When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Started another thread which is a result of the ARAPWB thread that I was discussing. I now have Wideband visibility in my ALDL stream so I can correlate more.
I did a few runs on a 1/2 mile decommissioned airstrip just to get some datalogs which I can send to anyone willing to help out.
- I can say that the issue is that I am lean for most of the time with an average AFR of 16.5.
- The lean problem only happens as soon as the car goes in closed loop.
- Coolant temperature would be around 180. Battery voltage around 12.7V.
- In open loop I am at an average of 13.5 AFR at 40% throttle. (Didn't try more throttle during open loop.)
- When I fully depress the throttle (around 86% of the 5V signal) AFR is quite good. It is around 13 AFR with 180 MAF, 237 LV8 and 128/128 BLM/INT..
- At idle AFR is around 15.
So I can say that the problem occurs during closed loop and not at full throttle. I think this is called Part throttle. I cant really figure out the BLM and INT values as they are quite inconsistent during the lean, part throttle condition. What I saw is that in the 80% TPS regions, BLM and INT stop refreshing sort of it stops using them?
Would it make sense to lower the injector constants as during WOT, it is fine? Should I touch the MAF tables increasing the values till the 130 MAF figure? I also saw a program called AFRTuner which I managed to get working on an old XP VM but it modifies only based on RPM and cooland temperature. My coolant temp is quite steady at 178 - 180.
So if i understand you correctly.... you say it seems to be closer to 16.5 afr during most rpms/LV8 loads while normally driving? 0-60 ish % throttle regions?
wot doesnt use closed loop.
If my assumption is correct, you will want to add to the maf tables to get afr in closed loop closer to 14.7 or whatever’s you want. That should be closer to blm of 128, depending what the o2 swing point mv values are. Typical 450-550 mv
after you do that, open loop may be abit richer so back off the open loop afr change vs coolant or vs lv8. There should be two tables you can use
Didnt know that WOT doesnt use close loop. What would be considered WOT? Anything above a certain TPS %?
I am using stock ARAP MAF tables. I read that these tables should be scaled up so that there is an overlap between all the 6x tables. When you say add to the maf tables, what are we looking at? Mine start 8.3 till 22.2 in table 1. Would I need to add a % in all of them? If it is the case, would decreasing the injector constant from 22 to say 19 have a similar result?
Sorry never did this in reality. I always stopped till the theory here.
Re. open loop, the AFR % change vs Coolant is all 0 above 68 Celsius (150F) so if it is rich due to this table it will be only for a few minutes.
Thanks for the information. Always nice learning new stuff.
I believe the blm int no longer update when in power enrichment mode which is based on min lv8 and min tps settings. Arap is likely 50 lv8 min and 70% tps min above 3600 rpm.
You will want to note where you are lean in part throttle and the grams maf reading at that point. Say 2000 rpm its 16.5 afr at maf grams sec of 30. Go to your tables and see where 30 would be, in stock arap thats table 2 around the 1.74 volts cell. I would increase that cell a few % at a time like say try 5% increase. May even increase the cells around it some to smooth out the shape. That adds fuel
open loop table may read 0 but still be too rich so you are allowed to use negative values to decrease fuel
Using my datalog and some monitor charts, I can see that during closed loop anywhere below 120 Airflow is all running at 16AFR irrespective of which RPM I am in. In this datalog I was between 800 till 3600RPM.
That is all until I hit 60% throttle and 210LV after which AFR go suddenly down to 12.8AFR which I think is a good number for WOT.
You can see the graph attached showing the sharp drop in AFR with %TPS, LV8 (Yellow), RPM and Airflow.
Does this mean I have to try increasing all values in the first 4x MAF tables?
You're actually lean everywhere compared to stock. A stock $6E would be about 10.8:1 at WOT. Not that 12.8 is particularly problematic - I would be aiming for about 12.5 myself. But with the ugly lean tip-in and ultra-lean cruise you may go into PE with a very hot cylinder and that will promote detonation.
Thanks for your input GD. Just to clarify, the WB goes up to max after the WOT session due to myself completely releasing the throttle. You can see the rpm % in red.
