Mark_ZZ3

HI,

I am attempting to tune a rather radical cam with a AUJP base. I figured I'd see if I could make it work. I may have to leap to S-AUJP for more granularity in the VE tables.

Car: 1991 Trans Am R7U/1LE - race car it's whole life, now coming back to street life.
Engine: 383, Accel Superam, 24lb injectors, heated o2, 160 temp, Dart Pro 215 heads, Comp Cams 12-902-9, headers, 3 inch exhaust, Richmond 5 speed, 3.23 rear gears.

The cam is huge - 260 duration at 0.050 lift.
Oval Track Mechanical Roller Camshaft
Chevrolet 262-400ci 1955-1998
Lift: .630"/.630"
Duration: 296°/300°
Lobe Separation Angle: 106°
RPM Range: 3600-6600

Thus far, I the VE tables working well. Idles OK 950rpm. The engine has lower vac and sits in the 65-95 range of the VE tables (except for decal of course). VE tuning wasn't so bad, but I can see the challenge with lower vac, you simply have less cells to work with.

So far, it actually runs well, and goes like bat crap above 3000 RPM.

Here is my challenge - below 2500 rpm, and part throttle. Seems to fall flat. The datalogs suggest it goes lean after the initial tip in. I know the cam is not designed for this range, so this is the challenge to make it as good as I can.

- I've adjust the AE Delta TPS Scale factor (0x535) and set it as 0.28 vs. 0.12. This really made a difference and got over the chug chug chug when you tried to get going. But what I see in the data suggests the engine goes lean after a second or two. So the AE pump works to get me going, but not long enough to get the engine past 2500 rpm.
- I've tried adjusting many of the AE tables, but no significant changes. While it might seem smoother, it still falls flat. You leave the lights and rev it up ... let the clutch out and it pulls for a second and then kinda chugs. If you rev it up higher and the let the clutch out, it will overcome the bog.

Here is another test. While cruising along, leave it in a higher gear, RPM in the 1700-2200 range, and give it throttle. The AE will give it a shot (so to speak), and then it kinda goes numb.

I've done some high way tests. While cruising at 60mph, I slow down to 40mph, and then lightly accelerate with 20% throttle (RPM in the 2000 range). Initially it pulls for a second, and then it kinda runs out of gas.

So here is question (also talking to myself as I write this).

- AE is simply to get you over the transition from initial acceleration into the running of the VE tables. Some call it a pump shot in carb terms. It only lasts for a second or so? If true, then I'm beating my head against the wall trying to use this to overcome the issue.
- PE mode is more of a longer transition mode or it can be setup as power under throttle. Perhaps I need to use this to engage extra fuel to get me over the chug chug zone? I could set the %TPS down to 15% or so and see if that comes online after the AE transition.

Some general info:

- BLMs move in their range. Nothing stuck. They sit well under 128 as an average ... 112-116. But depending on where I am, I've seen them down to 100. (Which is odd as in the same time period, the data log says the O2 reads lean under light accel).
- Cell - I am not stuck on cell 4. It moves to different cells. I wonder if the range is wrong for my setup. So when I transition from coast to part throttle, I need to be using a different cell (IE I go from rich, to a lean condition, but stuck on the same cell).
- Timing is setup basic for now. As I am in the top range of the table, not much to work with. Timing is more like old school weights. Rarely do I see any knock - but it has shown up (so it does work!).

I don't have my wideband O2 set up yet. I was figuring I'd need that for WOT tuning. But this part throttle has been a challenge.

Sorry for the longer note. This has been one of the more challenging tunes I've done. I'm somewhat pleased at how well it does run!

Any tips on where to look in the data for gems?
And are there any switches to tell you - for sure - that you are in AE or PE modes? (I can watch the desired AFR change when I enter PE mode).

... in wring this note, it helps with the thought analysis. I may dial down my PE mode to come on early to see if that solve the issue.

Cheers!
Mark.

