Tuning my 347 with rear mount turbo

[Dono]

Quick update.....Slow going.
I have not put much energy in to this, and have concentrated on drive-ability.

For some reason, finding the sweet spot on the crank fuel and air so that my truck starts easily at all temps is proving to be a challenge.

I do want to say thanks to @4pointslow for his help in tuning.
He turned me on to histograms, and creating user defined pids.

Here's a link to 4pointslow's description on user defined pids.
https://www.explorerforum.com/forums/index.php?threads/my-2004-explorer.441838/page-10#post-3567031


One of the histograms he created is to tune by ltft and let the livelink histogram do the work. It works amazingly well.
Tonight, over a beer, I created a user defined pid adding my two long term fuel trims together, and dividing by 2.
Then, I changed my histogram to use the user defined pid of LTFT so that it allows me to tune out of wot using an average of the 2 banks. Its pretty awesome to be able to take a couple things someone teaches you and marry them both together. Thanks 4pointslow!!

Ok, I give.....Once I have created the user defined pids, how do I save them in the configuration so they automatically show up the next time I open up my datalog config?

 

[4pointslow]

I have tried to save them in a livelink configuration file and it seemed to work until you change something like adding 1 more pid etc.
Adding 1 more pid can change the order and mess up the calculation. For now I just add the calculated pids after the datalog is finished.
If you think about it, having to calculate something on the fly while recording lots of information from all the other pids would slow the computer down anyway.
Going racing today, wish me luck!

 

[Dono]

Thanks, I see what you mean on saving user defined pids now. Also, I guess even a simple calculation could cause response issues. I'll play with it and see if my old core2duo laptop can handle it.

Were all looking forward to your results. Vroomzoomboom is chasing your times. He's itching to get back out to the track, although we are shut down here for the year.

 

[4pointslow]

Today was around 70 degrees F, not cold like it will be soon in months to come.
My best time was 13.40 today.
When it gets down to 40-50 degrees F, I should hit 12's again.
I am working the AM shift and will have off on a few Friday nights so I might have a chance to get to the track again if it is colder out.
We will see.

 

[Dono]

Nothing wrong with that time. Have you run your numbers thru the wallace calculator for estimated horsepower? Considering your weight, it has to be right up there.

Hopefully next year I can get out with Tim and have neither one of us blow up. I'm behind Tim on the mechanical stuff, and wot testing. I'm not sure that I have done a wot pull to tune off of at all this year. I know. I'm terrible. Now its almost time to put winter tires on.

I removed all that under rated electrical tape with the upper intake off, and forgot to re-torque my lower intake while I was right there. I wasn't completely happy with my self.
No problem though, a 'misfire cyl 8 code popped up', and off came the upper again. My wiring harness had had different injector connectors spliced in by a previous owner, and they had given me trouble in the past. #8 injector connector was clicked in, but intermittent when checking with my multimeter. I had already replaced 5 of 8 connectors, and now this makes 6, lol. And yes, I re-torqued my lower intake this time, and it was about a quarter turn loose on each bolt.
On the tuning by ltft, it has really helped me. My truck is so finicky on its starting that I'd think I have it down, and a few days later it would not start any more without touching the gas. I'm guessing (hoping) that with the ltft's pulling a couple percent in that area that this was my issue. Even though after a good drive, I would tune my stft's for idle, my computer see's things slightly differently. I do think tuning via ltft will give me a single target to fix my starting. I'll know in a few days.

 

[vroomzoomboom]

I'm behind Tim on the mechanical stuff, and wot testing. I'm not sure that I have done a wot pull to tune off of at all this year.

i dont think that its your behind, i think that its more like i care less if i scatter the motor all over the road then you do, because its not my daily driver lol

 

[Dono]

lol.
I did some wot tonight coming home from work...hesitation, cough, sputter, etc.

I made some maf transfer adjustments under higher load, and I'll see what I get. With a sputtering vehicle, its hard to get a a/f reading that's actually usable. Seemed to be about 5000rpm. More to come on this.

 

[Dono]

I'm pretty sure this is my issue. I did manage to make it worse. lol
This curve worked with tho old turbo.
I'm going to try taking that out by inserting the values from my base tune from just before that spot and re-curving the inserted upper curve to match what I had already done down low. I only needed to multiply the base tune curve by 1.05, so lets see what the next pull produces.

after:

 

[Dono]

Re-curve was a success. Truck pulled clean. I have not been able to pull past 5000rpm yet due to road conditions.
The transition under wot to the richer a/f ratio still leaves a hump in the curve though.
My question at this point is, do I re-normalize the maf points so that I get a few more spots to tune the a/f ratio as it transitions from stoich? Maybe if I moved the data point just below the hump up a bit, and the data point above the hump down slightly in maf add counts? That might help the resolution thru transition?
What do you think 4pointslow?

 

[4pointslow]

That would be a perfect example of why you might change the datapoints on the left, to fix that trouble spot.
That first picture almost looked like the one dot was lower than the previous dot. Make sure the numbers are all going bigger as they go up the MAF transfer function like in the last picture.
Honestly that little ramp doesn't look so bad and may not need to be changed, unless you are still experiencing a running problem there.
Test it and see how it feels both ways.

