Bottom line up front... there's more nuance to it than just the obvious injector constant and voltage offsets (which have been much more well documented in recent years).

I'm continuing to dive deeper into the calibration and trying to learn more... The gurus and experts can chime in here and correct me if I'm making incorrect statements, but just stating what I observed based on my still comparatively very limited knowledge even after ~20 years of playing around with tuning. I'm also running 7730 $8D, but from what I understand, this is also applicable to other ECMs and Broadcast Codes.

Doing a bunch of reading on some old threads (which still have loads of valuable insights). Came across this thread where the OP was describing a situation that's similar to what I have been observing over the years. That is, WB readings fluctuating quite a bit despite (in my case) very good BLM readings... obviously just during closed loop operation. For reference, my WB and NB are on opposite sides of the engine.

https://www.thirdgen.org/forums/diy-...need-help.html

Unlike the OP in that thread, I wasn't able to detect anything like surging or missing or anything like that, so I always thought it was normal. The car overall runs very well (even more so having switched to the DA3 VE and MAT tables. But I chased that WB swinging observation all the way up to having my O2 swing points at over 850mV at high map values and 650 mV at low map values, in order to maintain what I thought was the most stable WB readings I was ever going to get. But still, I'd see the WB bouncing off 16:1 at times especially during moderate acceleration. Switching the NB and WB O2's side to side resulted in the same thing, so again I figured there was nothing wrong with the engine itself and it was just normal (also did the same thing with my old ZZ4, before I dropped the 383 in).

Any rate, after reading through that thread, I came across this statement from RBob that has to be one of the most under-the-radar statements ever made on this board (in terms of its importance for bigger injectors, and even more so in terms of putting a complex topic like proportional gain into "English" that a layman like me can understand..... almost made me fall out of my chair... lol...(bold emphasis mine)

I swear, if it wasn't for RBob, Elky, Grumpy, SBFormula, JP86SS, and the other gurus on this site (forgive me for not mentioning everyone by name), there's now way in hell I'd still have this Miniram on the car, much less running as well as it is. I don't even know what else to say... lol I would have gotten too frustrated and yanked that manifold off in favor of doing something else (worst case going back to a carb!).

The main table in question is (as labeled in the S-AUJP V7 xdf)

Loop Closed Param, NOT Idling Prop Step Size Gain Factor .vs. Air Flow

The factory settings are 1 for all airflow values.

Based on that thread, I took my injector size, 30lb, and divided by the stock 24lb, resulting in a factor of 1.25. I then divided all the values in that table by 1.25, resulting in values of .8 across the board.


Holy smokes.... I cannot believe how much more stable and predictable the WB AFRs have become, and how precisely tune-able the closed loop AFRs have become using the swing points. I was able to pull the swing points back down to nearly factory settings for DA3 LT1s, but just slightly richer for the performance benefits. Rather than being in the 650 to 850 mV range, I'm now down into the 500 to 650 mV range. I watch the WB during the same kind of non-PE moderate acceleration and the AFR's hang right at 13.5:1 ± about 1/2 a point. It's crazy...

Like was said in that thread, the ECM was likely chasing itself due to "inadvertent" over-corrections on the fuel mixture because of the much larger injectors.

Bottom line...

We've all been focusing on the injector constant and the voltage offsets as what change when changing injectors. But this proportional gain table is every bit equally as important to change as well.


There's another table mentioned by RBob...

Loop Closed Param, CL o2 Proportional Gain Counts .vs. % Adjusted SLOW o2 Voltage Error

But I haven't played around with that yet as I haven't fully understood it yet... going to continue reading.

 

Could you post the address for the above?

I am talking about 0x???

The reason is, the XDF title may differ from XDF to XDF. I use my own custom XDF. The only way I can retrieve anything is by using the address, not the title.

Thanks

 

$84CF & $84DE

I think its these except replace the $8 with 0X

 

Thanks
Like usual I figured it out after asking the question lol

 

Ok, so a question after reading some more...

I understand how the proportional gain works from reading RBob's explanations.

My question is, can I bias the settings such that it corrects more on the lean side, but less on the rich side?

Example... if I'm trying to attain 13.5:1 nominal AFR during moderate non-PE acceleration and the the AFR dips to 13:1 or something..... for maximizing non-PE performance, I'd want to tell the ECM that I'm cool with that and to not correct so much the other way.

