ULTM8Z

I can't believe how difficult it's been to change injectors!!

Sorry for starting a new thread, but I'm trying to summarize all of this so it's easier to follow:

1.) Changed injectors to Bosch III, 30 lb, M-9593-BB302 (0280155759), and input the correct voltage offset tables from some earlier posts by RBob as well as datasheets that matched that I found on line.

2.) Put in the LT1 VE tables assuming that was a good starting point for my Miniram with slightly larger cam. I started tuning via the injector constant to whip the BLM's in line. Got up 36 lb/hr before the BLM's uniformly came in line to around 128. It was basically a scaling operation since the injectors appear to be very good and consistently running.

3.) Started noticing a horrendous lean condition above 3500 rpm on the WB O2.

4.) Lots of playing around with the calibration yesterday.I have some other threads running regarding what I've been trying in terms of PE values and O2 R/L threshold. Nothing I changed altered the lean condition (literally maxing out the WB O2 lean scale above 5000 rpm).

5.) Reading up about PW limiter, I transferred all my calibration changes from ANHT base tune to S_AUJP v5 base tune, which I read had eliminated that issue. That didn't solve it either.

6.) Today, just for the heck of it, I lowered the injector constant back down to 30 lb/hr. I figured that's the one thing I start playing with. VIOLA! The lean condition disappeared! I command 13.2, I get ~12.9-13.1. Doesn't get any better than that (and I can definitely feel the improvement in the butt dyno).

Tunerpro data indicates the following:

At 36 lb/hr injector constant, PW tops out at 6 ms

At 30 lb/hr injector constant, PW tops out at ~7.1 ms

Now here's the rub...

At 30 lb/hr injector constant VE's drop down to below 108 all over the place. If I scale the VE tables accordingly to get that back to ~128, I end up with ridiculous numbers like 75% VE at 5600 rpm in the tables, which makes absolutely no sense.

Anyone have any ideas on what I'm missing here?

I may try swapping to another O2 sensor to rule that out as the culprit for the BLM's, though I doubt that's the problem... since I didn't have this issue with my old 24 lb Ford blue tops... and the data shows it oscillating nicely between .1 and .9 V, with lots of cross-counts...

But any other thoughts?

ULTM8Z

It's not the O2 sensor. Just swapped out for the NB I used on the passenger side prior to the WB. Same BLMs as before.

O2 ground is ground at pin E15 checks out.... ~just a couple of milliohms to main engine ground. O2 sensor case ground has a ground strap clamped to it and going straight to the main engine ground, so I know that's good as well.

Idle O2 thresholds are still factory AUJP.

Air divert O2 mV offset is zeroed out.

I've attached my .bin.... I don't know... if someone has time, maybe they can take a quick look at it and see if I'm missing something obvious...

ULTM8Z

Found something interesting in S_AUJP hex...

According to the sticky on $8D PW limiter fix... the locations 4C81 through 4C88 (no offset) are the following:

Factory AUJP => DE F1 BD E4 2C FE 84 1C

What JP86SS says they should be to fix the PW limiter => FE 84 1C BD E4 2C DE F1

What's in S_AUJP_V3 => FE 84 1C BD E4 2C DE F1

What's in S_AUJP V5 that I'm running (as-downloaded) => 4F 97 E6 97 EC 7E CD CB

I have no clue what these values mean... but I guess I was suprised that the S_AUJP V5 values are different than V3...

Should I change to the V3 values and then run the hiigher injector constant again to get the BLM's back in line? I'm a little timid about changing hex since I'm not at all a hexidecimal guy...

bigal55

iTrader: (1)

I've been following your posts on this problem and also read the attached link about BPW limit. I did see the part about reversing the data in the equation for BPW, but also remember something about not taking the address as gospel as it may be different for different bins. There was something about doing a search for the hex sequence and verifying the data. The hex values you show in S_AUJPv5 does not make sense as the opcodes don't match. Try searching for the sequence, either the original or the swapped values as the search will only pick up the exact match. I'm no tuner by any stretch of the imagination but I do like to read and learn. Hope I could be of some help! GL!


P.S. As the bin gets modified with maybe a patch or some other data inserted or possibly moved around, the following code needs to move to make room for the new code so the equation for the BPW is probably farther down into the bin.

ULTM8Z

Thanks. Yeah, I read about that too.... "don't just take the values verbatim".

In S_AUJP V5, the closest thing I found to the V3 sequence is located at 4E82 through 4E89...