Orr89RocZ

iTrader: (20)

thats what i am thinking, ae is suppose to account for the rapid influx of air from throttle transition. Ve takes it in steady state once ae decays out. Wideband will help here

how lean is it going? Thats what you need to see and perhaps adjust the ve cells more to command a richer value where its having trouble

thats a pretty big cam for sure and being a manual trans car it likely wont be happy under 2500 rpm. Fueling and timing changes will need to be experimented with to find the best drivability


could also try open loop to verify the fueling table with the wideband and also see if that helps at all.

Mark_ZZ3

Yes it's a crazy cam ... so I thought I may as well try and tune it. I knew it was a roll up the sleeves sorta thing.

The VE tables work at a cruise state, no real load. BLMs on average are in the 110-115 range. Under load is where they don't work (or so it seems). Adding AE solved the big chugga-chugga it had to get going. I thought I was on to something. But then I can't seem to get over the longer period when the engine gets up to 2500 rpm.

I am going to just try the PE enable and set it down to 12% or so and see how it reacts. Maybe it needs to enter PE mode sooner at lower RPM vs higher RPM (kinda backwards to a stock engine).

Trial an error.

Mark.

Orr89RocZ

iTrader: (20)

Well your 2000-2500 cruise area should be 50-70 kpa correct? Some throttle should be 80-90’s kpa right? Thats more load and im curious what the ve tables are doing for fuel ratio there on wideband, ignore closed loop values just to verify things.

having pe mode come in that soon seems like the wrong approach to me.

do you have an AE duration factor vs coolant you can try to adjust? Also what pulsewidths are being commanded?

Mark_ZZ3

Cruise only ever goes down to the high 50's and 60's.

I have "AE Delta TPS Pulse Width Factor Vs. Temp. " or 0x554 and "AE Delta MAP Pulse Width Decay Factor Vs. Temp. " at 0x591 which reference temp. These tables are pretty close to a stock type setup.

In many cases ... from 1600-2500, throttle in the 16-25% range, the O2 goes rich for the first two seconds, and then drops flat lean for the next 5 seconds. BLM is at 122, and INT climbs up to 160. Cell jumps from 4 to to 14, then to 10 and 11 during 10 second zone I'm watching. PW starts out at 2.29 at 1600 , and then climbs proportionally to RPM to 4.21 at 1900 RPM (still in cell 4), and then it drops to 3.80's when the cell changes to 14 and BLM goes to 118. Kpa during this part were 65-85.

Keep in mind this is one view but I think represents what I am seeing. And I might be chasing a ghost by watching the O2 (vs. a wideband).

I haven't found which dial best fits this problem.

Mark.

Street Lethal

iTrader: (16)

Was in the exact same boat with my cam, fun isn't it...

Mark, can you post up a picture of the lower VE graph...?

- Rob

Orr89RocZ

iTrader: (20)

i would try playing with the pulsewidth decay factor vs temp, if you are certain the ve table cells are good in the areas it hits under mild throttle acceleration

try reducing it

Mark_ZZ3

Keep in mind my vac keeps the engine in the 60-95 range. so 1/2 of the table I don't use. The lower blue is where I am sitting.



Here is the history table from last run ... This is to demonstrate where the engine lives. The down side to the MAP range from 20-100 with a big cam ... I'd rather see the range 50-100 so I can better separate out the cells.

Mark_ZZ3

Hmm ... I get it ... let the AE hang around longer is what this should do. It is worth a shot for sure ... as this would be a bum-in-seat measured change. Really ... I want to keep AE around a bit longer to get the rev's up.

Mark.

