypass all the time. Is there built-in support for PWM fuel pressure control, or can I set the Link up to run the pump like this?

The Link really has me excited, as it solves several issues with the motor swap - cruise control, DBW, dual Spartan3 widebands over CAN, and more. I just want to make sure it will do exactly what I want before dropping a fair bit of change on it and spending hours wiring in (another) ECU.

Here's the car as it sits now:

Thanks!

[Laminar]

Dug around a bit more and found my answer for fuel pressure control...nice! Also great that it supports a variable voltage 4-button cruise setup.

[Vaughan]

7 hours ago, Laminar said:

nother user has it working on a 3.7 V6, which shares the same cam tooth pattern. From his tune, it looks like he's running the "Ford Duratec" wheel setting, which I'm not 100% sure on. "Duratec" referred to both the 2.5/3.0/3.5/3.7 V6 engines, and then later to the 2.0/2.3 I4 engines which are totally unrelated to the V6. The cam has four teeth on it, arranged like a 6 tooth cam but with two teeth missing, maybe like a 3+1. 

The thread about the 3.7 didn't mention any VVT issues so I'm assuming it will work. The only difference is that the 2.7 came with a 60-2 crank wheel instead of the 3.7's 36-1. I have both wheels in hand but I figured I should use the 60-2 if it's supported. Thoughts?

The "Ford Duratec VCT" Trigger mode is for 36-1 trigger setups but I don't see any reason why a Multi/Missing setup with Trigger 2 set to "Cam Pulse window" wouldn't work.

7 hours ago, Laminar said:

Secondly, poking around the 3.7's tune it looks like VVT control is 1D, where cam position is tied only to RPM. The factory control is...maybe...4D? Based on driving conditions the ECU picks a "mode" (max power, economy, warmup, etc.), then based on that mode it uses load and RPM to select a cam position. I dug through all of those tables and came up with a single load vs. rpm for each cam, selecting values from economy or power based on what I want in that region, at least as a starting point until I can get it on a dyno:

For a turbocharged steet car, VVT can definitely affect spool and what's right at 2000rpm for cruise may not be what I want for 2000rpm and WOT. Can Link do 2D VVT control?

All the VVT target tables are '3D' (two axes), typically engine speed and throttle position are used on the axes. If an extra axis is required games can be played with putting a math block or a value from another table on one axis. You can also switch between up to 3 different target tables if desired.

7 hours ago, Laminar said:

Third is boost control. The 2.7 uses its own style of dome control. The driver's side intake cam drives a vacuum pump that's piped into the boost control solenoid. When the solenoid opens, it generates a vacuum within the wastegate control tubing, opening the wastegates. Closing the boost solenoid bleeds pressure back into the wastegate circuit, closing the wastegates. The boost solenoid is PWM controlled at 250Hz and the circuit has its own pressure sensor within the wastegate circuit. The optimal control system would have a bias table with a desired wastegate system pressure per load and RPM, and then PID control to dial in the boost target from there. I see Link has a normal 2-solenoid dome pressure control system, but would it be able to handle a single-solenoid system like this?

This should be able to be done with regular boost control, you are just switching between pressure and a constant vacuum instead of between pressure and atmospheric. The Link Closed Loop Boost control has tables for expected required wastegate duty cycle and applies a PID over the top of that as well as having 3 stages to allow a fast spool and precise control on approach of boost target.

[Laminar]

Thanks so much for the reply, sounds like this is the direction I want to go. Time to start putting together my IO list. Also wild that Link even supports the canister purge valve...no need to vent that thing to atmosphere like my last build 

[Laminar]

Okay, maybe a silly question - what's included with the ECU? I found an unboxing video that showed that the USB cable is included, but do I need to purchase the CAN harness separately?

[Vaughan]

2 hours ago, Laminar said:

Okay, maybe a silly question - what's included with the ECU? I found an unboxing video that showed that the USB cable is included, but do I need to purchase the CAN harness separately?

https://linkecu.com/g5/voodoo-pro-specs/ - has a what's in the box section at the bottom of this page.

One CAN port is in the E plug and the other is on the black connector by the Wifi antenna plug. A plug/loom, B plug/loom, E plug and the CAN 1/KLine plug are available separately to the ECU.

[Laminar]

I think that's different from the one I'm looking at, the G4X XtremeX.

[Vaughan]

For the G4X Xtreme X one CAN port is in the B plug on pins shared with DI 9&10 and the other CAN port is on the round black plug below the USB plug. A plug/loom, B plug/loom and the CAN 1/Serial plug are available separately to the ECU.

The ECU comes with a USB tuning cable, mounting bracket, quick start guide and some stickers from memory.

[Laminar]

Gotcha - looks like another $50 if I want to use CAN1 then.

Next question - if I'm using Closed Loop Pressure Control for the fuel pump, do I still need a fuel pump relay or is PWM control safe enough on its own?

• This should be able to be done with regular boost control, you are just switching between pressure and a constant vacuum instead of between pressure and atmospheric. 

Following up on this - the vacuum is variable depending on solenoid position and there's never any pressure, it's a single-sided wastegate so there's only vacuum or atmospheric. The wastegates are fully closed at atmospheric pressure, they begin opening around 7"Hg, and they're fully open by around 17"Hg. This also means that more duty cycle = less boost. Is there a setting to invert the output? Everything I see in the Boost Control settings seem to assume more DC = more boost.

[Laminar]

And another one. From reading, I understand the difference between MAP and MGP, but I'm not sure how to scale my tables. I think I now assume atmospheric is 0kPa, vacuum is in the negative, and boost is in the positive? Then if any of my tables still use MAP as the axis, I'm back to ~100kPa as the crossover point between vacuum and boost. Is that right?

[Vaughan]

7 hours ago, Laminar said:

if I'm using Closed Loop Pressure Control for the fuel pump, do I still need a fuel pump relay or is PWM control safe enough on its own?

You'll need a solid state relay or some kind of pump controller between the ECU and the pump, the ECU can't supply enough current to control the pump directly but you don't need a another relay to turn that on and off.

7 hours ago, Laminar said:

Following up on this - the vacuum is variable depending on solenoid position and there's never any pressure, it's a single-sided wastegate so there's only vacuum or atmospheric. The wastegates are fully closed at atmospheric pressure, they begin opening around 7"Hg, and they're fully open by around 17"Hg. This also means that more duty cycle = less boost. Is there a setting to invert the output? Everything I see in the Boost Control settings seem to assume more DC = more boost

You can flip the polarity by changing the selected output's active state.

1 hour ago, Laminar said:

And another one. From reading, I understand the difference between MAP and MGP, but I'm not sure how to scale my tables. I think I now assume atmospheric is 0kPa, vacuum is in the negative, and boost is in the positive? Then if any of my tables still use MAP as the axis, I'm back to ~100kPa as the crossover point between vacuum and boost. Is that right?

Atmospheric is 0kPa MGP but ~101kPa MAP at sea level. Boost control is all done using MAP, boost is when MAP is greater than BAP and vacuum is when MAP is less than BAP. MGP is positive for boost and negative for vacuum.

[Laminar]

Awesome, that all makes sense. 

[Laminar]

Okay! Big milestone! Finally got the engine pulled back out of the car and on the run stand so I can seal up the oil system and crank it over.

I was able to set the base timing, it required adding 180 degrees to my expected tooth position.

I took a couple of trigger scopes with the plugs pulled.

Then I put the plugs in and sprayed some starting fluid and it fired right up!

I ran it a couple of times, the final start attempt it decided to blow a plug out of the intake, so I'm guessing it's syncing on a random cycle and then backfiring through the intake. I was in a super time crunch at the time and didn't have a chance to look at any of my logs. I'm reviewing them this morning and trying to figure out what all went down. Going back through this thread, I see Vaughan recommended Cam Pulse Window, but I was still set to Cam Pulse 1x so that wasn't helping me. Based on the trigger scope, I'm having trouble seeing where the missing teeth are to pick out my Sync Tooth and my Window Length. I figure I can use the lone tooth as my sync tooth as it should have room for VVT to move it around without losing sync.

Also, should my Trig 2 Edge be set to rising so that Trigger 2 is only active while the tooth is passing by?

 

Trigger Scope - 2024-02-2 2;29;24 pm cranking 1.llgx Trigger Scope - 2024-02-2 2;29;47 pm cranking 2.llgx PC Datalog - 2024-02-2 2;40;42 pm fire1.llgx PC Datalog - 2024-02-2 2;42;15 pm fire2.llgx rx8-xtremex.pclx

[Laminar]

Here are the fireups from the "run 2" log

https://www.youtube.com/watch?v=3X5K_6AlEew

[Adamw]

Is trigger 2 on an intake or exhaust cam?  If it is an intake cam, do you know if they only advance from the lock/home position or can they both advance and retard?

[Laminar]

It's an intake cam, and I'm about 90% sure it only advances from the home position.

[Adamw]

Some of these modern hall sensors do a very short pulse when they see a tooth, the sample rate of the scope is not great enough to capture these reliably so that's why your capture looks odd and it is difficult to see the gap etc.  However we can use "trigger 1 state" to indicate where the gap/tooth 1 is.  A cam pulse window sync mode set to sync tooth 6 with a length of 80 deg should work.

I put some notes on your scope to help explain how it works, although it is not a great example with all the missing trig 1 teeth (I use a engine angle overlay app to approximate tooth locations and angles).  The general idea is we want a trig 2 edge to always occur within the window on one crank revolution and never occur in the same window on the next crank revolution.

Your offset may or may not change by 360deg since it was syncing/mis-syncing randomly before.  

   

[Laminar]

Makes perfect sense, thanks Adam. 

[Laminar]

I made the changes to my sync tooth and cam setup. I got the fuel system hooked up to see if I could fire the engine of its own accord, but no dice.

I haven't yet hooked the timing light back up to confirm timing. I previously tested the injectors and heard them all click, and I confirmed they're getting 12V now. I pulled a plug and it wasn't particularly wet, maybe a tiny bit damp? I tried adding 50% to the applicable Fuel Table 1 cells but no change. Injector 1 shows active but duty shows 0%. I was playing with running the fuel pump independently, but at the end of the log I definitely have good pressure.

I see my battery voltage dropping to 8V during cranking, I need to make sure I have the charger on the battery, it might have popped off.

Is there anything else in this log that would be a no-go for startup? Not even the tiniest chuff?

I'll check my timing and may have to clean the injectors, they've been sitting for a few years with old gas in the rails, they could be clogged.

rx8-xtremex.pclx PC Datalog - 2024-02-5 6;32;25 pm crank no start.llgx

[Adamw]

On 2/5/2024 at 2:18 PM, Adamw said:

Your offset may or may not change by 360deg since it was syncing/mis-syncing randomly before.  

Did you try both options? 

Are they really only 310cc injectors?  They wont get you very far with E85.

Your first few cranks had very low fuel pressure so with the FP comp enabled this has resulted in a very long PW.  In the last couple of cranks the fuel press is better, but still lower than expected, this coupled with the lambda target of 0.78 and the 120% VE in the fuel table is still giving you are very long PW so I would say its getting too much fuel assuming all hardware is working.   

Otherwise it looks like all the basic necessities of life are present.  

 

 

[Laminar]

Yes - factory PI injectors intended only to supplement the DI system. Once I prove this thing will run on just PI I’ll get some real injectors that will meet my horsepower goal. 

[Laminar]

Got the engine to fire up under its own power. I'm a dummy and sucked some junk up into my fuel pump in the process. After disassembly and cleanout it's still hurt, so I'm only making about 140kpa of fuel pressure. New one is on the way (with a sock this time). After a garage session I'm firing the engine up for a few seconds just to test (still no coolant), but I'm working on finding good e-throttle and post-start enrichment settings.

This last time I got two chuffs out the intake before I got a real fireup. On the two chuffs, I see trigger one drop, status goes "Blank time" then "Blank teeth" then "Second tooth." On the actual fireup it fires before that happens. Did I mess something up by switching Trigger 2 to Rising?

 

Full datalog:

https://drive.google.com/file/d/1VAB4OFKh8jZ7lB-d1_5AwLvXVL5Pdfg-/view?usp=sharing

rx8-xtremex.pclx

[Laminar]

Latest datalog:

https://drive.google.com/file/d/1VFtPmPx3jF_EgX8KGufgq8rJ0f3xtH2R/view?usp=drive_link

It still seems like it occasionally picks the wrong cycle to sync on. It will sometimes fire right up and other times pump raw fuel out the exhaust and then ignite it. 

Double checked timing today and confirmed it's good.

[Laminar]

I double-checked my firing order configuration and that's correct, I want to go back and test each coil to make sure I wired everything properly. The engine should alternate firing on each bank, but in my video I can see bank 1 hitting twice before I see another spark from bank 2. It definitely doesn't sound like I'm getting all 6.

The other odd thing I noticed is that I'm only seeing about -6kPa at idle, I'm not seeing good idle vacuum. I see -50kPa on overrun but not at idle. I assume this could be a function of a bad firing order or maybe VVT out of control. Next time I get it running I want to do the cam calibration and see if I can make VVT happy. I need to get the radiator plumbed up so I can run it for more than about a minute a day.

[Laminar]

I was able to run the cam calibration last night, incredibly slick process. It detected the four teeth per cam and found my tooth angles.

I confirmed all 6 ignition coils are properly wired and put in fresh plugs. Still running poorly and the MAP is suspiciously high.

@Adamw I'm interested in your thoughts on the following:

1. In the log below, there's a failed crank event followed by a successful crank event. Do I need to adjust my sync tooth and/or window? It's still 50/50 on whether it starts or just chuffs and/or blows fire.

2. I calibrated my MAP sensor (factory IAT/MAP combo sensor Ford GV2A-9F479-BA "Germany") based on a known good MAP sensor and got 0V=10kPa and 5V=304kPa. The factory tune (via HPTuners) says the MAP sensor has a slope of 65.12kPa and offset of 0.47kPa. Using those numbers, I would calculate 0V=0.47kPa and 5V=326kPa. I can hook up a vacuum gauge to the manifold and compare it to the MAP reading while running but I haven't yet. All of that to say I'm not sure why my MAP signal is so high compared to what I would expect at idle.

I think I want to get my piston stop back out and double confirm my timing marks vs. TDC on #1, and possibly do a compression check on all cylinders. Anything else I should look at?

https://drive.google.com/file/d/1VJoKPYPfH62Uue69jWKKOgxdu3-3m-0g/view?usp=sharing