The Vortec swap nightmare saga continues.................

[pmkls1]

I figured I would post a new thread rather than revive an old thread so that people don't have to sift through dozens of posts to get down to the nitty gritty. For those just tuning in the important info is; '89 Firebird Formula 5.0 L03 TBI w/700R4. Replaced 305 with used L31 shortblock and cylinder heads from different L31. New HV oil pump, GMPP single roller chain, 7qt deep sump oil pan, custom fabbed front half of exhaust from original 2" outlet manifolds to 3" midpipe and flowmaster muffler. Edelbrock Performer RPM 4bbl intake and Holley 4bbl-TBI adapter and reused ALL other hardware from the 305. Base fuel pressure ~30psi due to idiot mechanic (me) installing way too much of a fuel pump long time ago. Later replaced 1228746 ECM with 16197427 PCM. So, in the last episode I realized that I had installed the incorrect dizzy gear and that the RPM signal to PCM was fluctuating and determined that it was time for a new dizzy.......

Fast forward to the present and I have made progress and also determined that I am destined to rebuild this car piece-by-piece wether I want to or not. I finally got a new GM distributor and also a new Accel cap & rotor kit and installed them. The GM dizzy comes with a cap and rotor installed, but I simply prefer the Accel parts as they are more robust and just as affordable at ~$25 for the set. I trimmed and re-routed the existing Accel 8mm spiral-core plug wires as they are in ok shape and I found a spare set of terminals for the dizzy end of the wires in the basement. I also replaced the low-mileage plugs as there was some fouling from a rapidly progresing misfire. After all said repairs were made I restarted the car and found that I had cured one ailment only to discover another. The old dizzy was having issues that I believe to be rooted in the aftermarket ICM. The new dizzy did clear up a lot of isues and also required a remap of the ignition timing after trial and error showed that the old map parformed very differently with a new and "proper" dizzy was installed.

The problem was, however, that upon fixing one issue the next issue only stood-out even more. Cutting to the chase, I'll leave out the boring bits and get right to the isue. After some driving and tuning, I discovered the MAP sensor to be going bad. I always monitor it in my logs and quickly noticed that it would not register numerically lower than 38kpa anymore. I instantly recognized this as a problem knowing full-well that at idle in park it usually reads closer to 30 and while decelerating it will drop into the teens under the right circumstances. I quickly sourced a known-good used late model MAP sensor and connector and installed it directly to the TBI unit with a short section of hose. Upon startup I knew that the issue was resolved before connecting the laptop. That led, however, to yet another issue..........

This time the issue is mostly programming related. I had been logging and using the VE correction spreadsheet as the MAP sennsor was going bad (unknowingly of course) and had been tuning IN a terrible deadspot. Once I had replaced the MAP sensor, the deadspot only became more pronounced and it took a little watching to spot the issue. So, to correct the issue enough to drive the car I had to throw in the old VE tables and start from there. The car drives pretty good now, but there seems to be something wrong with my DE and DFCO settings. What is happening is that while decelerating the BLM will continue to drop until it switches into DFCO. Then, upon exiting DFCO, the current BLM cell will be all jacked-up for lack of a better term and the car will lean out to the point of stuttering until the BLM counts back up into the proper range. I can bypass the deadspot by nailing the throttle and entering PE mode, but this is obviously not a solution.

So, I will attach my latest BIN and XDL for any to take a peek at. The VE tables are way off at this point because I reverted back to some starter tables after replacing the MAP sensor and using the VE correction spreadsheed will make the problem worse due to the current issue. Using these crude VE tables makes the car drive the best at the moment until I fix the DE/DFCO issue so take note of that while looking things over. The spark tables are pretty close at the moment and the knock counts are due to several different verified mechanical noises (non internal-engine damage related noises) at the moment so the knock retard has been disabled temporarily. I think there's more, but gotta go for now. Thanks in advance for taking the time to read through this post and reply.

EDIT: All of my recent log files are too large to upload so I will need to make a shorter drive and upload that file.

