S10 L31 going lean in PE mode

[E55Amg]

I don't have the log/tune files with me at the moment, but wanted to solicit thoughts. I'm VE tuning a 96 L31 I put in my 99 S10 ZR2 right now. I followed the generic instructions for tune setup, and have a few drives logged. It feels quite powerful at low and mid throttle, but goes very weak on the top end. Looking at the logs, the NB O2s show extremely lean at the exact moment of transition to the PE mode (I have set to enable at 65% throttle w/no delay). As soon as I let off the throttle to get below the enable, the O2s come right back into the switching zone.

I have considered air leaks upstream of the O2s, but they seem to switch fine just below that point, as well as at mid and low-throttle driving. Also wondered what kind of output the stock V6 fuel pump has, but I keep going back to the fact that it occurs EXACTLY at the PE switch point at varying RPM (it will do it at low RPM if I go heavy throttle).

Will post tune/logs later this evening if anyone is interested.

Jon

[cobaltssoverbooster]

nb o2 shut off under pe you can only tune wot when using a wideband. it is highly recommended to install one before making any more wot pulls

[E55Amg]

I was not aware that O2 sensors could or would bottom out just from PE mode entry... any value to changing the target stoich in the loaded areas and turning off PE mode so I can see?

Also read different year L31s had different injector flow rates. The tune I based mine off of says 23, but some say 19 is the number. This might explain my heavy underfueling...

I don't see an attach option for my files??

[E55Amg]

99S10hybridfile2.hptnewrun.hpl

Nevermind, I was in some sort of "lite" mode for the forum...

[2xLS1]

Saying O2s shut off during PE is kind of misleading. They just stop providing feed back. They still operate and you can look at what their voltage output is to see a trend what your PE A/F is. They will just not give you an accurate A/F number. But you should see them in the 900-950 MV range generally in PE. I haven't looked at the log file but no they shouldn't bottom out.

[E55Amg]

Saying O2s shut off during PE is kind of misleading. They just stop providing feed back. They still operate and you can look at what their voltage output is to see a trend what your PE A/F is. They will just not give you an accurate A/F number. But you should see them in the 900-950 MV range generally in PE. I haven't looked at the log file but no they shouldn't bottom out.

That was my understanding... but they are in fact bottoming out (0.050mv range). That is the strange thing about it, is that I'm nowhere near maxing out the injectors (4000rpm at 100% tps showing just over 50% duty cycle), but the moment the TPS tips into PE territory the O2s drop from regular cycling all the way to the bottom. I assume they'll stay there, but I'm not willing to test that out really. I'm wondering if it is a combination of poor VE mapping and the injector constant being too high. I still have the black box from the original 96 truck, and it would run the engine, so I might plug that in and see what I can see. I haven't tried reading or logging from it, so don't know if I will be able to see anything at all.

[2xLS1]

Put a fuel pressure gauge you can see while driving on it. I bet you are losing FP.

[E55Amg]

I'd think that too, but with it being RIGHT at PE mode every time, and it doesn't seem to matter what RPM (it'll do it at 1500rpm or 3500rpm), I'm not sure. I do have a fuel pressure gauge, might be interesting to rig it up to be visible though..

[Fast4.7]

I don't have the log/tune files with me at the moment, but wanted to solicit thoughts. I'm VE tuning a 96 L31 I put in my 99 S10 ZR2 right now. I followed the generic instructions for tune setup, and have a few drives logged. It feels quite powerful at low and mid throttle, but goes very weak on the top end. Looking at the logs, the NB O2s show extremely lean at the exact moment of transition to the PE mode (I have set to enable at 65% throttle w/no delay). As soon as I let off the throttle to get below the enable, the O2s come right back into the switching zone.

I have considered air leaks upstream of the O2s, but they seem to switch fine just below that point, as well as at mid and low-throttle driving. Also wondered what kind of output the stock V6 fuel pump has, but I keep going back to the fact that it occurs EXACTLY at the PE switch point at varying RPM (it will do it at low RPM if I go heavy throttle).

Will post tune/logs later this evening if anyone is interested.

Jon

Stock V6 pump will easily handle a L31 with a factory intake manifold. The pump itself if it is in good shape has a higher capacity than the 8 injectors do.

The MFI injectors are 22.1 lb/hr and the stock poppets 23.1. The difference is the fuel pressure the regulators for each are set at. Putting a MFI spider in place of the poppets will cause the fuel trims to go positive on an otherwise stock L30 or L31. Bumping the regulator on the MFI up to the Poppets factory 66 psi fuel pressure fixes the issue. MFI spider was set at 60 psi.

Which MAF are you using? The V6 and V8 are different sizes and have different calibrations. The V6 MAF is the same 3" MAF used on the B-car LT1s. The V8 MAF is the 3.5" used on the F/Y car LT1s.

Is this a stock L31? Does it have headers and exhaust?

I have found even a decent set of headers and exhaust will lean the factory tune out when running the black box. The VE tables are used heavily in conjunction with the MAF on the Black Box and both need to be corrected for it to run well.