FE 84 1C BD E6 47 DE F1

again, as compared to what's in V3

FE 84 1C BD E4 2C DE F1

Again, I have no clue what this all means...

bigal55

iTrader: (1)

Well, it seems S_AUJPv5 is already swapped as the sequence is the same (relatively). I say relatively because the opcodes seem to be in the right order, it's just the address of the data to be manipulated has moved relative to the position of the equation. Like I said, I'm no tuner or programmer so I am just guessing as to the differences in the sequence.
I would think that the BPW limit should not be a problem with S_AUJPv5. This is just my opinion though.

bigal55

iTrader: (1)

To confirm what was stated above, here is an excerpt from the thread with the fix.


Originally posted by ED89
I reviewed the hex file/code for ANHT and AUJP and found 2 differences between the two. One is location (thanks for the advise JP86ss) and the other being $24 vs $2C.

ANHT LOCATION
4C79-4C80
DE F1 BD E4 24 FE 84 1C(original code)

aujp
4C81 to 4C88 no offsetoriginal code)
DE F1 BD E4 2C FE 84 1C

Since I am just getting going with my first patch. How big a difference are the two values? What does that difference mean?Will the patch still work?

Thanks in advance

ED89 As quoted above, these are the original op codes as they would be seen in a hex editor at the listed locations.
You can see an actual picture in the post above by 11sOrBust.

Change it to read FE 84 1C BD E4 2C DE F1 instead for aujp.
Change it to read FE 84 1C BD E4 24 DE F1 instead for anht.

The difference is due to the location of the 16x16 routine being different in the two bins and the actual code is in a different location too.

bigal55

iTrader: (1)

Something else I was thinking about as I re-read your other thread on this subject. You were questioning the difference in the R/L threshold versus Map for SD (vacuum) and the LT1 bin, R/L threshold versus grams/sec (air flow). Maybe I'm thinking this wrong, but aren't Map (vacuum) and grams/sec (air flow) the inverse of each other? The higher the Map means less air flow which creates the higher vacuum?

htrdbmr

You have experienced the same issues I had, and corrected them the same way (resetting the constant and then adjusting the VE's). I too switched to S_AUJP, but when I still had the lean-out, I went back to 8D_MOD. I'm not as experienced as most of you guys, so I've only been lurking here in the background, and absorbing what I can from what you post. Some day I'll get it right.

Orr89RocZ

iTrader: (20)

I briefly scanned through the thread but imo you always want to give engine what it wants. Set injector constant to what your injectors flow at the pressure you have and adjust ve tables to get everything inline

For pe mode, use pe enrichments to get wideband air fuel right

ULTM8Z

Higher MAP has to mean higher air flow. At WOT, I have maximum airflow and I'm at ~100kPa.

It seems to me having the injector constant at 30 and getting the correct PE is the right setting. At 36 lb setting, the ECM (while "flying blind" in PE) is delivering an AFR that it “thinks” is correct based on the injector constant; but in reality it's lean. At 30 lb setting, it’s delivering an AFR that is very close to correct. To me, this rules out bad injectors or an incorrect injector constant. The injectors have the correct p/n on them too.

That said, the issue has to be somewhere in how the ECM or engine is operating in closed loop. Forwhatever reason, it either the engine is running horrendously rich or the the ECM thinks it’s running horrendously rich.

In reality, the car doesn’t run too shabby once the NB side settles in at the 106 value (S_AUJP minimum BLM is 96 I believe) at which point the WB side is reading around 13.5-14:1. So that seems to be telling me that somehow the engine runs super rich in closed loop and the ECM is simply doing it’s job.

So then the question is, why is the injector constant just right in open loop, but too rich in closed loop?

ULTM8Z

For the most part I agree. But when the VE's are this far off from other Miniram bins I've seen and LT1 bins, that seems to point to something being wrong.

I'm gotten tired of tuning around malfunctions only to have to start over once the malfunction is discovered and fixed. So I'm more inclined now to figure out what's causing this discrepancy and fix it before I get into re-tuning for the injectors.

Orr89RocZ

iTrader: (20)

I wouldnt worry about what other bins are doing. Not all setups are the same or act the same. Different injectors act differently and may require completely different fuel curve

ULTM8Z

It just makes me wonder if the VEs are off that much, what else could be off?

One other interesting thing to note that I just realized… (but need to figure out what it’s telling me)

Apparently (for the given target AFR I have) 6 ms BPW is too lean for 30 lb injectors, but 7.2 ms is hitting the target AFR almost precisely. That’s a difference of 20% on the BPW.

