Hey Everyone,

I’ve recently “finished” tuning my car to 23PSI using the open loop boost function and a 4 port mac valve and the plan is to expand to closed loop boost control setup to run 5 different levels of boost based off a rotary trim switch and sensed ethanol content. I’m not sure if the 4 port mac will prove to be too sensitive for the closed loop boost operation to function accuratly, but that is part of the experiment and I will easily swap back to the old faithful 3 port if it proves too touchy.

The car is a semi peripheral port 13B-REW, large street ports on the irons, FD intake ports opened to cosmo spec, pinned, balanced/lightened/clearanced, LIM peripherals bridged to both primary and secondary runners, full 4” open exhaust, CDI ignition, 6X ID2000s, 2X external 044s on a sumped tank, John Gleason Gleasemanufacturing Big Runner Manifold with 2X Turbosmart Gen 5 Comp 40 Gates, and a Borg Warner S476 Turbo with the 1.25 hot side housing along with a large number of other supporting go fast parts. Its currently tuned in TPS-Load (Alpha-N) fueling using a RPM X TPS base fuel table along with a MAP compensation table so it is a bit different than you traditionally see, but that doesn’t really have an impact on the boost control scheme.




The 4 port mac valve is the standard (but still genuine MAC) P/N: 46A-AA1-JDBA -1BA purchased on ebay for $48.49 and plumbed using all fragola PTFE 3AN lines as shown in the diagram below.



I’ll be using a Motec 4 position (DST-4) rotary trim switch to set the target boost level. I know Haltech makes their own 12 position switch, but I really don’t need that many positions and believe that for my application I’ll be able to setup this 4 position to run 14 PSI (spring, not selected but defaulted to for low E content) 18 PSI, 23 PSI, 28 PSI, 32+ PSI.

Motec DST-4 4 Position Trim Switch





Before starting the closed loop stuff, I fully tuned the car to the 14 PSI spring pressure on regular pump 93 without using the controller at all and established a safe fueling and timing baseline pump gas. Next, I filled the tank (starting 1/4 tank of E10) with E85 which ultimately ended up being about E55 when all blended. At the E55 level, I then re-ran the same 14 PSI spring file (with the ECU setup to sense ethanol content via flex fuel sensor and acccount for the flow difference) and verified that everything still ran on point as intended.

With the car running dead nuts on target AFR to 14 PSI it was time to start turning up. Big thanks to Monsterbox for coming down for the weekend to ride shotgun and help with the initial turn-up-boost tuning. Having a tuning copilot who knows what they’re doing behind the PC helps a TON and makes things go so much faster than being the driver/tuner and tuning alone on the street. If interested, here is the link to that tuning session.


So, at the end of that session we end up with a table targeting 23 PSI on E55-56. As expected with the MAC valve, 0-20% duty didn’t actually do anything, but I believe it was around 22% that we were able to lift it up to about 18 PSI, then by 27% it was running dead nuts on 23 PSI.

I’ve driven the car a good bit at the same level since then and refined the boost table by adding duty at low rpm to increase spool time, and dialing back down the duty to 25% to hold the same 23 PSI as the temperature dropped since the tuning runs with David and I was seeing 25-26 PSI at the same 27% duty.



Again, given the large increase in boost from a small duty cycle change on the 4 port I’m perfectly aware it may prove too sensitive for closed loop control but if needed I’ll swap back to a 3 port or cold even retain open loop and simply setup the trim **** to run open loop boost at each selected position.

First off, I’m wiring the Motec DST-4 trim switch to AVI 10 on my harness (previously TPS connector, now unused as I’m DBW). It’s a good fit for that AVI as the old TPS connector already has a +5V voltage supply, sensor ground. Motec gives the output of the switch between the 4 positions as shown below.



So in the haltech, I’ve assigned Rotary Trim Module 1 function to AVI 10 and setup the calibration as shown. This being the case...

Position 1, trim switch gives 1.0V, ECU reads <1.25V and Outputs Position 1

Position 2, trim switch gives 2.0V, ECU reads 1.25V<2.0V<3.25V and outputs Position 2

Position 3, trim switch gives 3.0V, ECU reads 2.25<3.0V<3.75V and outputs Position 3

Position 4, trim switch gives 4.0V, ECU reads 3.25<4.0Vand outputs Position 4





After enabling the closed loop boost under boost control function, the first thing to do will be setting up the target boost pressure for each switch position. When tuning a flex fuel car across a large boost range, I always setup the wastegate duty table as a function of flex fuel to ensure you don’t accidentally get a tank of low content E and full send it.

