PDA

View Full Version : How accurate is Auto VE?



GMCtrk
March 26th, 2010, 06:38 PM
So, did some logging at the base tune I was running on wasn't bad at all. This is basically just a LQ4 tune with 12.5% added to the MAF table above 5000 Hz. AFR is around 11.8-11.2 at WOT with 24-25 degrees of timing. Thought it was going to be real easy to get it dialed in at WOT, but the AutoVE doesn't seem to be accurate at all. I put it into SD mode, got several really nice logs (blew through 1/4 tank of gas) and the end result was an AFR in the mid 13s at WOT I set it up to run 12.8 at WOT in open loop. It seems like the BEN correction factor is overshooting the target AFR? How accurate have you found the autoVE to be?

I'm attaching some tunes/logs. First is the base tune along with a log for it. The next tune is the SD tune, basically just 12.5% added to the VE table along with 12.8 commanded AFR. The first log for that tune looked okay. Log3 is the result of 2 successive logs with autoVE BEN correction factors made to the VE table. End result was AFR in the 13s. ???

Thanks for any help.

joecar
March 26th, 2010, 08:14 PM
Hi GMC,

Do you have FlashScan V1 or V2...?

What vehicle/year/model and what mods...?

Several things:

- your H02Sx1 voltages show trimming... set B4206 to Disable.

- in C6001 you have P0101,2,3 set to "Not Reported", and in log file AutoVE3.log there is no presence of any of those 3 DTC's... i.e. the PCM is still using the MAF... one of those DTC's has to be present for the PCM to ignore the MAF... in C6001 set P0101,2,3 to "A: 1 Trip".

- at 91% injector DC, you're getting close to running out of injector, keep an eye on this, if it hits/exceeds 100% then tuning won't add any more fuel.

- adding an initial 12.5% to the VE table might be too much (unless you have big mods).


See if you can go thru the throttle range more progressively/smoothly... this will give you more useful data that the transient filter will keep.

If you have a stretch of deserted safe highway, you will be able to correct the VE table in one or two passes... going uphill is good, it loads the engine more consistently.

Be sure to apply the throttle transient filter to the BEN map before multiplying it into B0101.

GMCtrk
March 27th, 2010, 03:14 AM
Hi GMC,

Do you have FlashScan V1 or V2...?

What vehicle/year/model and what mods...?

Several things:

- your H02Sx1 voltages show trimming... set B4206 to Disable.

- in C6001 you have P0101,2,3 set to "Not Reported", and in log file AutoVE3.log there is no presence of any of those 3 DTC's... i.e. the PCM is still using the MAF... one of those DTC's has to be present for the PCM to ignore the MAF... in C6001 set P0101,2,3 to "A: 1 Trip".

- at 91% injector DC, you're getting close to running out of injector, keep an eye on this, if it hits/exceeds 100% then tuning won't add any more fuel.

- adding an initial 12.5% to the VE table might be too much (unless you have big mods).


See if you can go thru the throttle range more progressively/smoothly... this will give you more useful data that the transient filter will keep.

If you have a stretch of deserted safe highway, you will be able to correct the VE table in one or two passes... going uphill is good, it loads the engine more consistently.

Be sure to apply the throttle transient filter to the BEN map before multiplying it into B0101.

Thank you very much for this reply. I'm using flashscan V1. Vehicle is a 2000 fullsize truck, 6 liter, 1 7/8 headers, 3600 torque converter, 4.10 gears.

Regarding B4206, it's interesting they make no mention of that in the AutoVE tutorial. I followed all the instructions there to set up SD mode.

Regarding C6001, I set all of those to not reported because it was causing my transmission to command full line pressure for some reason.

mr.prick
March 27th, 2010, 03:37 AM
You're logging 26 channels this slows the log speed.
Log 24 channels max this will give you a little smoother BEN map for the VE table.
Turn on PE, don't rely on {B3605} for fuel enrichment.
{B3605} can be 1.00 EQ/Lambda @ 176F and below w/PE on.