[EagleMark]

XDL file size has been increased. But a shorter log with issue is always easier to help with.

Welcome to my world! Try tuning a vehicle that has MULTIPLE mechanical issues... I've honestly only done one in 2 years that did not have at least one issue! It was for a tech/owner of a shop.

Something that may be needed at this point is a new bin to start with. We sometimes change so many things that it is impossible to tune out what we tuned in, or something bad happened to bin? I've found compare difference to undefined items? Just a thought that has worked for me. Fueling was off due to MAP sensor, spark was off due to aftermarket EST/ICM. Possible 1000s of changes to bin. See where this is going?

But this stands out from what you said.

The VE tables are way off at this point because

Everything fuel related is an adjustment from the VE table. So if it is not right then neither is the adjustments like DEFCO, DECEL, AE, PE, etc.... You've got a combination engine, wet intake that is aftermarket for a carb, which is probably not an issue but could complicate one. For instance can't seem to get tune right? First more fuel, then less fuel? Well did you adjust fuel after a long idle or after a WOT run? WOT run would be a fairly dry intake where after a long idle would be wet spots.

Other then making an engine run till fully warmed up, then driving a little with a WOT run, I don't even bother to log data or look at wide band. Data is wrong! Then disconnect power to ECM to clear learn. Then start to get VE in order from data. I've just recently found this trick to save hours if not days of chasing my tail. Once VE is in order most other issues go away like your DFCO. If VE is right then it will run just as good without O2 sensor because fueling is correct.

HTH!

[pmkls1]

Well, the real problem here is the combination of parts as you said. I have basically created a one-of-a-kind franken-vortec (speaking in terms of mixing parts) that I have no known good pool of data to draw from. I also suspect that my unique and excessive fuel pressure is creating fueling inconsistencies that are just complicating matters. As far as my VE corrections go, this is how I was using the spreadsheet. I would log a drive that usually began recording after the car was warmed up and would still provide ~45 min worth of recording time. In TPv5 the data history tables populate as you record or playback and you can select running average, which I did based on the wording of the VE spreadsheet. I then copied and pasted to and from the spreadsheet as instructed. Using the running average option will do as implied and average the value of each used BLM cell over the course of however much data you run through it. Since the cells that count down while decelerating are the same cells used when reapplying the throttle, those cells would have a low BLM average because of the time spent off of the throttle. If you enter that value and compensate the VE table accordingly then you end up with a very low VE value in that cell so when you actually need the fuel with open throttle the BPW starts out very low and must slowly climb as the PCM detects an insanely lean o2 signal and slowly counts back up the BLM value. What you end up with is an EXTREMELY lean condition upon reapplying the accelerator after a couple of seconds of decel. The car will buck until the PCM can compensate and add more fuel. Living in a tourist town with lots of stop and go traffic creates a very unpleasant and also detrimental driving experience.

For the record, this issue began before the distributor change and possibly before the MAP sensor began to malfunction. I do not recall any particular event that spurred the issue and it began randomly during a drive home one day after the majority of the drive was over. I also agree that it may be time to start with a fresh bin as I have found that sometimes parts of the bin mysteriously get corrupted after thousands of tweaks and saves and transfers etc..... The issue here is that it seems for some unknown reason the PCM is not entering DE mode so while decelerating it sees that the o2 readings are going rich and instead of going into DE it simply begins lowering the fuel trims for the current cells. It does enter DFCO as you can feel it, but it appears that the BLM will still drop. This is why the BLM's are all over and I haven't wanted to modify the VE tables yet as while on the throttle the BLM's go high to correct the lean condition. I could be way off here with my assumptions, but this is far past the point that normal diagnostics go for a dealer tech such as myself. Especially considering that my training is continually focused on new and emerging technology and as EFI designs age the knowledge, support, and amount of vehicles returning for repairs diminishes. For example, I have never used my timing light on a customer vehicle and my distributor wrench doesn't see much more use. It has also been over a year since I have used a Tech 1 scanner either and The Tech 2 doesn't work on many new vehicles.