 

[Dono]

I sent my tune to Don Lasota to have a look at while I have my motor lifted waiting on getting my header fixed.

Don said my timing is to low at wot, creating to much heat. Not the reason my header cracked as it was clearly crimped from the exhaust shop pushing the pipe away from the frame.

Anyway, I think the smartest thing to do is drop the boost to around 6 to 8lbs and up the timing in the short term. Long term, I'll be working on installing my water/meth kit.

 

[Dono]

ok, back at it.
New motor, new hope.
The tuning seems like it has not changed much. Here's a pic showing the SCT histogram average for the long term fuel trims.
I really prerfer tuning the maf this way, as adaptive learning will just adjust its self to this anyway. Thanks 4pointslow. Love it.

Also, you see the low maf count (124)? Im wondering if this is a long time issue I was having. I would get my crank fuel pulse width so my truck would start, then after a few days of driving with adaptive learning on I'd have a hard start issue. This makes sense to me, so I have multiplied 124 and 108 by 1.06. Lets see if this fixes the weird hard start issue.
I also tried lowering my idle in park and drive just a hair to see if I can dial in the bottom of the maf curve better.

 

[Dono]

I'm lovin the J&S Vampire knock control unit.

First, I thought it wasn't working as it would light off really bad at lower loads and then throw the check engine light on with multiple codes about coil failure.

I logged where this was happening (Around .45 load and 2500rpm) and made some timing adjustments.

Wen't for another drive, and problem gone. I couldn't hear any detonation, but obviously it was there before making the timing change. I pulled about 5 degree's timing out and normalized the surrounding area using sophisticated wild assed guessing.
The vampire unit was obviously trying to pull so much timing it gave the pcm fits. lol No more check engine light.

This will be time consuming tuning as you log, make note of where the J&S gauge lit off, and review the log at that point in time. Look at rpm and load, then reduce timing in that area. Upload new tune. Rinse, and repeat.

@4pointslow, do you know of any way to make easy note of a particular spot when logging in livelink? It would be awesome if I could just reach over and hit the space bar while driving and have my log marked for review later.

Its interesting that I could not hear detonation and actually had factory timing in that area of the tune. I guess the turbo is pushing more air earlier than I would have expected at lower loads.

The knock detector is at least as important as a wideband o2 in my opinion. It's a tuning tool that should be installed in any forced induction application (Even temporarily for a few weeks while tuning is performed). I'm trying to get a couple of degree's below where the knock sensor is set off.
I may try to create another tune later that under load it just barely registers knock and hopefully is only retarding a couple of cylinders. This would be max power.

Just finding that lower load detonation issue makes it all worth it.

Once my V-band clamp comes in, I'll get my wastegate fixed up so I can control boost and continue WOT tuning. Drive-ability of my truck is starting to near a factory vehicle.

 

[CDW6212R]

Great to hear, that sounds like a positive tuning method and tool.

 

[4pointslow]

Hook up the output from the Vampire to the analog input to the X3or 4. That way you will see the rpm and load where the detonation took place in the datalog.

It is a shame that some vehicles did not come with a knock sensor from the factory. My 1998 did and it has proven to be one of the most necessary things along with the wideband O2 setup.
My 2004 has factory knock sensor but it doesn't datalog correctly with the datalogging software, I hope that SCT/Derive gets it working soon.

 

[Dono]

Here's me being lazy.

When I take my ohm meter, do you know if I'm looking for the wire that's on the tip of the 3.5mm plug? I'm pretty sure that's standard.

 

[4pointslow]

There is a dedicated out put for datalogging, I will look it up......
Edit: From one page it says the tip is 12 volts, don't use that one.
The signal is the "ring" area or middle contact part of the 1/8 jack.

 

[Dono]

Appreciated. for some reason, I struggle with finding documentation on this.

Edit: Done. An exercise in frustration. I should have bought a new cable and worked with it. The wires inside that cable are crazy small.

I had 3 wires to pick from. ! was red, so I made the assumption that was 12v. The only way to get a reading with my meter was to scrape a wire with a knife first. 50/50 chance. wrong one. ****. a slight bit of electrical tape, and back in the sheathing. I opened up a different area of the cable and found the right wire. Now to solder to it. lol. Painful. Success though, my ohm meter tells me everything is good, with no shorts either. I need a beer.

I tried removing vacuum to the vampire and rapping on the knock sensor. I couldn't get a reading on livelink, but I couldn't see the vampire gauge either....So I have no idea if its working. Seat time logging will tell me. Back to that beer.

 

[4pointslow]

If I remember correctly the Vampire will not retard knock under a certain RPM, 1700 I think.
That may be why it doesn't prove out at Idle?
Beer sounds good.

 

[Dono]

4pointslow, absolutely correct. I forgot to wedge a board in between the seat and peddle to get rpm's up.

success on the date log. The dark blue is knock. Load .65 at 3500 rpm, and the big one was at a load of .76 at 4000 rpm.
There we have it! The guessing is gone. I see a histogram being made for this coming shortly. Did I mention I love histograms? Again, thanks to 4pointslow.