Actually, Open Loop will use the VE tables as they are with no further correction coming from the 02 sensor. So if the VE tables are off, you won't be close to 14.7. PID controls (Proportional Gain, Integrator and Decel) are only invoked in Closed Loop. The issue with large overlap cams is the fact that more of the intake charge at low rpms goes out the exhaust before being burned. In C/L, the O2 sensor sees this as lean and continues to add more fuel. That's why it runs rich at idle.
A way to address this is to lower the O2 sensor voltage window during idle and at the airflow where say 2200rpm takes place. Not sure about your def file or which editor you are using but look for 02 Terms for Rich/Mean/Lean. By changing those voltages and/or the o2 sensor bias', the o2 sensor will "think" its ok to run leaner at those points of throttle opening and airflow.

 

The rich/lean threshold vs airflow table (or similar, if available) should be suitable for tuning in the presence of excess air for certain flow conditions.

The PE vs rpm discussion was a different conversation regarding L98s with a 165 ecm as to whether or not PE vs rpm enrichment was still avalailable in open loop. It is. The mixture, however is not directly compensated by O2 feedback during PE in either closed loop or open loop, since the BLMS are prevented from being updated since they would be chasing a likely richer than stoichiometric target which wouldn't make sense with a narrow band O2.

Historic BLM values still offer mixture compensation even during PE, but they are not being updated in real time while in PE or open loop. This compensation can occur both in open and closed loop if non 128 blm values are stored.

 

you guys are confusing me. (that's not hard to do!)

first, on my ecm, datamaster never triggers PE mode when it's in closed loop, period. i can start it on the dyno and immediately go wot, and no PE mode. once it goes into closed loop, then PE becomes active when i meet the wot threshold for tps.

you're saying PE is active in open loop too? is datamaster misreporting it, and if so, should we contact them to change this? if it's active in open loop, it's applying PE enrichment on top of 12.9 in the afr table at wot.

second, i can't seem to find anything that changes O2 voltage limits. if that exists, you're saying it modifies how far from stoich the ecm will allow the voltage to read before attempting to correct? if so, that sounds like the perfect fix, if it can be applied vs whatever rpm (or airflow) we choose.

-michael

 

MSR,

Sorry for any confusion. The PE discussion applies to L98s, as they only have basic open loop vs coolant temp, open loop vs load and PE vs rpm and PE vs coolant temp tables and a stoich constant to determine target AFR. The PE enrichment is applied on top or either the open loop target AFR(variable) or the closed loop target AFR (14.73)

Your car very well may be different.

 

Just to be accurate, PE occurs in O/L. And is usually dependent upon the TPS/MAP/RPM thresholds in the cal. You likely have an O/L AFR table that is used for commanded AFR when in O/L. However, how accurately your actual O/L AFR tracks to the commanded AFR depends on how accurate your VE tables are to generating a 128 BLM. This is because the O/L AFR is calculated by the ECM since there's no 02 sensor feedback. If the BLM at the "point of departure" from C/L was > 128, the ECM will richen up the mixture depending on how far off the BLM was in C/L. If the BLM is < 128, the ECM does nothing.
This is one reason why you guys doing dyno runs may find that your WOT AFR varies between runs. Unless you have tuned for Part Throttle correctly, you won't have consistent WOT AFR.

 

Are you using TC or TP as your bin editor? However, the answer to your question is that the O2 Term Windows will "define" the O2 voltages that the ECM will use to consider where stoich is. These are usually 3D tables so that the O2 voltage can vary depending on airflow.
The 7747 ECM, for example, used a Constant O2 bias for Idle that modified the O2 Window Terms when the motor met the conditions for what was defined as "idle". Other ECMs may have a separate set of tables for Idle and Cruise. You'll need to find a deffile that has these table already in it or get the addresses so these items can be added to the def file.

 

i'm using TC. i'll look again, but didn't see anything like that in the definition file from TC.

do you happen to know whether or not TP has that in its def file?

thanks.