OL is sensitive to ECT & IAT so don't be surprised if one log is near perfect and
the next is leaner/richer than the last.
Using LTFTs for tuning on the streets is easier IMO.

GMCtrk
March 27th, 2010, 07:00 AM
Got it all figured out. Thanks a ton Joe.

Here's a log after a couple revisions today. Looks like the stock injectors are maxed out.

joecar
March 27th, 2010, 11:32 AM
Thank you very much for this reply. I'm using flashscan V1. Vehicle is a 2000 fullsize truck, 6 liter, 1 7/8 headers, 3600 torque converter, 4.10 gears.

Regarding B4206, it's interesting they make no mention of that in the AutoVE tutorial. I followed all the instructions there to set up SD mode.

Regarding C6001, I set all of those to not reported because it was causing my transmission to command full line pressure for some reason.For headers, 5-8% initial increase on VE table is more suitable.

The LS1 Camaro/Firefird/Corvette calibrations all had B4206 disabled from the factory, these were that were being tuned back then, that's why the tutorial never mentioned it. :doh2:

In tables E0102, E0103, E0104 disable P0101, P0102, P0103... (they are midway down those tables).

Again, the tutorial doesn't mention these because they were already disabled.

joecar
March 27th, 2010, 11:36 AM
mr.prick has a good point on 24 channels or less...

upto 24 channels: pid update rate is 10 Samples/second
over 24 channels: pid update rate is 5 S/s or less.

GMCtrk
March 27th, 2010, 11:39 AM
Is it just me or does increasing timing really affect AFR? I added 2 degrees at WOT and it went lean (12.8 to low 13s).

joecar
March 27th, 2010, 11:44 AM
Also, do this (unrelated to VE):
- set D0960, D0961, D0962 to all 100%... see this: showthread.php?t=3336 (http://forum.efilive.com/showthread.php?t=3336)

joecar
March 27th, 2010, 11:54 AM
Is it just me or does increasing timing really affect AFR? I added 2 degrees at WOT and it went lean (12.8 to low 13s).Changing ignition timing can change the actual AFR... changing the timing changes the time that the mixture has to burn... (I don't understand it fully... or even a bit... :doh2:)... how much oxygen is remaining at the end of the burn (this is what the oxygen sensors see)...

timing-AFR effect depends on various things like combustion chamber design (fast-burn vs other)... some engine combinations show AFR being sensitive to timing, while other combinations show little effect.

You now need to do a quick re-touch on your VE table... but do make sure there is no spark knock... make sure there is no KR in the log, but also make sure you don't hear any.

joecar
March 27th, 2010, 12:00 PM
Yes, I would enable PE like mr.prick said.

joecar
March 27th, 2010, 12:11 PM
Yes, injector is just about at maxed out... 98%.

BEN is converging.

A little bit of transient AFR (transient lean coming on WOT, and transient rich coming off throttle).

:)

GMCtrk
March 27th, 2010, 01:18 PM
Well, I went back to closed loop and ended up getting ~1.5 KR around 50% throttle en route to WOT. Really upset me.

It seems at WOT my AFR is very unstable. Do I need to smooth my VE table?

WeathermanShawn
March 27th, 2010, 02:44 PM
Well since your back to closed-loop, I would assume that most of your airflow calculation is primarily MAF at that point?

But, I try to 'manually' smooth my VE Table. Why wouldn't the VE Table Airflow go up and as your RPMs and MAP change. Sometimes I will get some small peaks and valleys where I may be a different gear at the same RPM and MAP. Or I convert it to g/s (or your preferred unit) and see how it looks that way. But open-loop (street) can do that. Its very hard to filter, as are many methods of tuning.

Thats not a lot of KR. I prefer to keep a 2-4 degree spread between High and Low Octane Spark Tables, just in case. I would have to look at some of my WOT AFR logs to see if I fluctuate as much (WOT AFR). I mean a little misfire here and there especially with headers..occasionally I see a similar pattern. Perhaps yours fluctuates more than mine..and since you are running out of injector..might want to take a closer look at that.

