How to extend AE duration?

[Kitch]

I'm not having much luck figuring out how to extend the AE for my Vortec headed G20 van.
I'm getting knock counts with mild acceleration from a standstill, I can see some initial AE on acceleration then it goes lean on the Wideband followed by some knock counts.

What do you guy's do to extend the AE duration with an $0D mask?

[sturgillbd]

There are two AE triggers on $0D. TPS and MAP based. These tables are where I would start. Increase the time AE is active during the lean areas you are seeing in the log. The times are in microseconds. If adjusting the AE doesn't fix the issue, you may be lean because the VE table is off in that condition and AE is covering up the lean area temporarily while its active.

[Kitch]

Thanks for your reply.

Throttle position is normally around the 10% mark for mild acceleration and I'm still in closed loop at that.
Would having a lean VE table in that area still make a difference in closed loop? I can certainly see how it would in open loop.

I have a few logs where I can watch the van move away from a stand still. I can see the throttle open from 0 to 10% and I briefly get active AE that drops the AFR 10.8 and the RPM rises from 600 to 1250 but by the time the van actually start to roll out and put some load on the engine the AFR is in the low 15's.

For example:

0.0 open throttle and AE is active almost straight away
0.7 of a second after that AE is off and AFR is showing 10.8, throttle position is at 9.4% and at 58 KPa
1.15 seconds after initially opening the throttle the van starts to register vehicle speed, 45KPa
2.3 seconds after initially opening the throttle the van's up to 6MPH, AFR show's 15.5 and I've gained a couple of knock counts. 1300RPM, 53KPa & 26.5° of timing

So to me it looks like longer AE duration might help keep the AFR down enough to get the van moving.

I noticed your "AE vs MAP Difference" table has much higher values compared to what I currently have.
What vehicle was that table from?

Just reading over what I've written I'd still like more AE but I'm wondering if I should also look at reducing the timing around 1300RPM?
I am using 100+ octane unleaded fuel.

[1project2many]

Incorrect VE affects engine operation both in open and closed loop. The ECM requires time to deliver fuel, check O2, make corrections, then enter those corrections into short time and / or long term fuel trim. Here are some symptomes I use to help separate incorrect VE from incorrect AE

If acceleration sags or the intake pops almost immediately after the throttle is pressed then you probably need more AE.
If power is lacking or the intake is popping after throttle andle is increased then held you likely have incorrect VE.

AE gives a very fast shot to cover the change in air density as the throttle plates are opened or as engine load increases. VE provides correct fuel when throttle angle is relatively consistent or changing slowly.

Since VE can and should be tuned with WBO2, I recommend focusing on VE before AE. Otherwise you risk creating a calibration that has too much AE. A calibration like that performs erratically.

Good luck.

[Fast355]

TPS AE provides a short burst of fuel. MAP AE provides a longer shot. In carb speak. TPS is like the pump cam and squirter. MAP is like a larger accelerator pump.

[Kitch]

Thanks for the help guys.

I trimalyzed a couple of runs and the BLM's are generally on the slightly rich side now rather than the slightly lean.
Unfortunately, I'm still getting knock counts as described above.

I tried the "AE vs Map difference" values that sturgillbd posted but in a short test drive I can't say I noticed any real difference in AE duration.
Does anyone have a suggestion of something that should noticeably change my AE and I can then work back from that?

[tayto]

what plugs are you running? i went from a 44LTS to a 42LTS and knock counts went down from maxing out in a 10 minute drive to virtually nothing. If i recall after the plug change most of my knock counts were from startup. exhaust hitting the chassis? how are your engine&trans mounts?

[Kitch]

Thanks for the suggestions.
I'm currently using NGK TR55's, I changed from NGK BPR5EFS.
I've tried 44LTS and 42LTS but in my case I can't say I noticed much difference. From memory I think I felt it ran worse with the 42LTS so I changed them out.

