5.4L Big Cam Tuning Experience

[Milanski]

My fellow PCMTEC tuners, I thought I'd take some time to write about my tuning experience regarding my FPV BF GT 290.  I share this topic for public comment and also to see if anyone can gain something from my experience and possibly add things I may have missed.

I bought the GT (my second one!) early November 2020, 30,000km done on a newby engine and the previous owner wanted big cams.  I also point out the big cams - not needed for stock engines and unless you are doing 600+HP. They sound good at idle but have caused me headaches.  

I drove the car for a while.  Idling at 1000RPM doesn't help the ZF nor the brakes, it just wants to launch from standstill.  25-30L/100 not good for long distance driving.  I found a modified tune with many changes in all sort of areas i don't understand.  Changes to air flow at idle, which is a requirement of big cams to make it idle properly.  3 degrees of spark mods (not sure why it had to be 3) and few others.  Lots of MIL lights removed. 

The engine would never get to closed loop mode, LTFT were more than 25% at idle (negative) trying to pull fuel out, tail pipes were like a MACK truck and all 4 O2 sensors were stuffed because of the above.  In Open loop (power enrichment) the vehicle is exciting and couldn't fault it - so on a dyno I think they had this area sorted.  It was only at idle and low RPM that bothered me.  

To hand the car over to a tune house would take weeks to tune, I was quoted $5k to get the tune done.  I understand that its not about a few dyno runs, its going to take hours and hours on the dyno to make a change.  Let it settle for about 80km then make another change and so on.  So I understand why the quotes were significant.  Ford spend millions of dollars getting the tune in and its impossible for tune houses to do something like this in 5 days.  I decided to give it a go and I set out over the next few months - road tuning !.  As your reading this remember, that this tuning experience is to get idle into closed loop mode and good LTFT so that i'm not filling up every second day and gassing my garage!  The joys of big cams. (another discussion!)  

Requirements - new O2 sensors in the front.  Turn off rear post cat sensors. Install wideband AFR.  Got one of those PLX devices - for my job they seem to be OK.  I'm just watching for lean > 18 on idle.  In open loop there is plenty of fuel being thrown in and AFR is at 12 or less. 

Small recap before we begin:  LFTF 25%+ negative, IDLE 1000RPM. Lots of teary eyes in the garage after prolonged idle periods - had to buy a fan for the garage! Fuel burn not good!  

I set up my scanner with LFTF vs Load.  remember that load is a calculated load and not the same as the load used in VE tables.  For my interest the load was low - idle or when coming off the foot pedal.  The ecu went into closed loop mode, got to -25% LTFT and then into open loop.  Car splattered coughed during the closed loop but ran fine in open loop. 

For whatever reason the intercept at map0 table changed AUF0056, I put that back to stock and went from there.  What was clear that fuel needed to be pulled out at idle so I could at least get into closed loop mode.  I measured my AFR at various loads at low rpm.  This was acheieved by foot on brake and driving around my back streets for a few hours.  May I point out you need lots of data and at x/y points to make good judgment calls.  

First step was to research - looked at plenty of GT500 tune files and other big cam engines. Not much information sharing out there. Also I forgot to mention my cams - they are crow cams - engine built invoice states it - pictures shows the boxes, cams on the table show it,  but no part numbers (mild or aggressive??? probably aggressive).  Photos of the build show Crow.  So I have no idea whats in the the car or their overlap, if they are dot-dot then they are to fod spec +/-7deg!.  The lady I bought it from didn't have that info either.  Rang the engine builder in QLD to find they have closed down (bit nervous!).  I digress again.  Most tune files for big cams have a MAP slopes very high at idle.  So my fist step was to turn off LTFT (so it wont go to closed loop) bump up the MAP slopes to get it leaner at low rpm.  I did this until I saw at high load, low rpm my AFR being on the lean side.  Eventually I ended up with a MAP slope as:

Remember I'm not touching anything higher than 1500rpm.

Put the LTFT switches back in and we had closed loop for the first time (i think for a very long time).  

The Y-axis is Load.  This took a bit of time to find in the VCM Scanner software (my apologies Roland and PCMTEC, I'm using the data logger more and more now but at the time I knew VCM like the back of my hand. I hope you dont mind but there are a few more data plots not PCMTEC).

I cleaned by exhaust tips back in november and to date they are clean as a whistle!

 

What happened next was very interesting.  Whilst idling around - moving at low speed. I touch the accelerator and the car would continue to accelerate on its own.  I traced this back to the AUF0030  Aircharge in lb/min at idle speed.  This is an imporatnt table, its used be the ecu to control idle based on theoretical numbers from this table.  Higher number we take off automatically!  Weeks went by whilst adjustments were made to this table:

 AUF0032 are lower numbers

 

I have got the car to idle 780rpm using the above tables and it took a long time to get it to this point.  The car sounds lumpy and the brakes and ZF box are thanking me for it.  

I'm not going to show graphs for every point I changed, if you like those contact me.  Needless to say, the spark, STFT, LTFT load graphs were untidy!  I am suprised how well the ford ecu managed to keep the car running. 

I needed to do work around the low load idle without affecting the high load low rpm areas.  Not much information out there reagrding where to go to next then I found AUF16632 Multiplier for the slope of MAP per Aircharge BF.  Interesting table, it allowed me to lean out the fuel at specific loads vs rpm.  All of Dec and Jan were spent bumping the values at low rpm to lean out the fuel.  With my change previous I was now running in closed loop mode and I was simply using my narrow band O2 plots to work out if I was going lean or too rich.  Closed loop tuning is what some call it.  My Wideband was there just in case.  