Edit..I see you have MAF failed. My mistake. Might be more reason then to do some smoothing on that VE Table..

Same on Spark..You will be running off your Low-Octane Table anyway, so disregard that comment. I missed C2901, C2901, C2903, all indicating MAF Failed...:)

GMCtrk
March 27th, 2010, 02:51 PM
Yeah, I'm running SD. I figure it's easier to just run it SD than to bother tuning the MAF as well. I know the KR wasn't really a big deal, especially since it wasn't at WOT, but it was annoying considering I had just done several logs with no KR whatsoever.

I was gonna just go ahead and manually smooth the VE table, but I didn't want to end up screwing up my AFR I just got set up. I wanted to ask and see if most just leave it as it or smooth it.

WeathermanShawn
March 27th, 2010, 02:55 PM
I'll let the SD'ers comment. I manually smooth even when running CLSD. Those airflow jumps in the same RPM range or when your MAP levels out..might go better smoothed out a little.

We will see what the SD troops think..

mr.prick
March 27th, 2010, 03:38 PM
IMO
The spark advance is not that high in those areas,
try bumping PE to 1.176 EQ and see if that helps.

GMCtrk
March 27th, 2010, 04:15 PM
The motor does have 10.6:1 compression in a 4800lb truck, so I'm not expecting to be able to get the same amount of timing as a lighter car. I know that the low compression truck motors can take 28 degrees all day, but I was told 24 was gonna be pretty good for mine. Honestly, 24.5 feels the best on the butt meter, I also got the highest delivered engine torque with that timing (438). My stock cam is completely out of gas.

Now, we do have 10% ethanol in our gas here. I wonder what the ideal AFR is for that? I know pure gasoline 12.8 is pretty good, which is what I am targeting here. Maybe you're right and I should be shooting more for 12.5

joecar
March 27th, 2010, 06:11 PM
Try like mr.prick said on PE (1.176)... and set timing down half a degree (to 24°)....

When you smooth the VE, do only localized smoothing (small patches), remove the really spiky peaks/holes... don't do global smoothing... if there's valleys/troughs, leave those alone (they reflect some characteristic like wave harmonics, maybe).

SS Enforcer
March 27th, 2010, 08:17 PM
Running e10 maybe you should be shooting for 12.3 afr's stoich for e10 is about 14.2. my car runs lean in olsd when I fill with E10 so I add 5% more fuel everywhere to compensate.

cheers

joecar
March 28th, 2010, 09:18 AM
Yes, with E10 or E85 your stoich AFR changes... this has to be taken into account... search the forum for "E10 AFR".

GMCtrk
March 28th, 2010, 03:54 PM
I think the truck seems to like around 12.5:1 the best. Delievered engine torque seems to drop off higher than that (eventhough flywheel HP seems to go up with higher AFR - saw 411 peak). Speaking of that, how accurate have you all found the flywheel power and delievered engine torque PIDs to be? I seem to be settling around 404/436 and I think that's pretty much all I'm going to get out of the stock 196/207 cam.

I did a couple more logs and iterations of the VE table after the 1st log, but it seems like I preferred the first log best. My biggest issue is the variation of the AFR at WOT. I would like to get it as steady as possible. I'm guessing this is due to the peaks and valleys in the VE table? I've emailed my tuner all this info and I'll see what he has to say about it - Justin@BlackbearPerformance

WeathermanShawn
March 29th, 2010, 01:09 AM
I think it is. When I look at your log(s), your DYNAIR.DMA is bouncing all around at WOT. That seems a little unusual. Like Joecar said, taking out the spikes either by better initial filtering or manually helps a lot. Some hills and valleys are fine. But perhaps your higher RPM and MAPS should be more level.

Here are some examples of a CLSD Tune with 100% Throttle. My AFR does not bounce like yours. The VE Table was constructed entirely using the CALC.VE Table Tutorial.

joecar
March 29th, 2010, 03:09 AM
Also try setting your LC-1 to do 1/6th filtering.