[EagleMark]

All the more reaon to start with fresh bin, rough in your Vortec head timing tables, smooth out the VE, check/uncheck what is required and start to tune.

What exactly is your fuel pressure nowadays? I thought we got this back to a realistic number awhile ago? If still super high there may be so many changes needed to compenste it's not worth the effort to tune, espaecailly if that much fuel is not needed.

45 minute logs are skewed data for tuning. BLM learn has taken away the adjustments your trying to make. I can't believe how many hours and days I've wasted with this method over the years. Try this with a fresh bin, I'd really like to see a tech with tuning experiance try my shortcut. EGR, DFCO, PE, Air divertor off. Warm up vehicle and drive, do a WOT run, just a little. Park, disconnect power to ECM, start vehicle and when goes closed loop record data, 5 minutes tops, make adjustments, disconnect power, rinse and repeat. I like to do this with 3-4 TPS% runs to rough in a VE table, then smooth it out and go back to a longer log. You'll notice then if you look at low and high they will be close together, why take an average of 108 BLM and 172 BLM to tune? Which one is right? Since I started this I can get a VE table accurate for a stock, mild or built motor in order within an hour or so. Will run CL or OL just as well, same day...

[1project2many]

Not at home to look at disassembly but there are generally settings for locking out BLM during decel, DFCO, or for resetting after WOT or large throttle changes. Old ways still work... limit INT and BLM range to prevent excessive corrections from affecting VE tuning. What about your vehicle other than fuel pressure is radically different from Vortec TBI Van cal?

BLM issue you talk about is also part of $58 and is very difficult to tune out in that cal. Trick I use is to try and tailor AE to deliver slight extra shot of fuel after decel conditions. Have to work carefully with MAP and TPS AE to make this work and it never works as well as it should because stock $58 just doesn't provide enough control over AE. But the strategy should work better in the 7427 cals as there are more AE parameters to work with IIRC.

[pmkls1]

Not at home to look at disassembly but there are generally settings for locking out BLM during decel, DFCO, or for resetting after WOT or large throttle changes. Old ways still work... limit INT and BLM range to prevent excessive corrections from affecting VE tuning. What about your vehicle other than fuel pressure is radically different from Vortec TBI Van cal?

BLM issue you talk about is also part of $58 and is very difficult to tune out in that cal. Trick I use is to try and tailor AE to deliver slight extra shot of fuel after decel conditions. Have to work carefully with MAP and TPS AE to make this work and it never works as well as it should because stock $58 just doesn't provide enough control over AE. But the strategy should work better in the 7427 cals as there are more AE parameters to work with IIRC.

The timing tables are copied and slightly modified from the BJYL 350 TBI 700R4 BIN that I downloaded from the ECM info section and are a lot more aggressive than stock BJYL and other TBI BINs. Since the true VE tables will be drastically different from the typical TBI BINs due to the fuel pressure. I had actually widened the BLM range which may be worsening the situiation unintentionally. There are more AE parameters in the $OD mask than in the $61 for sure and probably more than the $58 as well. I had added more shot and a longer time of shot on DFCO exit, but had also changed something else in the BIN that made the car barely driveable so I just went back to a previous BIN and tuned from there leaving those values alone. As we speak, I am searching for the most economical MPFI manifold solution along with the best value headers to balance out the entire setup a little better and also deal with the probable injector consistency issue from running excessive fuel pressure. I think that a lot of my tuning issues are due to the injectors and the pressure they've been running on for the last few years. At this point I want to keep the car driving ok, but don't want to waste too much time tuning on the current TBI setup as I'm dying to ditch it ASAP. Once I get the hardware combo finalized and can get the engine running properly I'm wanting to switch back over to running ethanol blends as the sheer smoothness and noticeable torque increases get to be just as addictive as drinking alcohol is.