Also I dont get any pinging due to detonation. At least not noticeable.
Was looking at an article about the 7730 ECM related to AIR Diverter where the ECU subtracts 100mV from the O2 to compensate for the additional diverted air in the exhaust when AIR is activated. If the AIR pump and diverter valves have been removed (as also is my case), the engine will run lean as the 100mV would still be decreased from the O2 readings.
The problem is that I couldnt find a similar article for 6E on 165 ECM to make sure that AIR does not kick in or to stop the ECM from applying the 100mV decrease in O2.
Could this be the reason why I am running lean while in closed loop?
I also removed EGR (Block off plate) but that aspect I handled it by doing the below changes in the PROM:
EGR Enable if TPS > X & EGR Duty Cycle!=0 changed from 5.86 %TPS (0x0F) to 99.61 %TPS (0xFF).
EGR Enable if TPS > X & EGR Duty Cycle=0 changed from 6.25 %TPS (0x10) to 99.61 %TPS (0xFF).
Disable EGR if MAT < X changed from 0.50 Deg F (0x1E) to 304.25 Deg F (0xFF).
Last edited by aseychell; 02-04-2020 at 07:59 AM.
Reason: Revised the comment on Disable EGR if MAT < X
Disable egr if mat is less than 304 will disable egr
o2 milivolt values will control target swing/switch point for closed loop so it may want to correct for a leaner condition. You can change the o2’s around in the expanded 6e xdf
i think its called diff value to make up rich lean when divert air. Set to 0.102 v. Change that to 0
you could also try the fast and slow rich lean o2 mv offsets vs airflow. Might help some
Theres a couple of divert air constants/scalars you could also use to eliminate the AIR function i believe. I used if temp < thresh, divert air. Factory set to 14.75 i changed to 144. Or use if o2 > x divert air
I had in mind to modify "if temp < thresh, divert air" from 14.75 to something high but thinking about it, putting this higher would make the system try to divert Air all the time as we will be below 144 Celsius.
Is my reasoning right or am I screwing up the logic here?
Divert air is what you want. It diverts it to atmosphere and not into the exhaust. You want the ecm to think its going to atmosphere all the time so no correction is made
$6E was in the upper 10's when I used it on my '86 TPI 305. When I asked about this here on TGO the general consensus was that this was fairly typical of factory GM maps from this era. But admittedly I haven't tested a whole bunch of them.
$6E was in the upper 10's when I used it on my '86 TPI 305. When I asked about this here on TGO the general consensus was that this was fairly typical of factory GM maps from this era. But admittedly I haven't tested a whole bunch of them.
GD
yeah i never tried $6E on a non 6e based motor but not sure it would matter. My 89 5.7 was low 12’s stock motor with exhaust on the dyno. Another 89 with cam only swap with exhaust was in the 12’s if I remember right on base bin tune. I didnt have to do much with wot fuel. It didnt make much power over a normal bolt on L98. Surprisingly. The 87 32b i did heads and cam on the base tune was pretty close, so maybe some are richer. I just dont see why gm would have it that rich. I have heard newer motors are really rich wot
Went to test it today at the airstrip. The change on the AIR divert (increasing temp to max and removing the 100mV offset) did not produce any changes whatsoever.
I then tried modifying the MAF tables in steps of 5% until 15% and I honestly cant see any considerable differences in AFR. It might have increase a bit the power although I am not 100% sure. It surged a bit at a certain point but it was only for a couple of seconds. What is worrying me is that AFR is still swinging in the 15.5 - 16 regions no matter the changes I am doing.
I think it is fair to assume that fuel pump, filter and / or injectors are fine as when the necessary variables reach a situation for WOT (I think LV8 at more than 200 and TPS at more than 70%), the AFR goes down to 11.5. When engine is open loop, AFR is also fine as it would be around 13.5 - 14. Seems like some parameter in closed loop is forcing the ECM to supply less fuel.
You can see graphs for one of the runs while I was at 15% increase across all MAF tables.
This is with 15% increase across all 6x MAF tables
This is stock ARAP with the AIR modification
Do you guys have any recommendations re this issue?