Chevy366
March 29th, 2010, 04:58 AM
Funny 6.0L like 12.8 to 13.0 at WOT , 12.5 seems to lose power , different engines like different AFRs .
VE is best not smoothed , as Joe said bring some of the peaks (large spikes) and valleys (large dips) into sane numbers , use the fill cells with linear data .
Using LTFT and Narrow Band sensors (very inaccurate device Narrow Band O2 sensor) tuning the VE is not for all , VEs don't have to be smooth to be correct .
My VE is rock solid at WOT and it is not smooth .

WeathermanShawn
March 29th, 2010, 05:11 AM
Funny 6.0L like 12.8 to 13.0 at WOT , 12.5 seems to lose power , different engines like different AFRs .
VE is best not smoothed , as Joe said bring some of the peaks (large spikes) and valleys (large dips) into sane numbers , use the fill cells with linear data .
Using LTFT and Narrow Band sensors (very inaccurate device Narrow Band O2 sensor) tuning the VE is not for all , VEs don't have to be smooth to be correct .
My VE is rock solid at WOT and it is not smooth .

I disagree that narrowbands are inaccurate. I have found them very accurate at stoich. He needs them to run closed-loop (thats what he wants) Smoothed or not at WOT his DYNCYLAIR is jumping a lot. Thats hard to understand. A harmonic peak/valley makes sense. OP, if you feel like you need to at least smooth those spikes, just make the sure the Airflow values do not decrease when smoothing. Then try another test run.

Every car is different. Unless your car is simply running out of injector..something is causing those AFR to occasionally run 'lean' in your WOT. Those are the ones I might work one.

Good luck..

GMCtrk
March 29th, 2010, 02:11 PM
I think one of the problems is that I am not hitting a lot of cells when I'm logging. Especially in the higher MAP regions. Can someone tell me if there's any way to prevent my trans from shifting 1-2. I'm thinking it might be better to log (and safer) just in 1st gear. The problem is even with the lever in 1st gear, it automatically shifts out. The hold 1-2 table is already maxed out at 410kmh so that isn't what's holding it back. Maybe I should just set up the part throttle tables to never shift 1-2 in the tow/haul mode.

WeathermanShawn
March 29th, 2010, 02:31 PM
GMC:

I'll let the Auto Trans guy help you out on that point.

I think it is your 6000 RPM line that has the biggest drop (as you noted). If your VE Table at 90-95 MAP looked liked you VE Values at 100 MAP, then it would all make sense.

Hope you get it worked out. It is hard to find the right gear to do tuning safely on the street. But barring your ability to hit those 5600-6400 RPM, 90-105 MAP, I would just use the 100 MAP values (minus ~1-3%) for that dip you now have at 95 MAP. Thats just my opinion.

Good luck. Sounds like you are on the right track..

GMCtrk
March 29th, 2010, 02:49 PM
The problem is those 100 MAP values are not accurate. Values over 100% are not possible (okay, I realize it technically is possible, but not my engine). I think the issue is I am not hitting those cells enough with the autoVE and those cells aren't getting corrected. I think if I can get tuning in 1st gear i'll be better off. It will be hard because of all the wheel spin, but atleast I'll be going less than 50 mph.

joecar
March 29th, 2010, 08:37 PM
I think one of the problems is that I am not hitting a lot of cells when I'm logging. Especially in the higher MAP regions. Yes, this is difficult to get enough cell counts...

This is a special case where you lower the empty cell threshold count (say 10).

Logging a 1st or 2nd gear run isn't really representative, the mph is kept down, but the load is significantly less in 1st and 2nd and the engine just accelerates easier/quicker... the BEN is more accurate when the engine accelerates slower... (a loadable dynomometer can hold a constant engine speed).

See if you can hold 3rd gear, and do a couple of WOT runs like this, maybe provide extra load by riding the brakes for a few seconds... the thing is that you will see where the cells hit should be, the VE table follows the TQ curve (or vice-versa), so you can see the shape of what it should look like... when you get it right you will find your BEN stays at 1.0 during a WOT pass.