[pmkls1]

Just a quick note on the current setup.. The exhaust manifolds were intended to be a very short term solution as I ran into financial problems right in the middle of the swap and just needed to make the car driveable. The intake setup as it is currently configured was not intended to be used much longer either. I've since had major financial problems that have limited me to barely being able to keep up with maintenance so I have long passed the amount of time I had planned on using this configuration. I am now in the position to be able to swap to a sensibly priced exhaust and induction setup that will compliment the L31 longblock. So, the bottom line is that I am fully aware that I have had so many issues with this setup due to it being a sad combination of thrown-together mis-matched parts that are fairly aged as well.

[1project2many]

Since the true VE tables will be drastically different from the typical TBI BINs due to the fuel pressure.

With a specific induction system, changing fuel pressure should not require major changes inthe VE table. Volumetric efficiency represents the engine's ability to move air and is not based on the type of injector or the amount of fuel it delivers.* Injector specific values should be correct such as AE pulses, DE reduction, minimum PW, voltage correction, minimum asynch pulse, quasi asynch enable values. These values are designed to account for a specific injector's behavior. I recognize that getting these values correct can take a long time and that often the corrections find their way into the VE table instead but it is important to remember how the VE is intended to be used and to treat it accordingly.

* Wet flow manifolds will tend to have reduced VE compared to dry flow systems especially at low rpm as expanding fuel in the manifold displaces available oxygen.

[pmkls1]

With a specific induction system, changing fuel pressure should not require major changes inthe VE table. Volumetric efficiency represents the engine's ability to move air and is not based on the type of injector or the amount of fuel it delivers.* Injector specific values should be correct such as AE pulses, DE reduction, minimum PW, voltage correction, minimum asynch pulse, quasi asynch enable values. These values are designed to account for a specific injector's behavior. I recognize that getting these values correct can take a long time and that often the corrections find their way into the VE table instead but it is important to remember how the VE is intended to be used and to treat it accordingly.

* Wet flow manifolds will tend to have reduced VE compared to dry flow systems especially at low rpm as expanding fuel in the manifold displaces available oxygen.

That makes sense and I had not thought of any of that. I had convinced myself that the VE tables would be so different due to the limitations of the '746 ecm and the alterations I had to make to account for fuel pressure and also when tuning for E85 as the majority of my tuning experience has been tuning for E85 fuel as it does "behave" differently than gas. Gotta run for now.

[dave w]

I think I'd try setting the '427 PCM for Open Loop and tune / correct the VE Tables using a WBO2 sensor. Once I got somewhere close with the Open Loop tuning, I'd enable Closed Loop but I'd disable the BLM Update and continue tuning / correcting the VE Tables using a WBO2. Then I'd enable the BLM Update and keep a close eye on the BLM averages. I've seen where BLM averages and WBO2 data do not match, so in that case disable BLM Update.

dave w

[EagleMark]

I'd enable Closed Loop but I'd disable the BLM Update and continue tuning
dave w

D'oh!
Like disconnecting the battery? Man that has saved me so much time...

[1project2many]

I agree with Dave. Although there is more than one way to skin this cat, the goal is to tune steady state first then transients. Disabling closed loop, EGR, DFCO, and other functions mentioned by Mark above is a way to prevent transient corrections from affecting your steady state tables, which are the basis for the entire tune.

At times I've favored disconnecting the battery but there are plenty of times when I don't bother. I think there's a correlation between ecm family or ecm number and whether or not this is really helpful. I am firmly convinced C3 tuning can be finished quicker if the battery is disconnected. P4, not so much. And I have nowhere near enough experience with the 7427 and cousins to tell.

[RobertISaar]

with C3, it's almost a necessity if you don't lock the BLM min/max to 128.

with P4, either send a mode 4 command to clear out the learned BLMs or set the BLM min and max to 128.

[EagleMark]

with P4, either send a mode 4 command to clear out the learned BLMs or set the BLM min and max to 128.

I need to learn these mode 4 commands...

[RobertISaar]

they're different for every application, but almost all P4 and P6/P66 units allow datalink control over outputs and a couple of internal functions, like BLM and IAC reset, forced open loop or closed loop fueling, etc...