The engine mounts are new AC-Delco replacements. It could be knock caused by contact but the acceleration is quite mild so I doubt there's much torquing over going on.

I always get knock counts cranking, they can range from 800-900 when the engine is cold to 250 when hot.

At one stage in frustration I had an 80° thermostat installed, 42LTS, 100+ octane fuel and I took 5° off the spark table everywhere and it still showed some knock counts.
There's been the odd time when I've been just cruising along at 60MPH steady state with no change in throttle position, KPa or spark timing and it would throw up random knock counts!
In the end I decided it must be false knock but I put an inspection camera in the plug hole when I was changing to the TR55's and noticed what looked like shinny metal on the top of my pistons and then thought it might knock that's caused that?

At the moment I'd just like to try and get some more AE into it to see if more fuel changes anything.

The most disappointing part of it is that this should be a really torquey combination but the van can't break traction to save it's life!
While I know my vans heavy it should be able to rotate a tyre.
There's been may times I've thought I'd be able to get this thing sorted if it had a carb on it.

The van's cool but it's a stark contrast to one car I had that would lite up in top gear if you put your foot down at 60MPH.
I remember trying to accelerate so that I could pull back onto a motorway and merge in with the traffic doing 60, I struggled get it to stop smoking the tyres so that I could pickup speed....it was a fun car but not very practical.

[donf]

I always get knock counts cranking, they can range from 800-900 when the engine is cold to 250 when hot.


A lot of new people confuse knock counts with knock retard. Knock counts show up from time to time and its normal. My vortec head g30 van always had knock counts when starting also I don't remember 900 but there was always some on start. The computer is smart enough that it pulls timing with knock retard once it gets bad enough to worry about. That is what you want to look at. If you are pulling timing trying to get rid of all knock counts you are going backward. I did have to increase AE a bit with the vortec head swap, but not as much as some said. I got horrible mileage when it was pushed too far. When done right it pulled close to 10,000 lbs in van, trailer and mobile chassis dyno over some very steep terrain and got fairly good mileage. If you don't have a wideband stuck in the exhaust and able to log the data STOP until you do. It looks like you are trying to shortcut the whole process and go back to a carb when it fails. You have to do things in the proper order or you are just wasting time and the results are not what they should be.

[Kitch]

Yep, I have a Wideband in my exhaust and I'm logging with it.
I have racked up more miles in this van doing logging runs than I have actually using it.
I've been trying to get this engine dialled in for years now and I'm not saying a carbs are better than EFI.

As far as my ability to get AE working properly, I could get a carbed setup's AE dialled in with shooters and cams within an hour.
I spent days trying to get AE to show some change and for very little gain.

I don't get many knock counts under load, say accelerating up a hill. I do get about 200 when doing a WOT acceleration run including some knock retard.
Yes, I have been pulling timing on the spark table where in the areas where I have signs of knock retard. Then after gradually making changes and not a lot of knock reduction I end up reverting back to the original spark table.

Yes, I'm probably doing things wrong but I'm learning as I go.
If I could tighten up the rich, lean swing when in closed loop and keep the AFR closer to 14.7 I might not need more AE?

[84Elky]

My thoughts:
-First, it would really be helpful if you could post an exported .csv log file. Might also post your BIN and XDF.
-You don't state which AE is being invoked and then disappears (MAP or TPS). You can get AE-MAP even while idling as MAP fluctuates. Will only see AE-TPS for a short period when sufficient TPS% change occurs.
-Will generally always register knocks at startup due to vibration
-All things being equal, true knocks only occur under engine load. Hard to load an engine while in 1st gear under mild acceleration, even in a heavy van. If true knocks are really occurring under those conditions, you'd have them everywhere. So likely vibration. Load producing knocks would generally occur at low MPH in 4th gear and accelerating with as much throttle as possible without down-shifting, or doing same going up hill.
-Why do you think AE is the problem? In a stock or mildly modified engine, it's rarely an issue.
-Are you getting any stalling or hesitation when going lean while accelerating. If so, injectors may be an issue.