Can someone tell me if there's any way to prevent my trans from shifting 1-2. I'm thinking it might be better to log (and safer) just in 1st gear. The problem is even with the lever in 1st gear, it automatically shifts out. The hold 1-2 table is already maxed out at 410kmh so that isn't what's holding it back. Maybe I should just set up the part throttle tables to never shift 1-2 in the tow/haul mode.Try this:
Set D0960,1,2 to 100% (otherwise it conflicts with all other tables).
Set D0914 to all zeros (to hold manual 2nd).

Or better, manipulate the PT shift tables and WOT parameters to hold 3rd (i.e. allow it to quickly shift to 3rd and then hold it).

joecar
March 29th, 2010, 08:43 PM
If you drive it aggressively for 20 minutes on open highway, you should get most of your VE table... try to operate the throttle in a progressive manner, and try to hold it steady longer.

then go to the drag strip on open night and log a few WOT passes.

GMCtrk
April 1st, 2010, 01:05 PM
Joe,

might have a problem. Justin@Blackbear thinks my wideband isn't reading accurately since even after autoVE I have VE values over 100%. He says that EFI live misinterprets the analog voltage from the LC-1. e.g when the lc-1 voltage is 2.5 EFI live interprets it as 3v. Basically he said I need to rewrite the equation for the LC-1 AFR PID. He recommended getting a volt meter and measuring the voltage from the LC-1 with setting the AFR to 16:1 and 11:1 with the DVT.

dc_justin
April 1st, 2010, 02:36 PM
Joe,

might have a problem. Justin@Blackbear thinks my wideband isn't reading accurately since even after autoVE I have VE values over 100%. He says that EFI live misinterprets the analog voltage from the LC-1. e.g when the lc-1 voltage is 2.5 EFI live interprets it as 3v. Basically he said I need to rewrite the equation for the LC-1 AFR PID. He recommended getting a volt meter and measuring the voltage from the LC-1 with setting the AFR to 16:1 and 11:1 with the DVT.

Yes, this is a common issue with using analog output on most any wideband.

As for holding a gear, you can use DVTs to specify the gear you want and as long as you don't surpass the value in {C6152}, it will do what you tell it to do. I'd suggest doing more of your VE tuning in 2nd gear for higher RPM than first gear (on the street) as well, as you'll be able to spend more time in each cell.

GMCtrk
April 1st, 2010, 03:29 PM
I'm still confused why this inaccuracy exists. Is every LC-1 putting out different voltages from the factory that cause the EFI live PID to be off that much? How have you ended up modifying the PID in the past? In EFI live it is (AD1 x 3) + 7.35. I would assume all of the LC-1s are programmed to output 0 volts at 7.35 AFR and 5 volts at 22.35 AFR

joecar
April 1st, 2010, 07:10 PM
The default/factory programming in the LC-1 gives 0V=7.35AFR and 5V=22.35AFR, which is represented by equation afr = (AD1 x 3) + 7.35...

But it looks like you're seeing the problem of voltage offset...

Do this:
- connect FlashScan to vehicle,
- connect LC-1 to FlashScan V1,
- program LC-1 to output 3.5V, observe what EXT.AD1 says on scantool,
- program LC-1 to output 1.5V, observe what EXT.AD1 says on scantool,
- program LC-1 back to default.

Note:
- connect the serial cable to the LC-1 OUT connector, connect the DB9 end to your laptop;
- connect the terminator plug to the LC-1 IN connector;
- to program the LC-1 use the LM Programmer software from Innovate;
- after you reprogram the LC-1, you have to remove power for a moment, then reconnect power.

For example:
if scantool reads say 0.5V too high consistently across the range,
then the equation would become afr = EXT.AD1 * 3 + 7.35 - 0.5

This can be put in the calc_pids.txt file, we can show you how.

dc_justin
April 2nd, 2010, 02:18 AM
The default/factory programming in the LC-1 gives 0V=7.35AFR and 5V=22.35AFR, which is represented by equation afr = (AD1 x 3) + 7.35...

