Starting to Learn on 95 G30 5.7 for Towing

I am also only comparing to your comparitive numbers in your table form.

The TBI head and cam engine with rockers and headers made over 200 rwhp from 3,600-4,600 rpm. Kitch also saw very similar results. You both have standard 12cc dished pistons, mine were the 18cc dished but full disclosure here I had a 454 throttle body and 1" open center spacer under it with an edelbrock 3704 intake manifold that had been bored to 2".

The Express van hit 200 rwhp at 3,450 rpm and even with failing valve springs did not drop below 200 rwhp until over 5,200 rpm and even then it only did that because of horrendous valve float. I had never seen an engine fall off that badly before I experienced that. I have never seen such a mild cammed engine gain torque at peak, have the peak torque rpm drop 200 rpm and gain 27 hp with a 500 rpm higher peak from changing valve springs either.

OK sounds good. Next step is I will try pulling the AC compressor and removing the valve covers for less lifter preload and getting the AFR to drop about 3/4 of a point at WOT. Thank you for the ideas. John from 1320 Electronics sent me a file to try on the ALDL Scan for Android, so a big thanks to him as well.

One other thing I forgot to mention. I would try your timing increments in 2° increments.

I hooked in AEM X-series wideband this weekend. I am attaching the modified adx incase someone wants to use the newer series of wideband. The newer AEM wideband voltage chart does not list a 0v or 5v number, but using their formula you can generate them and then follow the instructions listed on this forum to come up with the math for the adx. I have found a few small things, but have not put the van back on the dyno yet. For some strange reason resetting the initial timing advance back to zero in the chip and leaving the altitude adjustment off, mismatch between what is logged and what is on the table still there on some cells, but most are really close. I am not sure if I will keep it at zero, just looking at cause and effect. I also goofed on the base injector flow rate at 19psi, leaning out all the tables by a little bit with that mistake.

I might have found a little clue for why the numbers in the data logging do not match up with the data log. The od TP5 v251XDF supposedly compensates for initial advance changes. The adx does not, reasonable since there is probably no communication between the two files. "Change last number(offset) to initial advance of engine, for correct display." Still adding a flat 9.8 to the log files do not exactly correlate see the pdf info posted on post #91 so "correct" may be wishful thinking. So far my OTC scanner has come the closest to recording the most believable timing numbers, especially when the main timing value goes to retard when there is a low total value on the main table and a high initial advance. If you see a crazy number in $od logs it's probably just the retarded areas, but the fact that it does that makes me think other areas may not be quite up to par either. One thing I am wondering about is the log file for the xdf. If the main timing table auto compensates for any initial advance on the chip, try changing it and only that from 0 to whatever you choose. Now look at the log file. You will see the change log made a note for the initial advance change but no timing table changes were noted. If I am understanding how this particular xdf works it should have knocked the value of the initial advance off and rewrote the value to the bin and logged the change too?

With emulation, it's not possible to make changes while running, correct? I had chip emulation set up on the dyno, but the one time I tried to push the change to the chip with the van running nothing changed then something happened and the file was corrupted the next time I tried to start up and would not start at all. It would be handy if that would work for changes on the dyno especially for part throttle timing.

I put off pulling the valve covers to re-check lifter preload hot and running instead of the static set. The one side is a huge pain in the van, but most everything else so far checks out or is fairly close to a reasonable value. So I guess that is today's project.

In the past I've been able to make live changes to my idle fuel mixtures and spark while running in emulation.
I've also had a problem with file corruption after emulation, it would end up in limp home mode until I could get a file back into it.

Quote:

---

Originally Posted by donf

I might have found a little clue for why the numbers in the data logging do not match up with the data log. The od TP5 v251XDF supposedly compensates for initial advance changes. The adx does not, reasonable since there is probably no communication between the two files. "Change last number(offset) to initial advance of engine, for correct display." Still adding a flat 9.8 to the log files do not exactly correlate see the pdf info posted on post #91 so "correct" may be wishful thinking. So far my OTC scanner has come the closest to recording the most believable timing numbers, especially when the main timing value goes to retard when there is a low total value on the main table and a high initial advance. If you see a crazy number in $od logs it's probably just the retarded areas, but the fact that it does that makes me think other areas may not be quite up to par either. One thing I am wondering about is the log file for the xdf. If the main timing table auto compensates for any initial advance on the chip, try changing it and only that from 0 to whatever you choose. Now look at the log file. You will see the change log made a note for the initial advance change but no timing table changes were noted. If I am understanding how this particular xdf works it should have knocked the value of the initial advance off and rewrote the value to the bin and logged the change too?

With emulation, it's not possible to make changes while running, correct? I had chip emulation set up on the dyno, but the one time I tried to push the change to the chip with the van running nothing changed then something happened and the file was corrupted the next time I tried to start up and would not start at all. It would be handy if that would work for changes on the dyno especially for part throttle timing.

I put off pulling the valve covers to re-check lifter preload hot and running instead of the static set. The one side is a huge pain in the van, but most everything else so far checks out or is fairly close to a reasonable value. So I guess that is today's project.

---

With a hot engine I would not go more than 1/4 turn preload.

I've set the lifters pre-load both static and running at different times. You're right it's a pain in a van, one side has the AC pump in the way and the other has the engine harness running over the top.
I do 1/2 a turn preload cold.
If I'm doing all the lifters then I just use the when X is rocking set Y method. When I'm just checking one bank (and I can't see the rockers under the other cover) I use an old dizzy cap that I hole-sawed the centre out of so I could see the rotor. So for example when the rotor lined up with where the lead for cylinder number one was, I'd set both the lifters for number one. I'd only use this method on an engine that's been running because I know the dizzy is installed correctly.