Should I give up on $6E and switch to SD? I almost have all the necessary parts by now....
Would it make sense to use my Wideband's simulated Narrowband signal in place of the stock non-heated O2 sensor? I am exploring this as I am running headers and the factory O2 might not be ideal for such a setup.
Anyone have experiences with using a simulated NB signal from a WB?
Plotting the NB voltage, for sure I can notice that it is in line with the WB signal just the inverse of it as apparently NB voltage goes high for rich and low for lean. So as far as functionality and realistic view, the NB seems to be reading fine when compared to the WB. Voltage swings between 50mV to 950mV (Purple line on graph).
Something is not right about the correlation of the WB and NB sensor values. They should cross each other at about 14.7 for the WB and 450 mV on the NB. At the time of crossover the WB is reporting lean and the NB is reporting rich.
Note that if you are trying to change fueling during closed loop, the MAF tables won't do it. Need to use the O2 window tables. They are what the ECM uses to target the desired AFR via NB O2 sensor feedback.
BLM tries to follow INT and INT is in the regions of 105 when the AFR reads leanest. I dont have access to my PC now as I would have sent the monitor with BLM and INT plotted... will do so as soon as I go back home.
With regards to where NB and WB cross eachother I would guess that sampling time between the two would skew where they intersect?
I also didnt know that MAF tables affect only Open Loop. To be honest I noticed it was a bit richer during open loop.
Any tips you can give me on the O2 window tables? Would really appreciate it.
Actually if you were to ‘imagine’ where WB @ 14.7 would intersect NB it would be in the regions of 450mV. As I never managed 14.7 AFR in CL it isnt shown in the graph... the only place where AFR drops below 14.7 is during WOT and as there is a sharp drop NB reading cant be relied on I think.
I was gonna say maf will tune so corrections dont have to be as much to reach target afr, but the o2 swing points are what sets target afr, but on ARAP i dont think they would be that lean. If i recall the swing points were in the mid 500’s mv range which is on the rich side of stoich
What are these swings points? I know there are slow and fast O2 min and max scalars. They go to 64gm/sec if i remember correctly but I have been looking at some more info on this aspect from the time Rbob pointed this out but cant find info. All I can find is Maf tables, Maf scalars and injector constants.
Attaching below the same run with BLM and INT graphing.
Note BLM and INT are frozen at 128 during WOT. As for the rest they swing between 110 and 148 with INT going down even to the regions of 90 while AFR swings in the regions of 16 - 16.5...
Airflow, RPM and TPS all follow a steady climb.
BLM = Thick Red, INT = Thick Blue, AFR = Black, NB Volts = Purple, Yellow = LV8
Theres 4 tables. Closed loop rich lean vs airflow. Upper and lower zero ref for slow o2. And Fast o2 rich lean threshold vs airflow.
not exactly sure which ones are what been awhile since i looked at these, but you can try adjusting them. I believe raising the values will lower the commanded afr
Fast O2 was set to 400 mv for this test with upper and lower 450 mv/350 mv respectively. Note: there is still a bit of a rich bias overall with respect to target.
Reducing the Fast O2 threshold from 540 mv to 400 mv only reduced the average afr error by about 1/2 of 1% from -1.2% to about -0.6%. More simply put, the running average WBO2 (with injector pw >= 1 ms) only moved about a tenth of a point from 12.8 to 12.925 vs the same 13:1 target afr.
Conclusion: the O2 thresholds have a very minor influence on the actual afr.
Last edited by tequilaboy; 02-09-2020 at 02:04 PM.
why are you suspecting a bad sensor? The graph you sent has a very similar graph to mine. Afr and nb signals are sort of inverted. NB range is between 50mV to 900mV. Its just your overall AFR is richer. PS i am using also PLX and using normal gasoline.
So if these O2 tables do not influence the AFR is it not possible to do adjustments on CL?
I suspect a bad sensor since the offset between your WBO2 and NB cannot be easily explained otherwise. The NB cycling suggests that your mixture is in the vicinity of stoich while the WBO2 is reading considerably leaner. Perhaps the sensor location is a factor. Are they installed on opposite banks, where a misfiring cylinder could influence one bank, but not the other?
The example data that I've shared is from a C4 with LPE long tubes and y-pipe. The WBO2 is installed in the y-pipe downstream of the NB which is installed in the driver's side header.
Originally Posted by aseychell
Hi tequilaboy
why are you suspecting a bad sensor? The graph you sent has a very similar graph to mine. Afr and nb signals are sort of inverted. NB range is between 50mV to 900mV. Its just your overall AFR is richer. PS i am using also PLX and using normal gasoline.
So if these O2 tables do not influence the AFR is it not possible to do adjustments on CL?
The WB is installed in the collector. The NB is on the driver side back at the edge of the shortie headers. I am tempted to use the simulated nb off my plx instead of the stock o2.
Analyzing a bit more my datalogs in an attempt to find something bad I can see that the block learn flag in the ALDL stream always shows as ‘disabled’ is this normal? What does this do?
Yesterday I found a 30 minute chance to try the car with the highway fuel disabled. I was suspecting maybe the ecm is going into highway mode earlier but unfortunately it wasn't the case. Am still running 16AFR during part throttle in closed loop. I have also scraped all the internet searching for some explanations on the O2 tables mentioned earlier to no avail. Also couldn’t find the reason why that block learn stays disabled all the time.
I was considering using the simulated NB from my PLX WB controller but from the waveforms on the datalog I can see that the stock NB O2 sensor is working fine with its average waveform value in the regions of 450mV and O2 cross counts are also occurring all the time. Don't want to mess us the wiring for nothing.
At this point the only option if no one else can help out on richening during CL is to either tempt my luck by going through a 7730 conversion hoping I hit the issue during the upgrade or else disable CL and fine tune MAF tables to get 14AFR for OL.
Actually to be 100% correct, during a 4 minute average, O2 mV are at 310mV and INT Average of 129 which should indicate a lean condition (310mV) yet no reaction from the ECM to try and add fuel (129 INT). I know that using Averages might not be the ideal measure but in such a short sample window (4 min) the average of 129 should indicate a stable AFR closer to 14.
To be honest I tried searching all over the net on what the O2 tables are but didnt manage to.
There are the following table in $6E ARAP and I cant figure our what to change:
1 Upper Zero Error Ref for Slow O2 Rich/Lean
2 Lower Zero Error Ref for Slow O2 Rich/Lean
3 Fast O2 Rich/Lean threshold vs Airflow
4 Low O2 Filter Coeff vs Airflow
5 O2 Error reduction Gain vs Airflow
6 Slow O2 filter CoEff vs Airflow
See Attachment. Dont know what Fast / Slow mean in the first place. Is it related to before its reaches the proper temperature or not?
Not sure what is causing your issue as i havent seen a stock code target that lean at part throttle
are you sure the sensor is reading accurately and no air leaks in the exhaust that could be skewing data?
if no leaks i would try maybe adding 100 mv to the top 3 tables you have posted. See if that does anything. Or atleast to the grans sec values where you are lean.
Went for a run today. Decided to load the MAF Scalars, MAF tables and related INT and BLM Tables / Scalars from an AUJM. The reason for this is that I noticed that AUJM has higher numbers in its MAF tables for Voltage vs Airflow thus it made sense to me that it should run richer albeit lately I learned that MAF tables are significant only in Open Loop. Anyway I gave it a shot but sadly enough it didnt really make any difference. To be honest I think that with the AUJM map, OL is a bit leaner even though MAF tables have higher values...Granted Scalars are also different thus it might be the reason for this. Closed loop still in the regions of 15.5 to 16.5 AFR.
So I then decided to fiddle around with O2 tables as suggested by Orr89RocZ starting with Upper Zero Error Ref for Slow O2 Rich/Lean changed. Increased by 5% > no change, increased by another 5% > no change, another 5% > no change. Total of 15% overall without any effect.
I then increased Lower Zero Error Ref For Slow O2 Rich/Lean changed by 10% in two steps but still again no noticeable change.