But it looks like you're seeing the problem of voltage offset...

Do this:
- connect FlashScan to vehicle,
- connect LC-1 to FlashScan V1,
- program LC-1 to output 3.5V, observe what EXT.AD1 says on scantool,
- program LC-1 to output 1.5V, observe what EXT.AD1 says on scantool,
- program LC-1 back to default.

Note:
- connect the serial cable to the LC-1 OUT connector, connect the DB9 end to your laptop;
- connect the terminator plug to the LC-1 IN connector;
- to program the LC-1 use the LM Programmer software from Innovate;
- after you reprogram the LC-1, you have to remove power for a moment, then reconnect power.

For example:
if scantool reads say 0.5V too high consistently across the range,
then the equation would become afr = EXT.AD1 * 3 + 7.35 - 0.5

This can be put in the calc_pids.txt file, we can show you how.


What he said. I forgot all about LM Programmer.

GMCtrk
April 2nd, 2010, 03:34 AM
The default/factory programming in the LC-1 gives 0V=7.35AFR and 5V=22.35AFR, which is represented by equation afr = (AD1 x 3) + 7.35...

But it looks like you're seeing the problem of voltage offset...

Do this:
- connect FlashScan to vehicle,
- connect LC-1 to FlashScan V1,
- program LC-1 to output 3.5V, observe what EXT.AD1 says on scantool,
- program LC-1 to output 1.5V, observe what EXT.AD1 says on scantool,
- program LC-1 back to default.

Note:
- connect the serial cable to the LC-1 OUT connector, connect the DB9 end to your laptop;
- connect the terminator plug to the LC-1 IN connector;
- to program the LC-1 use the LM Programmer software from Innovate;
- after you reprogram the LC-1, you have to remove power for a moment, then reconnect power.

For example:
if scantool reads say 0.5V too high consistently across the range,
then the equation would become afr = EXT.AD1 * 3 + 7.35 - 0.5

This can be put in the calc_pids.txt file, we can show you how.

That looks like a good idea. Only problem is my laptop doesn't have a serial port. I guess I could try those serial to USB cables though.

GMCtrk
April 6th, 2010, 12:22 PM
Bringing this one back up.

I put a cam in this past weekend and tried doing some autoVE'ing this evening. I got the thing idling pretty well by reducing fuel at idle RPMs, adding timing, and adding idle airflow (some other changes too).

Anyways, I went to go do the autoVE and I was idling rich rich (13s) and getting all kinds of AFR swings and real KR. Got my BEN averages and updated the VE table, fired it up and I was idling in the 16s! I would make a log and go back and everytime it would just switch from rich to lean and vice versa. It looks like to autoVE this thing, B4322, B4323, B4331, B4330, B4324, B4325, B4328, and B4329 all need to be set to 0. I think the computer was adding airflow when it was idling rich and then combined with the BEN changes made it just swing to really lean.

Any other tips for tuning a cam off the top of your head? How are you all setting up the VE table prior to doing autoVE? All cells below 1200 rpm I pulled fuel, but 1600 and up I should add how much? 15% across the board? 5% at low RPM and more up higher? Thanks.

WeathermanShawn
April 6th, 2010, 01:15 PM
I never had to fool with the Idle Learn Limits like that. I took out almost all the Low Rpm Airflow in the Learning Tables B4512- B4515 (per the Idle Tutorial). I guess I understand your logic, but I am not sure how the PCM will handle it when you limit the Idle learn Limits like that...

As far as fueling, generally a cam will require 10-25% less fuel in the lower 1200 Rpms. It all depends on your cams VE profile. I have less fuel from Idle to ~ 2800 Rpm/60 kPa, then 5-15 % above stock at that point. The tough part is from 1200-2800. Low kPa's, less, higher kPa's more.