[Kitch]

Hi,

I've uploaded my last BIN file and XDF, for some reason it won't let me upload a .csv file?
I've also added my ADX file that has my Wideband setup in it and the last log I made in an XDL, sorry it's from a cold start.
I have an 8625 PCM with a Moates APU1 attached and my Wideband inputs into the APU1, that's why I supplied my ADX.

Is there a way to specifically monitor what's calling for AE, ie MAP or TPS in TunerPro?
In the log I've attached I've reduced the "MAP difference to enable AE" value so I now get a more active AE. It was more as an experiment to see if it would make a difference rather than a long term solution.

Pretty relaxed about the knock counts at startup, my stock TBI 350 did it and when I changed to a TBI L34 it did the same. Just assumed it was noise from the starter or something like that. The increased knock counts when cold is probably just related to the fact it cranks over a little longer when cold.

I had got to the point where I assumed it was false knock because I could get random knock counts at light load and when there was no real reason for it.
The photos showing shiny parts on the piston tops made me start to reconsider.

Even assuming my knock is false, I'd like to find a way to stop the Wideband AFR getting to 15-16 in closed loop as the van is accelerating away from a standstill.

The engine in the van is a new L34 crate engine with ported Vortec heads, 9.5-1 compression, Comp 260XFI, Comp 1.5 Roller rockers, GMPP Vortec to TBI intake, coated long tube headers, ported stock TBI @ 20PSI, MSD6al2 and MSD leads.

The van can't break traction, if I stood on it from a standstill it kind of thinks about it and then goes. I don't have any recent logs to show what the AFR was doing at that point, I'd have to have a hunt around.

The injectors were stock replacements bought new about three years ago and I've done maybe 5000 miles with them, they could be an issue.

[84Elky]

FYI - MS data files are considered to be unsecure and contain viruses. Have to zip .csv .xls* and .doc*. I can export your xdl given you included the adx file. Don't need csv.

But what about:
-Why do you think AE is the problem? In a stock or mildly modified engine, it's rarely an issue.
-Are you getting any stalling or hesitation when going lean while accelerating. If so, injectors may be an issue.

[dud]

TPS Difference AE Coefficient vs. Temperature will stretch out the AE pulse time. I had a lean spot that occurred before the engine got heat soaked that I couldn't get rid of until I worked with this. I think it's because of the volume difference between the new vs. old intake manifolds. For me this also made the engine feel more responsive and smoother.

The smaller the number, the longer the pulse time.

Beware that you may have to adjust AE vs. TPS Difference accordingly.

Here's where I got the idea:
http://www.dynamicefi.com/Tune_Intro2.php
"The other method of creating delta values is by the use of a filter. By using a programmable filter value (coefficient) the tuner can change how much delta there is. A filter lags the change in a value based on both time and the value of the filter. The acceleration enrichment (AE) parameters have a table of filter values to create a dTPS and a dMAP value. By using a smaller filter value the delta's are increased. For AE this increases both the volume of AE fuel and the duration of AE fuel"

[84Elky]

Admittedly, I'm not overly familiar with $0d/TBI. I primarily live in the $8d/TPI world, but they are similar. My thoughts below looking at your log and BIN (see also attached .xlsx):
-Everything looks good and can see no real issues other than that below.
-Even in a heavy van, going from 61 MPH at 2900 RPM to 90 MPH at 3900 RPM is like accelerating with an anchor out (Samples 5738+). Also, during that time, injector PW is < 5.5ms which seems quite low, but again, I don't know what to expect in the TBI world. Further, the IAC is only 73 steps. It's higher elsewhere not in WOT. Generally, you would expect to see IAC open further as RPM increases. Is it possible the throttle body is not 100% open with the accelerator on the floor?