-michael

 

You should find these at www.moates.com or on the TP site.

 

thanks. the TP site doesn't offer the EE cal, but moates does. i took a look at it but the O2 voltages table isn't in there, either.

i think i'm about to have a crash course in modifying calibration files...

-michael

 

MSR,

You may want to post on 3rd Gen and ask if anyone either has a hac or an expanded version of the $EE def which includes the O2 Window Terms. If you get a hac, you can find the addresses and then add the items you need to the def. You also may want to ask John at TC if he would do it.

 

TC had it already!

i need to figure out what MAF flows correspond to idle and low rpm/light load. the stock setting looks like the O2 sensor voltage threshold is 381mV for anything above 8gm/sec airflow, and 460mV below that.

so, if i understood you correctly, i want something significantly lower, maybe 250-300mV?

is that threshold alone going to change much, or do i also have to play with methods to make it run leaner in the first place? (that is, find out how to make it run leaner, and the threshold changes will stop the ecm from attempting to richen it.)

thanks again!

-michael

 

Michael,

Check the def file in the Constant area to see if their is any bias for O2 Idle settings. The ZR-1 def has a constant that defines what airflow is considered "Idle" airflow. Probably not any higher than 16gms/sec. Drop the O2 voltages in that area in 50mv increments.

 

Hi, Mike

1. I use ambient air as ~ 1.3 gram/l . If you work the math, 8 g/sec MAF flow on a 5.7 l @ 600
rpm (call it a typical idle) = ~15% cylinder filling (from the perspective of the MAF). That would
be the right ballpark for a stock cam.

2. The problem with a radical cam is that (depending on your IVClosing compared to stock) the
stock MAF and VE tables are going to be way off from idle up to some large number. Combine that
with overlap, and I found that changing 02 setpoints really did nothing for driveability on a ~80*
OL cam that I was playing with a few years ago. The best thing I found for driveability was to crank
in a lot of extra timing for part-throttle conditions, to compensate for reduced cylinder filling.

3. I also corrected my VE tables to have an accurate base fuel map for that cam. I really never fooled
with changing MAF tables - they represent a derived VE anyway, but are complicated by the non-linear
transfer function of the instrument, compared to just using VE. Seems to me you need to adjust one
or the other to better reflect your actual fueling load - in addition to the revised spark. To bone up on
MAF functions, see the Delphi article here: http://delphi.com/pdf/techpapers/2000-01-0546.pdf

4. I work mostly in the VE realm - I just find it more intuitive. See my condensed version of VE fuel
calcs here: http://forums.corvetteforum.com/showthread.php?t=689344
And some work I helped with here - increasing VE table resolution:
http://www.thirdgen.org/techboard/di...ve3-patch.html

5. For my situation, I stopped short of running in open-loop-only to get the cam to behave. Instead
I had one made with wider LSA and steeper lobes (ala the LSx designs) to better fit my daily-driver
usage. Dave Vizard is great for racing engines, but I don't see where any of his experiments speak to
fuel injection for a more normal, street driven application. Of course the real answer is to just buy a
new Z0-6 and let GM figure all this stuff out... In the meantime you might want to talk with Alvin @
PCMforLess - he and his guy do a lot of tuning for the later C4s.

HTH,
DrJ

BTW, I took your advice a couple of years ago, and mounted Nittos all around for summer use - they
ride and handle very well, so Thanks!

 

Mike,


Your situation with cam overlap and tuning has been discussed on 3rd Gen pretty thoroughly. I think the concensus there is that 1. Of course the VE tables need to be "close", and 2. Nobody is suggesting that changing th O2 Window Terms can make a stock VE table "work", and
3. that modifying the O2 Window Terms does work particularly at idle with a higher overlap cam.
Worked very well for me on a cammed Xfire. My O2 Idle Window Terms were :

Rich 360mv
Mean 334
Lean 290

BLMs at idle were at 128 +/- 4

I would suggest you read this thread on O2 terms. It addresses what you are experiencing. There's several more if you do a search.

http://www.thirdgen.org/techboard/di...anual-car.html

 

first off, i'm amazed at how good the tech can be on this forum. thanks to you both, really.