Idle will always be tough as IAT and ECT swings can make day to day tuning of VE tough. I usually try to get an entire fuel map done in one day..keeping the ECT and IAT steady. If you decide to re-enable Trims after AUTOVE, I and others have more experience in that. If you stay open-loop, do you have any custom OS's available to you? That might help on Idle AFR swings.

I use a slightly different method, but I think if you can just log everything in one short tuning session, it is better than a morning vs evening run and then trying to combine the two results into one.

At least that has been my experience. Good luck..

joecar
April 6th, 2010, 01:19 PM
Idle:
What were all those tables initially (I looked up a stock 12212156 F-car file and all of those were something like +/-0.50, one of them was 3.00).

B0101 VE table initial bump:
15% may be too much unless you have also increased displacement significantly or added modest boost...

try 8% and do a trial run at light-medium load (pay attention to knock)...
- if you find the BEN's are all lowering B0101 by about 5 or more % then 7% was too much.
- if you find the BEN's are all raising the table by more than say 3% then 7% was not enough.
Yes, "skew" the initial bump up as you said and as Weatherman said (i.e. less bump up at low load, more bump up around peak torque).

The closer you can get your initial bump up to the actual VE, the easier it will be to get a good B0101 table.

What are the specs on your cam...?

:)

GMCtrk
April 8th, 2010, 11:26 AM
The cam is 222/226. Just added the 44lb injectors today too.

Just did a couple logs getting this thing dialed in. I simply cannot get the autoVE to work below 1600 rpms so I'm not using that data. Is seems that startup requires a lot more fuel than does a hot idle. Therefore, if I do a log and I'm idling at 13.8-14.0 (which I am), the autoVE of course pulls a good amount of fuel. Problem is, next time I go to start it with the updated VE table, the thing will barely start and stay alive because the AFR is in the 16s.

The other issue I'm having is a transient lean spike before going WOT causing knock. Otherwise it's going good and right now i have the AFR very rich and also very low timing (20 degrees across the board). I'm pretty comfortable that I now have a safe WOT AFR so I'll increase that timing and then increase the AFR into the mid 12s.

joecar
April 8th, 2010, 11:44 AM
Are those 44 lb/hr at 43.5 psi or at 58 psi...?

What is your rail pressure (with MAP-reference hose removed)...?

Would tables B3632, B3633, B3662 help with startup AFR...?

Transient lean spike can be due to injector tables and/or transient fuel/wetting tables.

joecar
April 8th, 2010, 12:08 PM
Try setting B3618 to a single value all across (sufficiently rich for WOT/load).

joecar
April 8th, 2010, 12:10 PM
Ok, I confirm MAF is failed (P0103) and CL/SOL is disabled (HO2Sx1 don't cycle).

joecar
April 8th, 2010, 12:13 PM
Did you verify that your wideband is producing a correct result...?

HO2Sx1 voltages are indicating wideband may reading a little lean, but we can't necessarily go by that.

Did you recent do free air cal on your wideband...?
When you first used this sensor did you do a heater cal...?
[ just checking ]

joecar
April 8th, 2010, 12:16 PM
Are you applying the filter to the BEN map...?
Are you using the average value in the BEN map...?

BEN seems to average 1.00 at load...
BEN seems to average 1.00 at cruise some of the time, other times it's a little under (see 2nd pic).

joecar
April 8th, 2010, 12:18 PM
Which bank is the wideband on...?

joecar
April 8th, 2010, 12:32 PM
You're not getting much knock... in a few cases it immediately followed a lean spike, but in other cases it looks incidental.

GMCtrk
April 8th, 2010, 01:28 PM
Let's see if I can answer all your questions here.

The wideband is on bank 1 (drivers side)

Also, the tune is running well at WOT. No knock whatsoever except for those lean spikes right before getting to WOT. What I don't understand is that I have PE set up to turn on at 50% throttle, yet it is not.

Yes, I am applying the filter and using the averages to modify the VE table.

The wideband was just recently installed, brand new, and followed the start up free air calibration and heater calibrations.