 

 

I have some more work to do around <20% load, this is when you release the accelerator or are coming to a stop at the lights and fuel is being pulled.  I havent worked out how to fix that area yet, but I guess getting to this point and being able to get 400km out of 60L is making my back pocket happier.  I have a good idle at 780RPM, sound lumpy and angry and yet purr's along at 1200rpm 60km/hr making for a good road trip with the family.

For those that like data:

I think I'm 95% of the way there, my ECU is working fine with <5% LTFT most of the time and if anyone has other suggestions and comments its probably where I will leave it.

There are more areas to look at with Big Cams not mentioned above,   like ECT adders, Spark Lambda, MBT but thats for another day.  

My final table of LTFT is shown as:

with the following counts:

I don't spend too much time at the high load - high rpm unless I'm at the track.  There is only so much one can do on a public road.  

I found very little open forums on how people have dealt with their issues, these are my experiences and I hope someone finds it useful. There is so much I can write about my experience but I hope this is a good summary for others to try out.   Let me know your thoughts.

Kind regards

Milan 

 

 

 

[Roland@pcmtec]

Could save yourself a lot of time if you use this feature! Sounds like you have it pretty close now anyway. The reason there is hardly any information on tuning cams is because hardly anyone actually knows how to do it properly in these cars. The main reason as you found out is a total of 0% of customers are willing to pay the tuner for the time required to do it properly. 

 

[Milanski]

Wow, that is very neat.  I did do anymore searches for large camshaft tuning since early december, kicking myself you had a write up in Jan.  This is very neat.  My Map slopes are no longer stock and hence if change to the custom OS as described in your write up will I lose my values? will they go back to stock?  Are there other tables that are affected by the change?  

[Roland@pcmtec]

16 hours ago, Milanski said:

Wow, that is very neat.  I did do anymore searches for large camshaft tuning since early december, kicking myself you had a write up in Jan.  This is very neat.  My Map slopes are no longer stock and hence if change to the custom OS as described in your write up will I lose my values? will they go back to stock?  Are there other tables that are affected by the change?  

This functionality has been available for over 2 years now!

FYI your current values may not make sense currently as that table you are modifying is throttle position (inferred load) not actual load which would explain why you are having funny results. Tuning correction is basically an Alpha N correction table.

All your other values will remain, you can always copy them back.

[Milanski]

I searched for cam setups, couldn't find much around.  Anyway, almost there.  

LTFT-B1

Not sure what Alpha N means.  But the table is getting me where I need it to go.  

Now I need to get my head around bank 2 which is way out.

LTFT-B2

Initial thoughts are the fuel delivery at the rails is over bank 2 may be higher pressure sitting up behind the injectors versus bank which runs round the plenum. Need to reach out for guidance.  

[Roland@pcmtec]

Alpha N means throttle vs rpm. Eg very old school tuning method. It is not load on the y axis so it absolutely does not line up with the load from the logger. 

It is inferred load calculated from the failed maf table which has throttle position vs rpm to generic a fake inferred load. 

Regarding bank2 being out it is probably a blocked rail/injectors. Seen old engines sitting around for months rusting up and then partially blocking injectors before. I would pull it all off and flow bench/clean the injectors/rail as these cars are getting on now. 

[Milanski]

Im putting it down to ford dot-to-dot and long chains causing some cams to be out by as much as 7% from some builds being reported.  Its a new engine about 30,000km.  New injectors.  When I get time I will swap the injectors from bank 1 to bank 2 and see if the problem had shifted.

Regarding load, reading the write up on speed denisty tuning the faulted MAF sensor table is a small part of the overall calc, which is why I spent a lot of time logging the inferred load and actual load and getting them within 10%.  But now that you pointed me into the Custom OS area, I will look at that also if things get out of hand.  I don't know how I missed your write up if its been there for 2 years.  I raised some questions relating to big cams months ago in the speed density calculations as well.

 

[Roland@pcmtec]

3 hours ago, Milanski said:

Im putting it down to ford dot-to-dot and long chains causing some cams to be out by as much as 7% from some builds being reported.  Its a new engine about 30,000km.  New injectors.  When I get time I will swap the injectors from bank 1 to bank 2 and see if the problem had shifted.

Regarding load, reading the write up on speed denisty tuning the faulted MAF sensor table is a small part of the overall calc, which is why I spent a lot of time logging the inferred load and actual load and getting them within 10%.  But now that you pointed me into the Custom OS area, I will look at that also if things get out of hand.  I don't know how I missed your write up if its been there for 2 years.  I raised some questions relating to big cams months ago in the speed density calculations as well.

 

Could the cam timing have skipped a tooth?

That is great news that you have the inferred load that close, almost no one bothers to do that. In that case it will probably do what you need. The fact this engine is not turbo and has no VCT really helps as well. The approach you are taking would not work on a VCT turbo 6 as the load can change drastically at the same throttle position whilst the turbo spools up, the NA v8 will have almost no transient change in VE for a static throttle position.

[john.z]

On 2/26/2021 at 1:13 PM, Milanski said:

Im putting it down to ford dot-to-dot and long chains causing some cams to be out by as much as 7% from some builds being reported.  Its a new engine about 30,000km.  New injectors.  When I get time I will swap the injectors from bank 1 to bank 2 and see if the problem had shifted.

I agree with this, ive ruled out everything else- fuel pressure, injectors, intake/ exhaust gasket leaks, exhaust cracks or pinholes, compression, leak down, spark, wiring etc. the only thing it could be is cam timing