Running at 106 BLM coincidentally requires a 20% reduction in VE to get back to 128.

RBob

iTrader: (1)

This isn't an issue, the curve of the 100 KPA column of the VE table will follow the torque curve of the engine. So at peak torque there will also be the peak VE value. Once past the peak torque RPM, the VE will be dropping off.

Before anyone jumps up & down about how that can be, as the engine needs MORE fuel as the HP is still increasing. Recall that the injectors are fired every engine revolution, which means that more fuel is injected as it is based on RPM (engine revolutions). The higher the RPM, the more often the injectors are fired.

RBob.

ULTM8Z

I see what you're saying, but the VEs are off by about 20% uniformly everywhere in the table (even at idle), not just talking WOT here. There's some global offset occuring somewhere that's making the engine run rich in CL but meeting target AFR almost precisely in OL.

Orr89RocZ

iTrader: (20)

Have to be careful with miniram and cam reversion effects. Trust the plugs and verify with wideband. Short runner intake could be seeing some cross talk between cylinders/runners that fudge narrowband readings.

May need to adjust o2 offsets at the low rpm areas or idle in OL to get blm right

Just something to keep an eye on.

I noticed changing from stock 19's to 36 bosch on my 305 it wasnt as easy as change injector constant. Ve needed some work

ULTM8Z

What I may also try is swapping sides with the NB and WB and controlling the engine from the other side to see if there's a difference. I didn't have split BLMs the last time I checked with NBs on both sides, but I'll check it anywya.

ULTM8Z

Trying to think about the airflow side of things... bad data on qty of air could also screw up the fuel calcs...

1.) Considering at air density vs temperature -

My Tunerpro data seems to indicate the air temperature readout is correct. I'm running a LT1 air temp sensor into the intake duct right before the throttle body.

The calibration switch enabling IAT has been turned on ever since I put in the cold air duct and it's also set now in my conversion to S_AUJP.

Tried altering MAT compensation counts table (plugged in the LT1 values) and it had essentially no effect on BLM at idle (and the values are markedly different, with LT1's being higher). Idle speed was all I got a chance to do last night at the very end. But I'm looking at idle as representative of what to expect elsewhere, due to the uniformity of the issue across all cells in the VE table.

Also tried both the AUJP and LT1 Inv MAT table as well, with essentially no effect on BLM.

Looking at temperature vs density, to get a 20% drop in air density (to explain the 20% rich condition) would require the ECM to be fooled into thinking the incoming air temperature was ~200°F (assuming an ambient air temperature of 70°F, which is about what's been the last few days). Tunerpro data shows IAT temperatures on the order of 100°F.

2.) Considering airflow

There are no restrictions in the air duct. At the last dyno day (with my old injectors), removing the air duct had no impact on manifold pressure even at WOT at 6000 rpm.

ULTM8Z

Ok, so I took the O2 swing points and moved them up and down by ~30% from factory AUJP settings. No noticeable change, at idle or cruise.

I shouldn't have to go any further than that to make it work...

There's definitely something wrong here besides just tweaking VE tables...

ULTM8Z

@#&%!!!!!!!!

I'm about ready to rip these @#!%-ing injectors off the car and put the old 24's back on!

I'm pretty much back to square one now... even at 30lb setting, the AFR's are super lean (i have no idea what happened yesterday when it seemed like the 30lb setting worked). But I can't repeat that...

The BLM's are fixed after having scaled the VE tables accordingly.... (for all the good it did me).

The only thing I can think of remaining is the injector impedance...

Bosch-III's measure out to 18.5 ohms

My old TPI-style Fords measure out to 14.5 ohms.

Could this change in impedance be somehow confusing the ECM or something?

Totally at my wits end....

84Elky

Elaborating on what bigal55 said. This applies to S_AUJP v5. The referenced code in S_AUJP v4 was located at 0x4D90 through 0x4D97 and the bytes were:
FE 84 1C BD E4 FB DE F1

In S_AUJP v5 the location of that code changed because code additions were made above 0x4D90. The v5 location of this code segment is at 0x4E82 through 0x4E89 and is now:
FE 84 1C BD E6 47 DE F1

There is no change in code functionality in v5 from v4. So why the byte changes? It's again because things moved in the code. The two bytes that changed in v5 represent the new address of a subroutine as bigal55 stated. In v4 the subroutine was located at 0xE4FB (E4 FB). In v5, it's located at 0xE647 (E6 47).

bigal55

iTrader: (1)

The impedance of the injectors won't matter to the ECM, only that it is able to drive enough current to open them reliably. Being a higher impedance than the old Ford style, the ECM has to drive less current to open those 30 lb. injectors. When you put them all in parallel, the difference is negligible, but they will be slightly easier on the ECM driver circuit. HTH!

bigal55

iTrader: (1)

Just a quick question, you say a BPW of 7.2 mS sets the AFR in open loop to about perfect, what was your BPW using the 24 lb. injectors?