Minor items:
-Knocks: Not an issue. Spark Table is in line with Vortec heads (30-32 deg max). You're seeing minimal knocks and no retard degrees. Unless there are calibration settings preventing retard from being computed, you're OK. The knock numbers look large, but it's not uncommon for the sensor to report knocks of 3 digits when they occur. Yours appear to be from vibration.
-PE: (Samples 4120+) Is active even at constant MAP and low TPS%. Not normal
-AE: The AE reported is AE-MAP (0x003D b6). AE-TPS is not reported but it could be (0x0005 b6). This is why you see AE Active with minor changes in MAP -- as it should be (samples 5689+, 5753+, etc.).
-CL Idle appears to be on rich side per WBo2. BLM is reported at 128, but that's because you're losing heat at the NBo2 while idling. 'BLM Enable Flag' is "Disabled" so going into OL. Need a heated NBo2.
-Lean Mild Accel: Don't see what you are seeing with lean WB with mild accel. At sample 2174, it's rich then mildly lean. Then at 3609+, it's rich and then mildly lean. As initially stated, AE is rarely an issue. This appears to be a lean VE table around 1100-1500 RPM, 40-50 kpa. Remember, AE-MAP is initially large and then is decayed out quickly. Again, unless getting hesitation/stumble, richen up the VE and forget it.

Biggest concern is the anchor. Others are a nit.

[Kitch]

Hi guys,
This is just wanted to say thanks for your replies.
I'll need to take some time to go through what you've posted, I'm sure once I've got my head around things I'll have some questions.
Thanks again for the help.

[Kitch]

Hi Dud,
Just trying to understand your statement "The smaller the number, the longer the pulse time".
If I'm understanding you correctly then reducing the uSec values as a trend should potential increase the AE duration?
I had previously assumed that the larger uSec value meant longer pulse time at the injectors.
I made up and AE vs TPS difference graph to for you to comment on as a visual example of what I think you meant, please let me know if that looks correct?

[dud]

Yes. Lower values = longer AE duration.

[Kitch]

Thanks for confirming that.

Looks like I've been going the wrong way, increasing AE vs TPS values when I should have been reducing them to increase the AE's duration time.

It potentially makes sense now, about a year ago I went back to a near stock bin which would have had stock AE vs TPS values and I remember thinking the van felt like it accelerated a little bit better.
Unfortunately the next day I increased the fuel pressure and tuned for that, at the time I probably thought I'd better increase the AE vs TPS values because stock values must be too low.
I've put my van away in storage ready for the winter but once I've gone through 84Elky's last post I might need to get the van out of storage and try some things.

Thanks again.

[84Elky]

TPS Difference AE Coefficient vs. Temperature will stretch out the AE pulse time. I had a lean spot that occurred before the engine got heat soaked that I couldn't get rid of until I worked with this. I think it's because of the volume difference between the new vs. old intake manifolds. For me this also made the engine feel more responsive and smoother.

The smaller the number, the longer the pulse time.

Beware that you may have to adjust AE vs. TPS Difference accordingly.

Thought I'd weigh in here. Apologies for the long post.

First, the table Dud referenced (0x4B5A=TPS Difference AE Coefficient vs. Temperature), applies to AE-TPS only.
Let's look at exactly how it's used:
-A coefficient (% change factor), is extracted from the referenced table based on coolant temp
-That coefficient is later used to lag filter the Current TPS%. Lag Filtering is a technique used to smooth-out values to avoid rapid changes.
Formula: NewFilteredTPS% = PriorFilteredTPS% + [(CurrentTPS% – PriorFilteredTPS%) * LagFilterCoeff]
The smaller the coefficient, the closer the new filtered value (NewFilteredTPS%) is to the old value (PriorFilteredTPS%). Plugging values into the formula will prove this.
-After the NewFilteredTPS% is computed, it's subtracted from CurrentTPS% to obtain a Delta TPS%, and to determine if TPS% is rising
-If Delta TPS% is not rising, AE-TPS is NOT invoked; or if currently invoked, it's exited
-If Delta TPS% is rising --AND-- NOT currently in AE-TPS --AND-- the Delta TPS% is > (0x48B9=TPS Difference to Enable AE [Factory = 1.17% .vs. the posted BIN here = 0.78%], AE-TPS will be invoked.
-Once invoked, an AE-TPS PW is extracted from the table (0x4B45=AE vs. TPS Difference) based on the Delta TPS%