I would move the PE enable up to about 90% throttle and disable the long term fuel trims. Disable the MAF by setting the fail frequency to Zero and the high maf value code to fail on the first failure. Tune the VE tables until you get the fuel trims in the -5 to 0 range across the board. Then enable the MAF and tweak only the MAF table until you get the fuel trims back into the same range. I have a scanner definition file I setup to log average short term fuel trims vs rpm vs map. I also have one for short terms vs maf frequency. I use an Excel spreadsheet to automatically correct the tables. Then smooth the areas that were not reached. I have found on numerous L31s they like the narrowband o2 sensors in the 930 mv range with the WOT timing about 28-32* by 4,000 rpm to make peak power. I had my Express van tuned this way for 5 years before I even put a wideband on it. WOT air/fuel mixture was practically spot on tuned this way. 12.2 on tip in, 12:1 at peak torque leaning out to 12.6 @ 4,600 @ max hp.

[Fast4.7]

One thing I also meant to ask. Have you made changes to the PE Multplier? In stock for it should be in the 1.22 range IIRC through a lot of the RPM range. With the Stoich set at 14.68 that is right at 12:1. If you are running on fuel with 10% Ethanol you should change the Stoich setting to 14.1 prior to tuning the VE and MAF. On a modified setup that is running as lean as yours I also find it easier to multiply the whole VE map and multiply it by 1.1 to add 10% to the fueling. Safer on the engine to start by running rich and bring it leaner.

[E55Amg]

I currently have the V8 MAF installed in line, but unplugged (vice fail freq to 0). The L31 itself is stock, has standard S10 V8 mid-size headers to fit the steering through the primaries, to a crappy y-pipe and through a fake flowmaster 40

I was looking at using the LTFT VE table tweak method, but have seen the STFT method. I was going to shoot for a target of -2 to -3 (with a focus on maintaining it slightly negative under all conditions). The MAF I'm a bit confused on it, was giving me some really weird indications (like almost no hz change going from 60% tps to 100% tps). I would LIKE to use it, but if it's not going to be consistent then I'm ok with an SD tune. For some reason my spark table shows up to 34 degrees of advance, so that's something I need to work on as well.

[E55Amg]

I set the PE multiplier to 1.2 throughout the range, but have not adjusted the stoich setting. I run 91 octane all the time, is it 10% ethanol as well? I DID already add 10% to the VE, which seems insane because it still says I'm running pretty lean. I'm wondering about exhaust leaks but seems strange with both banks behaving similarly.

[Fast4.7]

I currently have the V8 MAF installed in line, but unplugged (vice fail freq to 0). The L31 itself is stock, has standard S10 V8 mid-size headers to fit the steering through the primaries, to a crappy y-pipe and through a fake flowmaster 40

I was looking at using the LTFT VE table tweak method, but have seen the STFT method. I was going to shoot for a target of -2 to -3 (with a focus on maintaining it slightly negative under all conditions). The MAF I'm a bit confused on it, was giving me some really weird indications (like almost no hz change going from 60% tps to 100% tps). I would LIKE to use it, but if it's not going to be consistent then I'm ok with an SD tune. For some reason my spark table shows up to 34 degrees of advance, so that's something I need to work on as well.

The spark tables show values up to the 38-40* range in the tunes I have looked at but at WOT you will be in the 90+ KPA range unless you are at high altitude. The high map values and the PE adder are added together to get WOT timing.

[Fast4.7]

I set the PE multiplier to 1.2 throughout the range, but have not adjusted the stoich setting. I run 91 octane all the time, is it 10% ethanol as well? I DID already add 10% to the VE, which seems insane because it still says I'm running pretty lean. I'm wondering about exhaust leaks but seems strange with both banks behaving similarly.

Does not sound out of line to me. I had to put an extra 20% in places on the 99 Suburban I tuned with a F-car LT1 cam, headers, and CAI.

[cobaltssoverbooster]

Saying O2s shut off during PE is kind of misleading. They just stop providing feed back. They still operate and you can look at what their voltage output is to see a trend what your PE A/F is. They will just not give you an accurate A/F number. But you should see them in the 900-950 MV range generally in PE. I haven't looked at the log file but no they shouldn't bottom out.

i just dropped a general understanding to try and get the point across quickly that the narrowband is not going to tackle the task.
yes, the ecu still reads the NB voltage but it does not monitor the NB under PE or BE events which is why the NB trims go static. Under these modes the ecu runs the commanded PE value based on the SD or MAF airflow calibrations (whichever mode is active). It is not suggested to read NB voltage to determine afr because it is only sending a signal of 0-1V range which can cover between .8 and 1.2 lambda (low resolution compared to a widebands 0-5v signal range). The NB sensors also aren't linear responsive sensors and vary their output drastically on the temperature of the sensor which is typically not regulated. Why are they cheap, because they dont usually have heater temperature feedback circuits and they aren't linearly calibrated since they rely solely on a reactive element to generate a signal source when it reacts with the oxygen left in the system. I could go on explaining it but AEM does a better job in their 30-4100 Instruction manual so ill leave it up to the reader to go learn whats really go on in depth.
Yes, i made a partially wrong statement but you shouldn't even be relying on the NB sensor unless its for trim related uses under stoich control which is why i just told him they dont work there.
Make the investment to not risk the cost of failure.