Street Lethal

iTrader: (16)

Mark, still try to flatten out the unused area if possible, and smooth it out. I also agree with what you said earlier, in that you feel you may be chasing a ghost. When I resurrected my GTA, I was tuning with the heated narrow band and bigger cam, and of course seen the very same scenario of 160 INT, yet was running so rich my eyes were tearing. Kept pulling fuel ignoring the INT, and it settled lower near 128. I do use a wideband, but I did not have it hooked up to the ECU yet, was just going back and forth with it comparing what was being observed and making adjustments. I finally said screw it, and hooked up the wideband to the ECU to do a quick self learn just to see where it would take it (at this point I was okay with the VE), yet the wideband still made its' own adjustments when in Gear, so I said screw it and started smoothing it out in conjunction with the wideband using it as a lead (see pic below, the 50'ish kPa area from 1200-RPM and up). I live in the same kPa area as you, I just run an automatic so it is slightly different. This is before I made the change, and it made one hell of a difference off idle. Mind you, I only compensated for injector size difference with the AE tables, did jnot touch anything else yet. It's been a very long time since I even looked at S-AUJP, which injectors is this bin put together for, and if they are a different size in comparison to your 24#'s, did you compensate for that difference in the AE...?

- Rob

Orr89RocZ

iTrader: (20)

you seem to have less ve commanded in the higher kpa values than the 50-60 range at the same rpm. Slopes down not up. I think that is part of your problem. I agree smooth that table out cuz on high rpm decel in gear it will get into those 30-40 kpa cells and be rich. That Ve table def can be improved. Whats the 1800+ rpm look like?

Mark_ZZ3

VE upper .... keep in mind I only worked on those areas I've seen the engine using. I have not done much on decel and such. More cruise, off idle, part throttle and a few 3/4 throttle. Still got some areas to tune. Stuck on the chug chug issue. I connected the 1600 RPM values from the low table to the high and then ramped up from there based on BLM readings over various cruises. I do have PE mode working so ti comes on. And Stoich was set for 14.0 in my config vs. stock 14.7 (so that would impact fueling everywhere ... unsure how much thou. Seems wrong since narrow band o2's are set for 14.7).

Mark_ZZ3

Seeing your graph ... makes me think I've just stuck too close to idle VE numbers and should maybe have a more aggressive slope to higher RPM and also to less VAC. I was using the BLM numbers from my drives ... but maybe these aren't truely representative due to the short drives and types of driving. Basically my test circuit is a few miles of mixed driving. So I could be chasing a BLM number that does not represent what I really need.


I suppose I could use the idle as my anchor ... engine runs great in the 900-100 RPM in the 70KPa cel.

This is like cutting down a forest with a swiss army knife.

Mark_ZZ3

Here is the VE lower table that came with the car when I got it. Done back in 2002 era ... oddly looks like yours. The car idled like crap. I had to take the VE table from the 78 down to 35 to get it smooth and BLM at 128. I suspect with this much fuel ... there wasn't any dead spot off idle.

So ... I am going to have to mirror this type of slope but keep the idle cells close to what I need.

And then back at the AE/PE stuff again.

Orr89RocZ

iTrader: (20)

Yeah i think you could use more slope to it.

with a big cam like that i do not like closed loop at idle and low rpm. I would set it to open loop til about 1500 rpm. Tune idle by wideband and plug readings and idle kpa....idle vacuum and timing adjustments to try to get as high a idle vac as possible is where its ideal and happy. Overlap will fool o2’s

also try to get a bit more data in steady state in those 1500-2500 rpm ranges in the 50-80 kpa range. Limit ae exposure to see how the ve table it self is doing

Mark_ZZ3

Ok ... I re did the slope (I kept the low number at 900-1100 from 70-80 as the car idles well there). Added in more fuel above my idle. Also changed the upper table to be higher to match the new curve. The test drive was very different. The chugga chug is much much less. So this is the right track.

I think I am being misled by the BLM in CELL 4. That works for idle, but not much else. So I have been tuning based on the BLMs when I am in cell 4 ... so gotta stop doing that. And I need to get a wideband installed.

So for now ... I'm going to add back in some fuel in the 1600-2500 ranges. See how it responds. If it gets smooth, then I can work backwards.

Orr89RocZ

iTrader: (20)

Yup keep going with it, it should come around

ULTM8Z

Ok, so one thing you need to do (and this may send your AE tuning effort back to square one)...