Quote:

---

Originally Posted by Kitch

I've set the lifters pre-load both static and running at different times. You're right it's a pain in a van, one side has the AC pump in the way and the other has the engine harness running over the top.
I do 1/2 a turn preload cold.
If I'm doing all the lifters then I just use the when X is rocking set Y method. When I'm just checking one bank (and I can't see the rockers under the other cover) I use an old dizzy cap that I hole-sawed the centre out of so I could see the rotor. So for example when the rotor lined up with where the lead for cylinder number one was, I'd set both the lifters for number one. I'd only use this method on an engine that's been running because I know the dizzy is installed correctly.

---

Anything over 1/4 warm and running is giving away torque as well as RPM. Any non running adjustment is also not accurate.

FWIW, this time around my L31 is getting Comp reduced travel roller lifters and 1/4 turn preload. They are designed in a way not to collapse more than about 0.020" total. So 1/4 turn puts them right in the middle of their travel. The lifter can neither collapse or pump up enough to cause any issues. Running 150# at the seat and 375# open spring pressure.

The problem is usually weak valve springs that allow the valves to bounce on closing, introducing slack in the system causing the lifters to pump up and hold the valve open. With 1/4 turn preload your lifters can never pump up.

Edited this to save anyone from confusion. See below.

Quote:

---

Originally Posted by Fast355

The problem is usually weak valve springs that allow the valves to bounce on closing

---

https://www.summitracing.com/parts/s...1/applications are the springs I am running. The cruddy import vortec heads Jegs sent had pathetic valve train. So I upgraded to be safe, not as strong as what you have but stiffer than the stock specs for vortec springs and I have the fuel cut off set at 4600. I think stock was 4200. Installed spring height was measured and set, a few springs did require small shims, more poor machine work. I got side tracked and ran out of time so the valve covers stayed on today.

I am editing this so not to confuse anyone. Mostly I was just confusing myself by editing things with little explanation of how it works. It was a good lesson though and I didnt hurt anything.
The OD adx file in tuner pro and others shows crazy numbers in retard. It's not a huge deal but if you see it that's just what it is. The reason 0 in the altitude bias on the bin helps the crazy numbers is the bias is a subtraction so zeroing that one part out while leaving the matching table actually raises the timing. The ADX also doesn't show the initial spark advance if it's anything above 0 with out editing. So if you have some missing timing in the logs look there first. The ecm does look at the initial advance and is supposed to subtract that from the main spark table. I found a few older photo copied papers that explained the biases vs the tables a bit better. Still there is no where near the explanations that you find on the later GM stuff but it helps.
This pdf I found online was too big to post here. www.gmcmi.com/wp-content/uploads/2014/05/How-to-Tune-a-GM-EFI-System.pdf The other one attached also explains the timing bias a little too.

Quote:

---

Originally Posted by Fast355

To know if the base timing is accounted for you would need to look at the equation for the spark advance in the .ads file.

I really need to pull the $OD Hack and look at it really well. Something tells me that zero'ing out the altitude spark bias is what is causing your values to be off so far. Where there is a bias table their is an offset value.

There are also other spark advance tables in play that would keep you from matching the Main Spark advance table. Abuse Mode spark retard, Coolant Temperature compensation, Catalyst Overheat spark, etc come to mind.

---

^^^ I was onto that bias table still having a value other than zero and zeroing the table bias as causing an issue.

You can easily edit the ADX to include the initial timing value in the datalogged value.

From what I remember Initial advance is actually subtracted out of the final spark advance values right before the PCM checks that the values are within the Min/Max values and uses it for EST calculations.

Whoever wrote that article straight up plagerized some of my own tuning information I put on a forum years ago not that I really care.

Post #7 in this thread.
https://www.thirdgen.org/forums/tbi/...i-buildup.html

Fair enough, there is a LOT of contradictory information here. So care has to be taken. Even some of my post would probably mislead, so its a learning process. I tried to edit out the most confusing parts. I did try the Lt timing table from the 1995 Caddy with the towing package. It seemed to increase responsiveness a bit but as soon as I hooked the enclosed trailer to the van and started pushing it the knock counts started to rise in several areas and that was at the winter air temps, it would be worse in the summer. I went back to the mostly stock vortec table for now that I had in the beginning. I will work on the table once the dyno gets pulled back out. The weekend weather has been crud for the last few weekends so I am not pulling the dyno out in that. I am pretty happy with the 40hp and the 50ftlbs of torque over the base line and it does pull the trailer better. Comparing different dyno readings is just speculation. Like I mentioned before, Dyno Dynamics has tested other makes of dyno's and said they sometimes show more than true crank hp at the rear wheels. I don't know how others owners calibrate theirs, but mine is super easy to check for accuracy. I checked it again, the dyno is within 1 percent of perfect, slightly low. 1996 literature said the l31 is 245hp and 325 tq with a much better intake system, flow wise and a slightly worse cam and and those numbers are not taken at the rear wheels. With the head swap and small cam. I really don't think there is anything wrong with the numbers I got. I also locked the converter at very low rpm for the dyno test this time.The reason is to be able to use the dyno brake to load the engine much lower for repeatable tuning and to keep the rpm from falling back, messing up the printout for the run like in the first test where it has three 3100 readings. I am sure that has some effect too. That's not the chip I run on the road though. The fuel table is more refined now. Its snowing outside but I think I will burn a chip again at a fixed number and see if what the computer says its getting is really what the crank says timing wise. I found this thread and it seems like something worth a look at. https://www.thirdgen.org/forums/diy-...ncy-table.html

https://farm8.staticflickr.com/7812/...55212e98_b.jpg