The wideband may indeed be running lean, but I would rather it read lean than the other way around.

Once I get these transient lean spikes figured out, I will probably just put it back in closed loop and call it a day. I suppose I could try and work on the WOT AFR and get it closed to 12.5-12.7 but according to Justin@Blackbear there's a negligible power difference between 12 AFR and 13 AFR at WOT...it seems better to err on the richer side for a street tune anyways.

WeathermanShawn
April 8th, 2010, 02:11 PM
If I understand B3643 (PE Fuel Step Size), 1.0 is actually the fastest fuel step (I think I read that on this forum.)..

Looks like you are going into PE mode. Your WOT AFR is your business. But no harm should come from running even a 12.2 -12.7. A lot of excess fuel acts just like spark retard. Gets a little snappier as you lean it somewhat. Even 12.5 AFR is a good compromise.

Good luck..

GMCtrk
April 8th, 2010, 02:29 PM
Thanks for that info Shawn. I went ahead and zero'd out B3644 (PE commanded fuel ramp in rate). I also set B3643 to .250.

I also fiddled with the Impact Factor (B3406) and increase those cells from 8.7 psi to 14.5 psi by 25%.

I'll try these settings out tomorrow and report back.

Also, you're right about the AFR. I'll try and shoot for 12.5:1

Thanks.

joecar
April 8th, 2010, 02:42 PM
...
What I don't understand is that I have PE set up to turn on at 50% throttle, yet it is not.
...Set B3609 to all zero.

mr.prick
April 8th, 2010, 03:05 PM
You can cure the cold start lean condition w/{B3605}
What I do is set {B3605} to be viewed in EQ in the properties section,
then use WBO2 Lambda as a multiplier instead of BEN.
If {B3605} is viewed in AFR or Lambda you can use WBO2 EQ as the multiplier.
This way stoich is always the target.

This will work until the engine gets up to temp so don't change the ECT areas you are tuning the VE table or MAF in, keep those areas @ 1.00 EQ.

GMCtrk
April 9th, 2010, 10:26 AM
Well, I made the changes, the problem isn't with PE (although it comes in nice and fast now). The KR spike and lean AFR spike is always before going WOT. For example if I hold it in 2nd gear and bring the throttle up to 25-50% that's when I get the lean spike. And I'm not exactly sure why that's happening. It's still commanding 14.68 there.

WeathermanShawn
April 9th, 2010, 11:24 AM
GMC:

While it is curious your are getting the KR right at throttle on..I guess I will be the first to recommend just lowering all your spark values 2-3 degrees in the mid DYNAIR flow values.

Not sure of your CR, but even though my WOT spark is a couple degrees higher than yours, almost all my mid-DYNAIR Spark values are about 2-3 degrees lower. Some engines need that rapid spark drop. I could throw all the fuel in the world at those cells, but only taking the spark down eliminated the pinging and KR.

In reality most engines are only in that spark zone for a few seconds. Eliminating that KR will then allow you to get to the high-load cells quicker.

Just a few thoughts..

GMCtrk
April 9th, 2010, 12:47 PM
Yeah, I think you're on to something there. That part of the spark table though is all bone stock. I'll definitely look into lowering spark there, because I noticed even sometimes with stoich AFR I was still getting KR. Spark was in the mid 30s.

Anyways, I'm pretty much done with this tune for now. Put it back into closed loop, LTFTs are hovering from -4 to +4 which I am content with. WOT AFR is around 12.50. Spark is around 25 degrees at peak torque at makes its way up to 29 degrees at 6400 rpms.

Post cam: Estimated flywheel HP - 474 Delivered engine torque - 456
Pre cam: Esimated flywheel HP - 390 Delivered engine torque 436

This thing hits pretty freaking hard from a 4-2 downshift on the highway roll in a 4800lb truck with a 4l80e

WeathermanShawn
April 9th, 2010, 01:02 PM
Nice...Congrats on your numbers..

I'll make sure to stay our of your way!!!:grin: