Pro

I drove the TPI truck today (16197427 ECM) for a few minutes just enough to get some data logged. I dont know tunerpro enough yet, only this little app on my phone that exports data to a CSV. Attached is an export of INT and BLM values. As you can see, Open loop, then closed and idle for a minute, then i drive for about 4 minutes. BLM takes a dive after going closed.
So I'm nearing 110 at an idle, and actually hitting 105 when cruising at 10mph in a residential. Exhaust is pretty bad too. EGR is disabled.
The injectors are the yellow south bay's, and I read somewhere that they should be ran at 40psi. I am currently at 45. So should I be purchasing an adjustable regulator and bringing it down a little? If so, which brand which you recommend?
I can attach the whole spreadsheet if you want to go thru all of the values, but its a pita to read thru, like doing accounting work line by line.

Tuned Performance

iTrader: (92)

Even with the aforementioned you won’t be able to go below 3bar. You should adjust your ve.

ULTM8Z

You could start by raising the injector constant by a point or two. I'm skeptical you could lower the pressure enough to get the BLMs back up to normal and not result in driveabilty issues due to too low of a fuel pressure setting.

ULTM8Z

This is assuming of course that there's nothing else causing the low BLMs.

Tootie Pang

iTrader: (2)

Though your BLMs are low, I don't think they are hitting their lower adjustment limit (what is it? Does anyone know?). Your engine ECM appears to be able to reach stoic by shortening pulse width.

Lowering your fuel pressure should help bring the BLMs up.

I don't think AFPRs go lower than stock unless you buy them that way intentionally and I probably wouldn't run less than 36-37 psi but that's just a wild guess.

ULTM8Z

Lower adjustment is 108 on $8D for BLMs (though you can make it lower in the calibration if you want). For integrator, I believe it's in the low 90's, if not the 80's. Not sure about other masks...

Pro

Tootie I think I read in another thread someone had said to be worried if are constantly running around 90-100 at all times. I'm not yet there, but somewhat close.
Attached is VE. I found a few calculators online. I dont know how accurate, but putting in my values for a stock 350 converted from TBI to TPI calculate to be almost half of what these values are.

Pro

I warmed the motor up and was able to duplicate the above results in my first post. So I then adjusted the VE numbers down a couple, and started it back up and let it idle for 5-6 minutes. If I'm not mistaken, do these numbers look much better? Previously I was hitting 112 at idle. Here, the last minute or two of idle was 119.

ULTM8Z

119 is certainly better than 112. Though your VE's seem quite high across the board...

In cases like that you could potentially scale everything down by increasing the injector constant (making the computer think the injectors are bigger than they are will cause it to decrease the pulse width... less fuel will bring the BLMs up).

Once you get close with the injector constant scaling, then you can fine tune BLMs on a per-area basis with the VE tables.

Pro

Thanks for the advice.
Is it normal to see INTs come down though? Why wouldnt BLM and INT be together in sequence (both go up/down together)?

ULTM8Z

Basically the integrator reacts to the O2 sensor and the BLM reacts to the integrator.

Integrator is the ECMs short term response to changes in O2 sensor reporting. If the integrator hangs low for an extended period of time, it'll start to pull down the BLM.

BLM is long term fuel trim. Integrator is short term.

For example, if the Integrator hangs down below 128 for a while, the BLM start to drop as well. As the BLM drops, the integrator will start to go back up to 128 because the O2 sensor sees the air fuel ratio becoming correct again. When the BLM settles in at the new trim value, it means the ECM is happy with the fueling and no more adjustment necessary for that operating condition. ECM then records that fuel trim for that operating condition such when you hit that condition again, it knows how to trim the fueling... you may get a mimor integrstor fluctuation again but it'll be short enough that the BLM won't react to it.

Pro

Do most people send an ECM reset, or pull power to the computer before testing a new chip? This last go around I didnt.

ULTM8Z

If your BLMS are already well established and you're making incremental VE changes, I'd avoid resetting the ECM. It'll just take longer to reestablish the new BLMs.

Pro

Attached are a couple of tests from yesterday and today. My previous post was with all various with 4 points subtracted, and BLM ended up around 119. Then I ran with 6 down and ended around 123. And tonight, tried 8 down and it spiked to 131-133. I can only assume that 7 may be the magic number. But thats interesting that 6 to 8 would jump BLM that high.

I went ahead and did 7 down. It leveled at about 127. I assume not many people get a perfect 128 all the time, and 127 is just fine?

That spike at the beginning on 7 was me backing out of the garage. The fumes are horrible and I dont know why.

ULTM8Z

General rule if thumb... if you have a BLM in a cell that's say 119, you can multiply the VE value in that cell by the quotient of 119/128.

If you have a BLM of say 136, the multiply the cell by 128/136.

RBob

iTrader: (1)

The INT shouldn't be oscillating like it is. Need to work on the proportional gains to correct this.

RBob.

Pro

It shouldnt be oscillating at all, or by that much? How much should it be? It looks to be a - + 5 here.
Added O2 just to show the obvious coordination with the O2 sensor.
One video I watched the instructor mentioned that normal short term fuel trims should oscillate between two and five percent. Mine appear to be about 6 percent. So that doesn't appear horrible. But his was OBD2 if that even matters.

Pro

Cool I didnt even notice that.
BTW this helped a ton to understand fuel trims
And he mentions in there that ECMs must be reset after fixing issues affecting fuel trims.

RBob

iTrader: (1)

During steady state engine operation the INT shouldn't oscillate at all. Note in the graph that the INT is what is causing the O2 cross counts. That is the job of the proportional gains. Maybe the 'instructor' isn't familiar with this.

The engine will run smoother and sound better once the proportional gains are correct.

Not always. Many ECMs have a power up initialization of the BLM cells. Just set the value(s) to 128 for this feature. There are even some ECMs ('8746) that set the BLM cells to 128, hard coded, at every key-on.

RBob.

RBob

iTrader: (1)

Here you go, for the $0D mask in the '7427 PCM:

Set those to 128 to reset the BLM cells to neutral at key-on.

RBob.

Pro

Cool so thats a nice way to do a forced reset in this ECM.
Where do I need to go for those proportional gains? Not only do I have the screeshot posted above but also what fast355 has here:

https://www.thirdgen.org/forums/diy-...ml#post3153057

But this is only for idle.

RBob

iTrader: (1)

Did you check the xdf for those parameters? There is also a hac of $0D available.

RBob.

Pro

I'm not worried about it anymore (in terms of the 2 percent fluctuation in short term trims). Everywhere I have read, these shops/instructors on youtube say thats within spec.
What I am worried about though is when I went driving, I was coasting downhill at around 25mph, foot off the gas, and rpms were jumping up and down 600 to 800.

RBob

iTrader: (1)

What does the data log show for fueling and SA?

RBob.

Pro

Hopefully what I have attached is enough. If you want me to go duplicate this in a different manner I can easily.
But what I did here is from a dead stop, accelerate slowly thru first gear. Once midway thru second, I let off the throttle and let it shift into 3rd by itself. That happened at about line 47 (hence the 2 RPM spikes at the beginning). Then, after line 47, i just coasted downhill, with the foot off of the throttle. I reached about 25mph, and then at line 142 hit the brake and slowly came to a stop. The small spike at the end of the graph was most likely the downshift into 1st.
But from 47 until I hit the brake at 142, the RPM is up and down.
My thought is that maybe its in 3rd trying to lock up the TC unsuccessfully. But I dont think theres a PID for that anywhere so I cant prove it.

RBob

iTrader: (1)

Need to do some more tuning on the VE table(s) and likely the prop gains. Note how the BLM is as low as 105, and varies then when the cell in use changes (different cell).

Best to also place the O2 sensor value in the .xls SS.

RBob.

Pro

Here are a couple of runs around the block in 2nd gear. The first at 2000, second at 3000. Both end up with knocks. The ECM logged 4 knocks total. The 2nd graph highlights with red dots where the knocks occured. I assume its because its running lean.

RBob

iTrader: (1)

How is the VE table coming along?

RBob.

Fast355

iTrader: (2)

I am assuming this is a stock cam/stock head TBI 350? If it is, you will likely need to pull some timing in the 2,500-3,500 rpm range in the 70+ KPA range where the TPI runners do their thing. I found from 4,000 rpm and beyond I could add some timing to the stock TBI timing maps up to about 26° total advance on 87 octane. With 93 octane you can generally add timing across the board and make more torque. Most TBI engines I have messed with at 90-100 kpa like about 10-12° flat up to 1,200 rpm then around 20° from 1,600 through 2,400 and then a slow rise to 26° at 3,600 rpm through fuel shut-off. When you add TPI they like the same low-end timing bump, but drop to 16° from 2,400-3,600 and bring up the timing to 26° at 4,000 rpm to fuel shut-off. At 3,200-3,600 rpm you will find the timing 8-10° retarded from the TBI setup. You can find a thread on here where one of my timing tables was posted and another member based his TPI swapped TBI 305 setup on and found worked well for him too.

Pro

The cam and heads are stock that I know of. Rob I have been modifying VE tables still, trying to get rid of the knocking. So far increasing VEs has fixed it. Yes i noticed that 2500-3500rpm the TPI seems to pull a little more than the TBI did. It feels much more like the camaro does.
Hopefully Ill be done with VE this weekend, and I think I have the trans shift parameters set nice as well. EGR is turned off. Ill probably have to pull the computer back out of the glove box and revisit every bit of this again in december/january when i fail emissions lol. Although at 1300rpm 15mph and 25mph which they are testing me at, according to recent results i just recorded from the ECM today, short and long term trims are right on 128 +- a couple of points.

Edit:
Going over more data and found more knocking. At one point the knock count got up to 4 and knock retard jumped up to:

0.703
0.527
3.867
4.922
4.57
4.219
5.977
7.383
8.262
7.734
7.207
6.68
6.152
5.801

RPMs range from 3200-4000, INT 128, BLM 138 or so.

Fast355

iTrader: (2)

I was a little off on what I said for timing from memory but not much. I forgot to metion my PE spark advance was set at 4° from 1,200 rpm and beyond. Richer mixture needs more timing to burn completely then the closed loop stoich mixture.

https://www.thirdgen.org/forums/tbi/522536-tpi-305-tbi-heads.html

What you are missing now to make that truck run like your Camaro is cam. The TBI cam is pathetically small. I have used the Mellings MTC1 in several non-roller TBI and TPI builds and not had issues getting them past smog. The stock exhaust manifolds are horrendous and even a set of shorty headers would wake the setup up noticeably when combined with a custom 2.5" to 3" Y-pipe and a free flowing 3" cat like a Thunderbolt high flow and non chambered straight through muffler like a Magnaflow. I bet you could find another 40-50 hp in that setup with only those changes.

Pro

I've already taken it apart enough. And it has 175k on it so I don't think it's worth it. Maybe in the next couple of years I'll put a new motor in it with the TPI on top. There's a guy on YouTube with a 383 and seems happy with it
I'll look for the spark tables tomorrow or the weekend and hopefully figure that out. Should I also run higher octane if that might help with the knocking?

Pro

BTW it already has borla shorties on it, for the past 20 years. and I have a couple of cats and mufflers, I just need an exhaust shop visit

RBob

iTrader: (1)

This is the RPM range where the long tubes start to do a better cylinder fill. As Fast355 stated need to reduce the spark advance in that area. Stock TPI BINs have a dip in the SA from 3200 - 4000 RPM.

RBob.

Fast355

iTrader: (2)

A 383 TPi is a great way to move something heavy around. I recently pulled the 383 out of my old 1983 G20 van to make room for an 8100 Vortec and it is on the stand at the shop. It is getting resealed externally, swapped to the 24x ignition setup and put into my 1997 Express van. The 383 is a 10.5:1 setup with heavily ported 906 Vortec heads with 2.05/1.60 valves, a mild Reed custom roller cam (260/264@ .008, 208/212@ .050, .480/.507" lift (.512/.541 w 1.6 rocker), 112 LSA, 108 ICL), 1.6 full roller rockers, ported edelbrock 3817 base, siamese ported SLP runners, ported plenum and LT1 throttle body with airfoil. In the Express it will be run with a late model LS PCM, 24x crank reluctor, Delco 585 coils near plug, and a vortec distributor base for cam signal. The air filter assembly and MAF will be from a 2002 Express van using ebay custom ported MAF ends for a LT1/LS1 maf. The Express already has Doug Thorley Tri-Y headers and a dual 2.5" exhaust system with Thunderbolt race cats. Backed to a 4L85E and 4.56 gears.

Pro

Youre defintely the king of vans. Youll be putting electric motors in them someday probably

Pro

I noticed in your link to the other thread, 1MeanZ, increased the timing in the lower RPMs. While I'm in this table (i assume the correct one - Main spark table - map vs rpm open throttle) I may go ahead and do what he did if it's worth it?
BTW, I brought down the SA values just a point or two less than what the ECM did automatically when it heard the knocking. 2800-4000rpm, 75-100 kpa.

Fast355

iTrader: (2)

This may sound weird but when I wrote the spreadsheet I use to for adjusting timing maps through a knock retard map, I only reduced the timing in 1/2 increments of the knock retard and was sucessful. If it pulled 4° by itself, usually removing 2° fixes it.