However if the AFR climbs to say ~14:1, I'd want the ECM to do something to get it back to my goal of 13.5:1 ASAP.

We have the Fast O2 R/L threshold vs MAP and then the tolerance bands around that with the SLOW o2 Idling Rich/Lean LOWER Boundary .vs. MAP and the SLOW o2 Rich/Lean UPPER Boundary .vs. MAP.

If I tightened the lower tolerance and raised the upper tolerance, would that mean the ECM would allow for slightly more rich mixtures than lean mixtures? Like if I allowed for say 25 mV below the nominal, but up to 50 or 75mV above the nominal?

Looks like GM went ±44mV above and blow the nominal for the O2 windows on AUJP. So I'm wondering if the code works the way I hope it does for a more performance oriented fuel correction.

I understand that mV on a NB O2 doesn't translate into specific AFRs... rather I'm just trying to generally err on the side of slightly richer side of Stoich for non-PE performance vs leaner. Now that I'm finding the average AFR's so much more responsive to O2 swing-point adjustments, I'm wondering how much more I can capitalize on this.

 

I changed the gain mult vs flow ( OX4CF) to account for my larger injectors, and noted a more stable WB reading, at 14.7 or so. I then changed the step size vs error ( OX4DE) Now I'm running very lean, 15.5 or so and the blm's are uniformly high, 134 or so. I noticed in the CL BLM/O2 all of the floating point tables allow for decimal point entries except for OX4DE, which is set for zero. Is this an error in the def.?
Thanks for your patience guys

 

Curious...

Is there a way to run two WB O2 sensors into the 7730? I.e., is there another unused circuit in there that can be mapped into TunerPro?

I was thinking of replacing the NB with a WB, temporarily running open loop for a bit just to check the AFR balance between the two sides of the engine.

What can I say... I'm just looking for something to tinker with on the car... lol...

I suppose the other possibility is to run both WB's into the same harness and splice in a DPST switch to be able to switch back and forth while driving, and be able to datalog both sides separately and compare later. The two sides would be logged separately, but I suppose better than nothing. Though it'd be nice to have both sets of WB data for the same exact time.

 

I'd personally wouldn't want to rely on the ECM to correct for anything. I agree in terms of being happy with 13.1 AFR at part throttle, and wanting O2 correction if and when it ever reached 14.1 AFR, but the question here is why is it even reaching 14.1 AFR in the first place? Even during Boost, most will ask if the code will correct for AFR during Boost in case it dips to 12.1 AFR, when it doesn't, you have to tell it to... and even if it did, I would never ever wait for the system to correct for it. Correction should be looked at as a last resort, you want zero correction, and VE immediacy. If the AFR deviates, the VE is wrong, or there is a mechanical issue. The other parameters are just band aids. I have a tremendous amount of respect for the guys you mentioned, but this was during a time when members were trying to get modded engines to run off of a code designed for a specific engine, with a range limitation set for 19# and 22# fuel flow. Some went Open Loop and were happy because this allowed the tuner to cut to the chase, others realized fuel needed to be adjusted proportionally and quietly corrected said areas and went on their way, while others blamed the code and went another way out of frustration altogether...

JMO of course.

- Rob

 

I've never run full open loop before, so who knows... if it runs well and the AFRs are even more stable that way, I may just run it like that permanently. At that point, it'd be actually cool to always be able to see AFR on demand at both sides of the engine. I suppose I can always weld a second O2 bung on the NB side, but that may be overkill.

Obviously fuel economy is not the priority here... lol.

 

I have Bosch EV14 35# injectors (they hit 82% duty cycle at 6000 rpm's) and I have changed everything in the in bin that uses pw in the formula. I've been running always open loop for about five years now. My combination is happy with 13.5 at idle and tapering off to 14.7 - 15 after 1500 rpm's. It runs so good that I quit trying to improve it. Hiway mileage is about the same and slow city driving it has increased a bunch because it runs so much smoother at low rpm's. I have a wide band with a gauge in the dash and it always reads where I want it. If I could find a way to keep my EBL from going closed loop only after 1500 rpm's I might try closed loop again. But like my Father used to always tell me, "if it ain't broke don't try to fix it".

 

Very healthy engine...

That's about 425-RWHP @ 82% DC, as those injectors flow about 38# @ 43.5-psi...

- Rob