ULTM8Z

I never ran the wide band on the 24s... so it's hard to evaluate what the BPW was producing.

84Elky

A quick look at the posted BIN shows nothing out of line, but the flow rate is 30# and VE values are not at 75% at 5600. They are at 89-91%. If you could post a Log that corresponds with the posted BIN, I'd be glad to take a look at it. Or any log and BIN that correspond will be OK as long as they show the lean condition. Please post a .csv or .xls that you've exported from the TPro .xdl file.

ULTM8Z

Hey thanks a bunch for doing that.

Yeah, that was the combination that had the BLM's correct and what I thought was producing correct AFR's (but still now appears to be lean PE's).

I'll get the bin that has 36# setting plus the lower VE's. That's what originally got me the ~108 BLMs and the lean PE's.

Also will post a tuner pro .csv output for the 36#/low-VE calibration as well as the 30#/high-VE.

I'm also going to go back and probe the ECM pinouts again for voltage vs the schematic to make sure nothing is wrong there...

84Elky

No problem. Still would like to get the BINs and csv files; but thinking more about this, it almost has to be something mechanical - like the injectors - since nothing appears to have changed calibration-wise to any great extent. At least nothing that would cause an overly lean condition.

Am arriving at this conclusion because reported BPW is a calculated value based on AFR (whether in OL, PE, or Hwy Mode [CL always =whatever set for Stoich]), Cylinder volume, BARO-adjusted VE, Injector Flow Rate, MAP, RPM and MAT-based air flow. The only variables here we can control are VE, Cylinder Volume, Injector Flow Rate and MAT air flow -- all of which are OK in your BIN. The calculated BPW is adjusted upward with some things, but given all are upward adjustments, that does not lead to a steadily more lean condition. That would seem to come more from declining fuel pressure or the actual fuel out of the injectors being less than the 30# or 36# specs.

Also, you mentioned somewhere not seeing commanded AFR being close to what your WB was reporting. That's likely occurring in PE for which you can calculate very closely what the commanded AFR should be using the decimal values from the standard or S_AUJP v4/5 extended Lambda W Table (either 0x617 or 0xB3A) and Lambda C Table (0x60D/E) as:
PE AFR = (Stoich AFR at 0x41A) / (Lambda W + Lambda C)

HTH

ULTM8Z

The thing that's making my brain hurt is basically the subject line of this thread...

I have the car running rich in CL but lean in OL. If I correct for one, the other gets worse!

Recall during initial re-tuning for injectors, I had the PE essentially turned off. It was still going way lean... and at the time I was thinking in terms of O2 constants (per my other thread). Turning the PE back on didn't result in any noticeable improvement which is what brought me to this thread and the resulting handfuls hair follicles that used to reside on my scalp!

So I do agree that it doesn't appear to be calibration related at this point....

The reported AFR gets progressively leaner as the RPMs climb above 3500. Like something is essentially capping off the fuel flow based on some malfunction.... engine takes in progressively more air with RPM, but ECM is either "unwilling or unable" to match that with more fuel.

If so, it's going to be a huge coincidence that the issue occurs just as I change from the old injectors to the new ones...

ULTM8Z

Ok, so some new information...

I can eliminate the lean condition by lowering the injector constant far enough (found out accidentally). So it's looking like I may not have a mechanical problem.

I had decided to start from scratch with an old TPIS Level IV bin that I got from a friend a long time ago.

It has a 12:1 AFR programmed in across the board.

I ran it accidentally without changing the injector constant to 30 lb/hr. It was set for 24 lb/hr by TPIS.

Gave me a 10:1 AFR on the wideband all the way up to ~5500 rpm.

When I changed to 30 lb/hr and the offset vs battery table, it started leaning out again...

So it seems somewhere between 24 and 30 seems to be what will have the target AFR match the actual AFR.

But here's the confusing thing (AGAIN!), I'm running at ~44-45 psi fuel pressure, but the injectors are rated at 39 psi. So you would think I'd have to increase the injector constant to get the tune correct??? The fuel pressure is holding perfectly steady during these runs too...