Set the flag for "Reset INT when AE active"

What you may be fighting against the ECM trying to keep you at stoich AFR when the AE is active. You program in some additional AE and the ECM sees that as a rich condition and then starts trying to trim fuel back out. You end up chasing your tail...

When you set that flag, the ECM suspends fuel trimming for the period of time that AE is active, which allows you to dial in the momentary AFR you want during an AE event.

Also, don't assume that more AE is always better. You can easily bog the engine with too much fuel as with too little. But you'll need your WB hooked up to see what's going on.

84Elky

Mark_ZZ3---
If you wouldn't mind posting a TPro .csv log that shows where the problems arise and maybe the Bin as well, I'd be glad to take a look.
Elky

Mark_ZZ3

For sure I am getting stuck in Cell 4 BLM outside of idle. So the Reset INT may help that when AE comes on. I'll give it a try.

Mark_ZZ3

I redid by lower VE table yesterday with some new insight. The result was a "better" off idle acceleration. Only a slight bog which was overcome if you rev'd up more when you let the clutch out. Seems this could be tuned out.

I will post up new log and bin once I do a check with new VE tables.

I am also thinking of putting my wideband in and running open loop to datalog it. See where things really are. I'm questioning the BLM numbers now. I got it setup to idle really well, but slope to off idle suffers. When I redid the VE tables, the off idle was good, but coming to a stop was bad.

It's coming along.

Mark.

ULTM8Z

EDIT: Sorry... it's not a flag, it's a constant (or scalar in Tunerpro).

In Tunercat => Force BLM Idle cell CCP DC% threshold

In Tunerpro => CCP, Don't force BLM Idle cell if CCP DC >=

Factory AUJP is 23.4%. Basically if your CCP duty cycle gets below 23.4%, it'll force your BLM cell to cell 4.

Mark_ZZ3

My setup was 19.92% for this parameter. Looks like I should lower this or set t to zero. My cam does not idle anywhere close to a stock cam. Ideally Cell 4 should just be used for idle and not above idle, correct?

Mark.

ULTM8Z

Yeah, I zeroed mine out as well. Idle cell 4 should just be for idle.

84Elky

The Calibration value at 0x522 (In S_AUJP XDF: CCP, Don't Force Idle in BLM Ceil #4 if CCP DC% >=) should be =0 to allow idle to float to any cell based on RPM and kPa. If other than 0, idle can be locked into Cell 4.

XDF Description: When this value =0 (S_AUJP Default), idle is NOT forced to BLM cell 4. This is important with a big cam that idles at kPa and RPM outside the range of BLM Cell 4.
S_AUJP Default = 0%
Factory Default = 23.44%

Mark_ZZ3

Quick update. I installed a WideBand O2, connected that to my APU1 and set open loop only to see what the tune was doing. To my surprize ... I had everything RICH! I had set the BLM lower limit to 80 to help with initial tuning, but that turned out to be part of the issue. It was down to 80 in some cell value. While this made it drivable, it actually caused other issues.

So for now ... I'm going with open loop tuning and then see what the graph looks like. See if I can keep everything in the 12-14 range ... right now I am seeing most values in the 10:1 range when it runs crappy.

But kudos to GM ... the BLM and cells actually cover off lots of fueling issues and makes it usable.

When I get my tune working, I'll post up a new graph.

Mark.

Orr89RocZ

iTrader: (20)

Good to hear. I always tune open loop and with a wideband first when possible. If you have a proper open loop tune, the closed loop wont have to work so much

Mark_ZZ3

Update.

Wideband tuning has been good. I've had to remove lots of fuel from the AE tables where "my" engine operates. With the low vacuum, it changes the look of the table compared to any stock one. I just don't have the range of 20-100 kpa, but instead live in the 50-100kpa range.

I got rid of the chugg chugg off idle. It pulls GREAT. Feels like a normal car. Full throttle is better as well. A couple things still haunt me ...