Hopefully you know what i know about the sensors and can understand why i stated it the way i did.

[cobaltssoverbooster]

I set the PE multiplier to 1.2 throughout the range, but have not adjusted the stoich setting. I run 91 octane all the time, is it 10% ethanol as well? I DID already add 10% to the VE, which seems insane because it still says I'm running pretty lean. I'm wondering about exhaust leaks but seems strange with both banks behaving similarly.

I hope im not misunderstanding your post but the difference from pump gas to e85 is a 30% increase in flow for ethanol over ethanol free pump gas. typically the 10% blend isnt even worth correcting for as the pump doesnt have to guarantee a 10% alcohol content. Federal law states that E10 cannot exceed 10% alcohol content and because of this most pumps typically test somewhere around 7-8% in my local area. Its not a guarantee just a max rating. Because the range of pump E blend can vary so drastically and the alcohol content is so low in most cases, the effects of the alcohol on fueling error are minimal and thats why people just leave the stock 14.68 stoich value even though they pump E10 fuel.
if i misunderstood your post thats my bad.

[Fast4.7]

I hope im not misunderstanding your post but the difference from pump gas to e85 is a 30% increase in flow for ethanol over ethanol free pump gas. typically the 10% blend isnt even worth correcting for as the pump doesnt have to guarantee a 10% alcohol content. Federal law states that E10 cannot exceed 10% alcohol content and because of this most pumps typically test somewhere around 7-8% in my local area. Its not a guarantee just a max rating. Because the range of pump E blend can vary so drastically and the alcohol content is so low in most cases, the effects of the alcohol on fueling error are minimal and thats why people just leave the stock 14.68 stoich value even though they pump E10 fuel.
if i misunderstood your post thats my bad.

I do not do that. I change to 14.1. I learned that way back fighting with crappy running TBI systems nearly 15 years ago when E10 hit our local pumps.

7-8% ethanol gives a substantial fueling error when the long term fuel trims are lean and you go WOT.

[Fast4.7]

i just dropped a general understanding to try and get the point across quickly that the narrowband is not going to tackle the task.
yes, the ecu still reads the NB voltage but it does not monitor the NB under PE or BE events which is why the NB trims go static. Under these modes the ecu runs the commanded PE value based on the SD or MAF airflow calibrations (whichever mode is active). It is not suggested to read NB voltage to determine afr because it is only sending a signal of 0-1V range which can cover between .8 and 1.2 lambda (low resolution compared to a widebands 0-5v signal range). The NB sensors also aren't linear responsive sensors and vary their output drastically on the temperature of the sensor which is typically not regulated. Why are they cheap, because they dont usually have heater temperature feedback circuits and they aren't linearly calibrated since they rely solely on a reactive element to generate a signal source when it reacts with the oxygen left in the system. I could go on explaining it but AEM does a better job in their 30-4100 Instruction manual so ill leave it up to the reader to go learn whats really go on in depth.
Yes, i made a partially wrong statement but you shouldn't even be relying on the NB sensor unless its for trim related uses under stoich control which is why i just told him they dont work there.
Make the investment to not risk the cost of failure.

Hopefully you know what i know about the sensors and can understand why i stated it the way i did.

I do know the difference, but I also happen to know that 930mv works very well as a target WOT narrowband sensor reading on TBI, TPI, LT1 and CSFI Vortec engines.

While a wideband is better you can certainly tune with narrowband only and have a great running engine that has a tune that is not very far off the mark.

Then again I started tuning with carbs and NO air/fuel ratio feedback at all other than plug reads, sound and seat of the pants.

[E55Amg]

While I appreciate any and all input, it was a misleading statement. I'm new to Vortec tuning and V8 S10 swaps, but not tuning or EFI itself. I am aware of how an O2 sensor works (several versions thereof), but not if any particular system blanks out an O2 when going into PE. In my case, it is useful to know that the lean indication is, in actuality, a real lean, and not a false lean provided by the OS dropping O2 sensor voltage when entering PE mode. Which sounds as reasonable as some of the other interesting engineering/programming I have seen in other applications, so not out of the realm of possibility, but also not true. Ignoring the readings is a very different prospect than zeroing the readings. I also know that I'm sensitive to the use of particular words (occupational hazard), so the misunderstanding is just as much mine as anything else. Thank you for the advice.

[2xLS1]

Never mind the fact that most stock GM truck calibrations have anywhere from .8 second to 60 second time delays before they will enter PE, you will hear that entering PE without a wide band O2 you will blow your shit up. LOL.

OP check your fuel pressure at the rail.