If there was some sort of blockage or restriction, I'd expect that no matter how rich I programmed the ECM it would run too lean.

Here's the bin and the run from today after I changed to 30 lb/hr... Wideband was reporting ~14-15:1 AFR for a commanded 12:1.

Orr89RocZ

iTrader: (20)

What is your pe mode fuel enrichment doing when going wot with injector constant set at 30?

If it runs rich at 24 lb constant, to me that means at 30 lb constant, you are not commanding enough pe mode enrichment to get wot afr right

ULTM8Z

No change to PE on this run today going from 24 to 30. 12:1 commanded in both cases.

Orr89RocZ

iTrader: (20)

That doesnt make sense lol. Tune ve tables to 14.7 air fuel then turn pe mode on, adjust enrichment til wb is reading where it should.
Pe mode should be doing something else its either not getting enabled or is not enough adder

ULTM8Z

I just back probed the ECM again and found that the PPL-WHT wire at D8 (EST-High reference) is showing 1.05V when the schematic says it should be 2.3V?

http://chevythunder.com/fuel%20injec...%20pinouts.htm

Which is correct? 2.3 or the 1.05 that I'm getting? Checked the wire itself and it's only .1 ohm from ECM to connection at the HEI module.

I also just found an old GM Port Fuel Injection manual that shows 2.3V for the EST high reference.

The tunerpro data has the RPMs slightly erratic... does the ECM depend on the HEI module for injector PW?

Edit: The WHT wire at C8 is showing 1.5V when the schematic says 1.3. the EST low is showing -50mV when it should be 0.

RBob

iTrader: (1)

Looking at the recent TPIS Level IV - 30lb-hr.csv file, TP samples 1281 - 1285. This is where the ECM transitions from non-PE into PE mode. The injector PW goes from about 5.3 msec in non-PE, to about 7.5 msec in PE mode. The MAP did increase slightly at this point, but not by much.

The commanded AFR also went from 14.7 to 12.0 when the ECM entered PE mode.

So far, it's good, the ECM is doing what it is supposed to be doing. Can also see the injector PW change back when slowly exiting PE mode.

Not for the PW itself, but the ECM uses the DRPs to time the injection events. Under load the injectors will be fired every 4 DRPs (once per engine revolution). If any of these reference pulses get dropped along the way, the injection event will go to every 5 or maybe 6 DRPs. This will cause a lean out.

If dropouts occur often enough I would except the see the RPM take dips in the data log (by half).

I also agree with you that the RPM values in the data log don't look right, too jumpy.

The EST low at -50mV is OK. You are measuring the voltage drop between where you grounded the - probe and the engine block. Ref Low is a ground reference that is grounded to the distributor. It goes to the lower leg of a differential amp.

The high leg of the amp coming from REF High.

This is a GM distributor and ICM, hopefully?

Do you have a known good ECM to try?

RBob.

ULTM8Z

Unfortunately I don't what is "known good" anymore. I've tried two ECM's already.

The distributor ICM is a GM one, but it's several years old. Pickup coil is new.

I put in a Wells aftermarket ICM and made the same measurement... pretty much the same.


I reattached another export of that same data log in that earlier post, this time with everything. Take a look at the BLM Cell though... I'm bouncing off of cell 4 numerous times at WOT and ~4000 rpm. 4 is the idle cell??

ULTM8Z

Here's another couple of logs as well as my old original modified bin, but with TPIS VE tables.

One log has injector constant set at 30lb. The other is at 32lb.

The BLM's seem to be doing well now at 30lb injector setting.

No longer bouncing off Cell 4... must have been an artifact of the TPIS chip.

But the leanout is still occurring... (also swapped in another ICM I had).

I'm almost ready to offer $100 at this point to whoever can figure this out...

EDIT: One thing I just noticed... it seems like when I punch it from very low speed (like a freeway on ramp), the lean condition kinda goes away, but the AFR gets a little erratic (I can feel the car lurching a little). It's when I'm cruising at speed and get into PE that it just seems to lean out above 3500 rpm...

ULTM8Z

Just want to give a big kudos to 84Elky... he did a really in depth analysis of my bin and log files and sent it to me. I haven't had a chance to really digest yet and it remains to be seen what the outcome is, but the time he spent on it is very much appreciated.

ULTM8Z

Ok, I hope I'm not jynxing myself by posting this... but it *seems* as though I'm on the right track here finally... I had the day off today, so I spent a few hours on this thing.

Last weekend when I pulled the distributor out to really examine it closely and test the components, I saw the gear was pretty worn. So I ordered a the melonized ZZ4 gear from GM (I forgot the origin of my current gear).

I also found that there seemed to be excessive lateral play between the shaft and the lower bushing at the gear-end of the housing. So yesterday I re-bushed the housing, installed the new gear, and got all the clearances correct with a new shim kit from Summit.

It seems to have corrected the erratic RPMs I was seeing. It also seems to have contributed to better stability in the reported AFRs (I'm thinking erratic spark timing could have been screwing around with the reported AFRs, whereas I thought it was strictly the fueling).

So this morning, I took the car out to do some more road testing and spent a few hours on it. I have this great 3 mile patch of freeway by my house where hardly anyone is on it and its a good uphill incline, so I can load the engine up nicely without picking up excessive speed.

Any rate, using the TPIS Level IV VE tables (slightly richened up at high map between 1600 and 3000 rpm), I used that as "baseline" to see where the injector constant wanted to be, as well as using the feel of the existing AE values.

I started to get the PE above 4000 rpm (the problem areas) to come in line when I used injector constants less than 30 lb/hr (starting with 26 and working my way up). Anything 30 and over, the reported AFR leans out REALLY fast above 4000 rpm. Don't know what's up with that... but I'm not sure it matters at this point.

I ended up settling on 29 lb/hr injector constant. At that value, the BLM's look absolutely fantastic. Just some minor tweaks here and there, and it should be perfect.

I haven't gotten the WB output to show up on Tunerpro yet (it's hooked up, just I haven't managed to get TP to recognize it and report the output yet).

But, it *appears* as though (from glancing at the WB gauge during WOT) I have to command the AFR about 1 point richer above 4000 rpm to get the actual AFR to what I want (i.e, if I want 13.2, I have to command roughly 12.2). Once I get TP reporting the WB, I can get a more definitive look at that.

What was mildly (albeit pleasantly) surprising is that the car doesn't seem to like the richer AE settings of the LT1 $DA3 any longer. I'm returning the AE tables back to $8D one by one, resulting in a crisper throttle (and a VERY crisp neutral throttle blip).

Going to spend some time now to get WB to show up on TP so I can fine tune... But so far so good... Again I hope I'm not jynxed now... knowing my luck I'll take the car out again and it'll be right back to where I was!!

Orr89RocZ

iTrader: (20)

Have you tried leaving ve table alone and adjust pe enrichment vs rpm?

ULTM8Z

That's essentially what I did.

Once the BLM's lined up with the existing VE's in the TPIS Level IV chip with the 29 lb/hr injector constant, I started playing with the PE enrichment. It seems to be very close (target vs actual) below 4000, but above 4000 is where I have to start commanding it quite a bit richer to get what I want.

ULTM8Z

Freakin-A.... The car is running extremely well!

I'm essentially dialing in BLM's via the MAT Compensation Counts vs MAT table now.

I changed to the LT1 (DA3) table and noticed a very nice improvement. The minor tweaks I was alluding to in the earlier post #39 above seem to be getting smoothed out with mods to this table (there seems to be some BLM variability with air temp). So I'm now running both the Inv MAT table and the MAT compensation counts table from the LT1 and the engine really seems to like it.

It was interesting result since I tried the MAT compensation counts table once before with disappointing results... but that was before the battery vs fuel pump voltage discrepancy was fixed, and now this Bosch III injector swap with the correct battery voltage compensation table and injector constant dialed in. I decided to give it another go this time and it worked out very nicely.

What's more, I picked up another ~1 to 1.5 mpg on the highway. My mpg calculator (which worked out to within a 1/2 mpg to the fill up mehod) is now showing avg 26-27 mpg at 70 mph on level ground when the highway mode kicks in.

One big lessons learned for me is that I wasted a TON of time trying to tune without a WB O2. One of the best ~$150 I ever spent. Many of my pre-conceived notions about what the engine wanted went out the window when I found out what the true AFR readings actually were.

I still want to release another Miniram start bin when I get it to where I can't get any more improvement out of it.

bigal55

iTrader: (1)

Great! Can you say "perseverance"? Glad you figured it out, anyone who follows this thread will surely learn a few things. Good job!

ULTM8Z

And a many thanks to the folks on this board who spend the time to help and literally make this EFI tuning hobby possible. I swear, if it wasn't for this DIY PROM forum, it wouldn't be possible for me to do this...