- 1800 Chugg chugg under "light throttle" or as you coast down from higher rpms (slowly). The engine bucks and jumps 100RPM (kinda like a surge I suppose). The commanded AFR is 14.2 at this time and wideband says 13.4. My guess is that I am too rich for the light throttle coast. I'm going to try and pull out a bit of fuel in the 50-60 KPA range of the 1600-2000 RPM. (I wish this had more granularity).
- Stumble when you come to a stop. I had thought this was fuel, but my spark seems to be jumping. I'm at 1000RPM and the spark will change from 28 to 22.15 ... almost digital. I am not moving. I am stopped. Engine idle. No TPS input. Spark table has all the same values around that zone ... 26. So I am puzzled at why the variance and where it is subtracting spark.

The beast is somewhat fun to drive now. 5th gear (which is 1:1 for me) at highway speed ... pulls! Just starting to come into the 3000 RPM power of the cam.

If you have ideas on the spark jump, please let me know. I'll be digging into it more.

Mark.

ULTM8Z

Thats probably the idle speed correction spark advance/retard. You can try zeroing out those tables.

Orr89RocZ

iTrader: (20)

Some of that chug at light throttle could be spark timing. Play with the advance and see if it comes around. It may want leaner too tho

Mark_ZZ3

I thought Timing as well ... turns out in my setup that was 34 degrees!!! I started with a stock timing table but have not dialed in all the zones. I dropped this area down to 30 degrees ... was a little better, but not gone. AFR is still 13:5:1 which is high on cruise and light load. More trial and error.

Mark.

Mark_ZZ3

Did some logging. Sharing info.

My BLM Cell reference numbers are as follows:



SO I am going to conclude my Cell boundaries are as follows based on my RPM and MAP values:



I am going to try to adjust the boundaries a bit to make use of the cell references. 0-850RPM is not good as my low idle is 950. So this where the engine will bounce at idle as it will jump in out of of two cells. I will have to see how much I can change here to improve. I Found the history table interesting to watch as it laid out the Cell "blocks" quite well. Cells 0 and 4 are pretty much useless as the engine would rarely go there ... perhaps why they use Cell for for CPP and idle force.

Anyways ... sharing info.

Mark.

84Elky

Perhaps better thought of as ranges:

Mark_ZZ3

Yes that is how I now understand it.

Is there any way to see what is "stored" in each cell?

Went out for a tune ... in open loop it ran pretty good. Still a bit of a chugg to sort out ... but on the right path. For the ride home, I enabled closed loop and figured it would even better ... BUT ... it ran worse! It was pulling too much fuel out. It seemed what ever was saved in Cell 9 and 10 was from an old setting (I was rich before) and it just went lean 16:1. So ... maybe I needed to pull the battery and reset the cell numbers as whatever was stored in them did not match what I needed.

Seems I need another test drive after disconnecting battery from ECM.

This is kinda like chopping a tree down ... one half at a time. You never really cut if all down, but get half way there each time you tune it.

Mark.

84Elky

PM sent

Mark_ZZ3

Update. I spent lots of miles with my wideband o2 data logging in open loop. It starting to run much more civil for the size of cam it has. Sounds mean as heck, but drives somewhat reasonable. I could actually let someone else drive it without instructions!

Here is my VE tables now ... keep in mind my vacuum is low, so I only use a small portion of the table ... I tried to see where it goes on decal and hills, but there isn't much.

It's pretty basic. My idle is 1000 rpm and Kps is in the 70 range.



Here is the upper table so far. I still need to do WOT data logging ... which isn't easy on the local highways (I'm doing 60mph at the top of 2nd gear ... this is a road course car not a drag car).



Again, pretty basic. The slope in smooth. The Wideband keeps me reasonable close to the target AFR ... but I have been trying to get rid of the rich condition at higher vacuum ... usually 1800-2000 rpm when you are coasting ... still needs work.

When I got the car ... here was the Lower VE table ... If you look at it ... the numbers are 99%!!! I could not figure out why it was so high ... and then the VE upper was more typical. What I found out was they programmed the PROM for 38 llb injectors BUT the engine had 24's with 60psi fuel pressure. So that would explain why they had 99% ... the 24s were not providing enough fuel as a 39 would. The graph looks better in the sense all the areas form the slope. But the engine never goes into those zones.