Increasing also Fast O2 Rich/lean Treshold vs Air Flow by 10% didnt result in any difference.
None of the O2 tables enriched or en-leaned AFR... The ONLY time that I see a decent AFR of 12.5 is when Airflow exceeds 130 grams and TPS is around 55 - 60%, LV8 around 220. I am assuming this condition results in WOT.
So now I am back to stock ARAP with my only way forward being either stop using closed loop and run all the time in open loop, or else embark in an SD conversion and hope for the best. Not sure if attempting to use the simulated NB signal from my WB controller instead of the factory O2 would make a difference.
I am still sinking in the information but I might not have modified the value enough when I added 15%. With 15% I would have practically moved the O2 sensor stoich mV value from circa 575mV to 660mV. This might not be enough and might be attributed due to the headers, lack of catalyser and big 3" pipes till the back making the sensor 'think' that mixture is right. An NB O2 indicates rich when voltage is high and a gasoline O2 would reach stoich at circa 450mV. Modifying these table seems like telling the ECM which mean mV value to accept from the sensor as stoich and if not, modify using INT and BLM accordingly.
So in my next test I am going to raise the average to circa 800mV and see what happens.
Furthermore, as you can see from another ongoing thread, I currenty have an issue with cooling fans and remembered that my tstat is at 180. My fan is kicking in at 180 and never goes off. Temperature never reaches more than 186. So there is another table related to O2 called 'Rick/Lean offset vs coolant'. Our cars are designed to run at circa 200 - 210 and funnily enough this table drops from 16 Offset to 0 at the 176 Temperature boundary.. From what I can understand, these offsets SUBTRACT from the 3x O2 tables thus lowering the mV value resulting in leaner AFR. This makes my situation even worse as I never reach 198 Temperature which is the row with 0 offset.
If all the above is correct, due to the TSTAT change I should change the offset at row with temp 176 to 0 so that I dont get any offset during my normal temp.
You have an instrumentation error. Pure and simple.
Either the NB is wrong or the WB is wrong, or they both are wrong (unlikely).
NB sensors are robust and have been well-proven in automotive applications for 40+ years and are pretty much fool-proof. Sue, they can be subject to error due to lead contamination, air leaks, insufficient exhaust heat, oxygenated fuels and what-not, and don't last forever, but they are pretty damn reliable in-practice with normal fuels. Your posted data does not suggest that anything is wrong with the NB or with the closed loop fuel control.
I have seen data from many cars with both NB and WB sensors and in all cases both sensors have all shown very good correlation to each other. Different sensors, different cars, different fuel, different intakes, different cams, different superchargers, different operating conditions, different climate, elevation and fuel, you name it.
If someone has trust-worthy data to the contrary, I would like to see it.
If you can't verify/validate your existing WBO2 sensor, try another sensor, test your sensor on another car or find a 3rd party (dyno) with an independent WB sensor to collaborate the results before re-tuning to a bad sensor. This is like a trick question on an exam. You have been warned. Good luck.
I agree with you that there is some sensor skewing the INT AND BLM controls and the o2 has a big influence on these. Before I got the PLX I was working with an LM1 (handheld) at the exhaust tip which also showed lean. I got the plx as its a permanent install and I was hoping that the lean condition was due to the placement of the sensor in the exhaust tip.
i find it hard to believe that both systems are faulty. Also to note that the WB does read rich during WOT and the car actually feels weak during part throttle and strong during WOT.
yesterday I also verified that its not a tunerpro conversion issue as the numbers match the plx display identically.
NB is located in the driver side headers just before it connects to the Y-Pipe. Headers are edelbrock.
WB is located in the collector a few inches from where the left and right banks join together.
Well it certainly is possible the wb is reading the average air fuel value of both banks and the narrow is only seeing driver side, so they could be reading different values if you have any air or fuel distribution issues in the engine intake, injectors etc
not as likely but i have seen it on occasion especially if injectors arent flow matched and with some short runner intakes. Could try using the simulated narrow band signal to ecm next to see what happens
02-03-2020, 08:18 AM