So why this detail? On-balance, the Delta TPS% will generally (note "generally") be larger with lower lag filter coefficients because TPS% - [lower]NewFliteredTPS% = [larger]Delta TPS%. This means the AE-TPS PW extracted from the 0x4B45=AE vs. TPS Difference Table will be larger (larger Delta TPS% = larger AE-TPS PW). But remember "generally". Delta TPS% is calculated 160 times a second. Due to varying engine conditions (TPS position), it's impossible to know at any point in time exactly what a given Delta TPS% will be.

IMO, it would be far better to change the PW provided at any given Delta TPS% by changing values in the 0x4B45=AE vs. TPS Difference Table. Now one knows exactly the AE-TPS PW that will be provided at any given Delta TPS%. Don't have to wonder what the Delta TPS% will be. Can't control it anyway. Other than for the coefficient, it's calculated with values we can't control.

Granted changing the table coefficients "may" work, but here's something that is supported by a multitude of tests over many years: Mildly changing lag filter coefficients (and there are many) rarely results in consistent output. Exact results are indeterminate. And radically changing coefficients can have unexpected results.

Bottom line, it's best not mess with them. The GM Engineers had significant data to support what they used. As above, there is generally a more concrete way to achieve a desired result than changing filter coefficients.

FWIW, Elky

[Kitch]

Hi 84Elky, sorry about the time it's taken to reply. I've had a vintage computer problem to solve and I didn't want to focus on this until I had that sorted out.
Thanks for taking the time to have a look at the csv. log, I appreciate it.
I don't normally look at the csv. file because of it's sheer size but I do now realise there is a lot of useful information in there.

Getting back to your original comments on my csv. log:

-Yes, the van is heavy and acceleration in high gears is underwhelming, but I guess that’s what I should expect given the RPM and the weight it has to move. I probably should have gone with a 383 for a bit more low RPM torque or a slightly lower diff ratio. I have played with the shift tables in the past and it was better.
Not sure why the IAC is sitting at 73, even at W.O.T. I just assumed that was normal?
I'm not sure what parameters control the IAC but I'm guessing it's not looking at throttle position once above a certain speed or in a certain gear?

-Knocks: thanks for the comments, it’s be a real monkey on my back trying reduce/eliminate knock counts. I’ve spent so much time over the years trying to sort it out and you’re right it’s probably just a rattle somewhere.

-PE: I reduced the “TPS % for PE” to get PE to come on sooner. So while you mention it not normal, it’s what I wanted to try.
I when accelerating up an onramp, manifold vacuum would drop to around 8" and at that point I wanted PE AFR’s not Stoich. Changing that value allowed me to at least be able to get PE to come on when I squeezed the throttle a little bit more.
If it was a heavy vehicle with a carb that's about the point I'd like the power valve to start operating.
I know early PE is not the best for economy but the van was staying at 14.5-15.0 even when the manifold vacuum was getting down to 5”. It normally cruises in the 10”-16” range at 60.

-CL idle is actually set to open loop. I’ve found the engine idles a bit smoother at 13.5-1 rather than Stoich so I’ve set it to go open loop when at idle. I run a heated Narrow band O2 sensor as well as a heated Wideband. I could see how it might have looked like the Narrow band O2 sensor wasn’t up to temp though.

-Yep, you’re right I do need to work more on my VE in the 1100-1500 range.

[84Elky]

Understand your comments. Explains several things. I wouldn't be concerned about the IAC. Only adj is to properly set it at idle. Stupid as it seems, I would check to see if throttle body is 100% open when pedal is on the floor.