My point in sharing is to help the next guy who is look up info. The big cam is working on the stock ECM and seems to run well enough. More testing. I may book a chassis dyno for some tuning. I expect the 24's to run out of fuel at 6000 RPM with the 383 (well that's what all the math says ... you need 36's or more for this level of HP).

Mark.

Orr89RocZ

iTrader: (20)

Looking good. Nicer looking graph for sure. High vacuum coast areas can be tough to see. Easier to do with full manual valve body transmissions or manuals lol. Else you can try free rev, bring it up to 3000 rpm and punch rev it so it goes over 5000 and watch the highest vacuum it hits. Likely wont see that on the street but can get close sometimes

once you get some cells figured you can extrapolate the rest

Mark_ZZ3

Had a couple track sessions to test out my setup. Best was a 12.9 at 109mph .. .with a 2.00 60 foot. I had another run with a 1.84 60 foot, but missed my shift. So .. maybe mid to high 12's.

AFR numbers were ok through the run. I need to add a bit of fuel in the 4800-5500 range. And it goes down to 11.0 in the 5500-6300.

Now a question ... my BPW at 4900 is 9.57 and at 6300 it is 9.86. AFR at 6300 is 10.5 on the wide band. So I can lean it out some.

Where is there a table that lists the max BPW for a given injector? I am running 24's on a 383 (thats what the combo came with when I got it - and it was raced like this ... so I assumed it worked well).

Mark.

ULTM8Z

I don't know that there's a pulse width limit for regular fueling, whether it's in PE or closed loop. I think the only pulse width limitation in 8D is on the AE fueling.

Besides, if I were trying to lean out the PE fueling, I wouldn't put a hard limit on pulse width. I'd attack the PE vs RPM table (if your VE table is already to your satisfaction).

84Elky

There is no such table and all 8 injectors are fired at the same time so nothing for an individual injector. You need to compute injector Duty Cycle %. If you were using v6 S_AUJP and the provided ADX, it would be easy. Just take the reported ("Final PW" * RPM) / 600 = DC%. At 6300 RPM, the 100% duty cycle is 60000/6300 = 9.52ms. But to avoid overheating injectors and allow for open/close time, need DC% to be no more than 85% (maybe 90-95% if drag racing).
At 6300 RPM:
---85% DC = (60000/6300)*.85 = 8.09ms
---90% DC = (60000/6300)*.90 = 8.57ms
---95% DC = (60000/6300)*.95 = 9.05ms (95% risky)
Anything above those values and injectors are static (open all the time)

BUT ---- if using any other $8d AUJP ADX, what you are seeing reported is BPW BEFORE voltage compensation. So to compute a correct DC%, need to take the reported PW ms and add the voltage offset at your ign switch or pump voltage (whichever used) in MICROseconds from the Table at 0x3FA, convert it to ms and add it to the ADX reported PW. Then and only then can you get a true Final PW and true DC%.

HTH, Elky

Mark_ZZ3

I'm using the base AUJP one, not the modified one people use. I never thought about the duty time at 6300 ... but I can see these injectors are at max. Granted ... apart from Drag racing ... they have not been there. Maybe back when it was a full time road course track car, it was. BUT ... back then, they had fuel pressure set at 60PSI which would bring this down. I didn't like running the injectors at 60PSI as the ones I tested didn't all flow consistently with higher pressure.

A Superam is such a pain to take apart ... changing injectors is a long process. And really ... a MiniRam might be a better intake for this CAM anyways. A winter project to think about.

For sure this has been a learning curve ... and it helps explain how alot of it works.

Mark.

Orr89RocZ

iTrader: (20)

If you are 10.5 afr at 6300 you arent maxed on injectors. You need to take like 20% out of the pe vs rpm table or reduce the ve table. The pulsewidth will come down and you can check duty cycle then

then you can try more pressure if you need more margin