This can be done a few ways, but I setup the Y axis as flex fuel content, X for RPM, and Z (4D) for the rotary trim module switch position. So…for position 1 the target boost level will be 18 PSI, but only when ethanol content is 50% or greater. If less than 50%, it will automatically run the 14 PSI spring pressure without any input from the controller.


Same thing here on position 2, boost target is 23 PSI, but only when ethanol content is 50% or greater.



Since I’ve tuned the car to 23 PSI on E55-60 without any issues, I’m fine with it running that boost level as long as the E content is above 50, but for higher boost I set it up to ensure I’ve got a tank of E75+. So now for position 3 the target for E75+ is 28 PSI, E50-E74 is 23 PSI, and again below E50 will default to the spring pressure.



And lastly, position 4 table to big boy boost level for a target of 32 PSI for E75+, 23 PSI for E50-E74, and as always below E50 is spring pressure.



I like doing it this way as it give you the flexibility to choose between the 4 boost levels using the trim switch, but at the same time there is a flex fuel fail safe that ensures that if you don’t have enough ethanol in the tank that you will always be at a safe boost level no matter what position the trim **** is turned to.

Next, the Closed Loop Base Duty Cycle table is setup like shown below using the known duty for 18 and 23 PSI and slightly scaling up for the higher 28 and 32 PSI boost levels.



Once the switch is wired and this is all setup in the file, it can be loaded to the ECU and checked against the position of the switch. With E50+ in the tank and the car powered on, you should see the little blue circle move from the 18/23/28/32 PSI target boost rows with each position of the trim ****.

Next, once there is E75+ in the tank and it’s time to turn up the boost from 23PSI, I’ll revert to open loop and increase the duty cycle 1% at a time tuning the fuel required for each higher increment as I go and saving the final open loop duty cycle baseline for the 28 and 32 PSI levels once finalized. With those values then known, it will be swapped back to closed loop and the known 28 and 32 PSI baseline duty will be added to the closed loop base duty cycle table.

My goal is to get everything up to this point sorted in the next few weeks. Given how fast the car is now on 23 PSI the limiting factor on how long this will take will be whether or not I end up needing a dyno to safely tune the higher boost levels. I’ve got to drive ~10 miles to the major 6 lane highway from my house on a curvy 2 lane road to be able to open it up and make a real pull as going WOT in 3rd or higher without ALOT of straight flat runway is extremely sketchy and borderline asking to die.

I expect I’ll have to play with the closed loop boost control offset point and PID values a good bit to obtain the right balance for accurate control when I start using the closed loop feature. Even if the car is tuned up to the 32 PSI in open loop, I’ll start on the closed loop control targeting the 18 PSI target so it will be at a lower safe boost level in the event I overshoot or the 4 port proves too sensitive for accurate closed loop control.

More to come….

Skeese

 

I ended up spending a good bit of time in the car yesterday and spent most of it fighting the sensitivity of the 4 port mac valve, but I think I came to a solution that doesn't require me to change to a 3 port...at least yet.

On the first pull I hit overboost cut in 3rd gear at 28 PSI at 6300 rpms running the open loop duty at the same 25% that had been consistently 23 PSI just a few days before. I made no changes and made a second pull in 3rd gear a few miles down the road it ran only 20 PSI all the way to 9500 rpms at the SAME duty cycle. I pulled over and checked the data which showed that the only difference between the two was the IAT temp was 90F on the first 28 PSI pull and it was 138F on the second that only net 20 PSI all the way up. To me this was such a large gap I thought something had to be wrong with the setup on the valve so I turned around and returned home think on it more.

Searching online I found the frequency to duty cycle range for the 4 port I'm using, and realized I was running the controller at the haltech base value of 33hz meaning the full scale of the valve was slightly higher than 23-46% duty cycle.

This made perfect sense in hindsight as the car would sit on the 14 PSI spring pressure up to 22% duty as the controller did NOTHING below that but by 23% it wants would run to ~18 PSI then by 25% it would run between the 20-29 PSI range I saw yesterday morning in those first two pulls. As the full range of the valve is contained within 23-46% duty cycle, it is no wonder why it was so overly response and unpredictable with each 1% duty increment.



Dropping the frequency to 15hz and the duty cycle to 13% to reflect the low end of the linear range given of the valve, the car ran ~17 PSI all the way up across 2 identical 3rd gear pulls, 14% made ~19 PSI all the way up twice in 3rd, then 15% duty net ~21 PSI consistently across one 2nd gear and two 3rd gear pulls.

This being the case, it seems that running the valve at 15hz nets a consistent increase of 2 PSI per 1% duty added above 13%. I'm not sure yet if the 15hz operation is going to be too picky for tight closed loop control, but I'm going to make some more runs today and see if this trend holds with more duty and will go from there. If it proves to be tight enough to consistently hit and hold each of these targets I'll make the change to closed loop, setup some conservative PID values and drop back to the 17 and 19 PSI targets and see how it handles the boost.

Skeese

 

Pulls from the weekend now have it running ~25-25.5 PSI with the controller outputting 18% duty cycle at 15 hz which is consistent with the increases from the lower duty cycle pulls from before. Will be swapping over to closed loop this week and beginning work on 1) getting spool lower and 2) having the controller hit and hold each target level consistently across all gears.

More to come...

Skeese

 

Awesome, I'll be following your lead as I'm looking to switch from 3 to 4 port in closed loop also.

 

Honestly, I'd stick with the 3 port after my experience with the 4 port and how sensitive it is. It is alot better running at the lower 15 Hz but it still isn't nearly as accurate as I'd like. You may could get there if you combine the low frequency with a really light spring (3-5 PSI-ish) to try and expand the whole control band to run over a larger range.

I don't really care to drop my spring pressure below the 14 PSI that is in there now, so I'm just going to make the swap to a 3 port or one of those fancy new fueltech dual valve setups. Regardless of what setup I end up with, the goal and setup for the selectable 5 boost levels via trim **** all being based on ethanol content is still the plan.

I haven't had an opportunity to drive or tune the car further since my last post, hopefully soon.

Skeese

 

I am nowhere close to your power levels, but my experience with the 4port solenoids is crazy, boost consistency is not there at all. From one pull to another, boost could fluctuate a couple of psi. Have you considered dual 3 port solenoids? I might end up going with the simplest setup, a larger wastegate spring, and a 3 port solenoid... SMH, cause I really wanted to run a "large range" of boost pressure.

 

Its been a while since I updated here, but I've been making some progress and now maxed out the haltech elite's internal 3 bar MAP sensor that flatlines at 29 PSI in my logs. I'm still running the 4 port MAC and have been able to achieve consistent repeatable results when running it at 15Hz. I picked up a 5 bar map sensor and will drilling and tapping a new map port on the UIM to accept the 1/4 NPT threads on the 5 bar, then I'll run it as a generic pressure sensor and log data from both the internal 3 bar and this new 5 bar in parallel to verify the new sensor's calibration is correct.

Here is a clipping from the last log where you can see I flatlined the MAP sensor at 29 PSI at 6200 and held it through there until letting out at 9200 rpm using 23% duty on the mac. I had the overboost set at 32 PSI, so I know that with the MAP reading pegged at 29 PSI the car saw somewhere between 29-31 PSI. The AFR held right on target through the region where the MAP was flatlined, so I'm confident the tune is on point through there. I've dialed back the duty to 22.5% which I'm expecting will hit and hold the target 29 PSI and plan to keep the boost here for a while while I sort out the closed loop and switchable boost control setup.



Going back through my logs driving the mac at 15 Hz:
13-16% duty cycle = 20 PSI
18.5% duty cycle = 22 PSI
20% duty cycle = 25 PSI
22% duty cycle = 27 PSI
23% duty cycle = 29-30 PSI



I'm going to drop back down to targeting 22 PSI and swap over to closed loop to start working out the settings needed for the closed loop function to spool as quickly as possible to the target level and hold it once reached. More to come soon.

I need to get a gopro so I can make some actual pull videos as there is no way to do it holding a cell phone and driving at this level. This damn thing really screams at 29+ PSI. As much as I want to turn up further, it is honestly pretty useless to do so for street use.

Skeese