Doctor J, the particular ecm i use ignores the VE tables completely if the MAF is hooked up. i have contemplated switching to SD and tuning just the VE tables, but have stuck with the MAF until now because it makes it a lot easier to get a new combination rolling quickly.

i'm glad you liked the DRs!

Dominic, before even reading this, i ended up with a mean of 331mV in my testing yesterday. it did make a significant difference along with all the other work. i think i need to extend the change to even higher flow rates, and maybe play with timing a bit more. the idle is amazingly good, i was able to drop it down to 900rpm, even. it still has a radical lope, of course, but it's smooth and consistent, and quieter somehow.

altogether, the drivability is not much worse than it was with my old 61-degree overlap cam, now. again, thanks all, i'll definitely read all the links and refine the tune, then report back when i'm "content" with the result.

-michael

 

Michael,

One other point I forgot to mention in conjunction with dropping the O2 sensor terms. Particularly on a cammed motor, maintaining flow velocity thru the throttle body is important. Take a look at the IAC counts at idle. Try to do the Min Air Adjust so that your IAC counts are < 10. I used to run at 0-8. This minimizes the bypass air and forces more air thru the butterfly maintaining air flow velocity. That's important for torque. Throttle response will be improved. You can also get into lmiting the Throttle Follower so as to minimize IAC air throughout the entire rpm range. You want to get more air thru TB at WOT assuming TBs are large enough to supply the needed flow for the respecttive motor.

 

i'm still refining the tune. one area i've been unable to address is tip-in behavior. at highway cruise speeds, it feels like there's a dead spot in acceleration as you tip in. the engine almost sounds like it's dropping cylinders.

it felt/sounded to me like a problem with the TPS, so i checked it. the voltage tracks very smoothly. i tried datalogging with the engine off to make sure the ecm was seeing the correct voltage. when doing this i noticed something that really puzzled me: the %TPS reported by the ECM remains at 0 until quite a bit of throttle is added.

i had always thought the ecm took a baseline and figured %TPS from there, but it doesn't appear to be working like that. i slotted the TPS and tried different baseline settings, resetting the ecm each time, and it consistently seems to use ~0.9V as the breakpoint from 0% TPS.

this contradicts everything i've read about the '94-'95 ecm, i was hoping someone else has some input. fwiw, i set the TPS to 0.88V (it was at 0.55V) and still need to drive it to see if there's any improvement for that. if what the datalogging reports is true, i can't see how that wouldn't make a big difference since at 0% tps the timing advance would be at least 10 degrees different from 1% tps!
-michael

 

i have actually done the opposite, but for a different reason. the rear geometry seems prone to changes from on/off throttle. on a busy auto-x course this can mean throttle transitions in areas that really exacerbate the handling difference.

i found the handling tamed quite a bit if i tuned the car to limit engine braking. turning off DFCO and adding a little bit of throttle follower made quite a difference in the transient handling.

-michael

 

Increasing the throttle follower and the lag filtters will cause thte IAC to react slower. I can understand why you would want to take some of the jolt out of throttle transitions.



MSR,

As for your tip-in issue, does your Mask ID have AE tables based on Delta MAP and Delta TPS%? If you do, these are the tables you'll want to adjust and increase based on the both the MAP and TPS changes where the stuttering occurs. If you are datalogging these events, is the NB O2 going lean when you increase throttle? As an analogy to the old "carb days" you may need more of a pump shot.

 

FWIW:

1. Re tip-in response: Your PCM probably has more in common with the later (LS1) designs than with
the P4 ECMs. Read the discussion here on VE vs MAF tuning for throttle transitions -
http://hptuners.com/forum/showthread.php?t=687
The frequency-out MAFs are way better than the older units, but in MAF-only mode you still have the
problem of gas compressibility (in the manifold) on a rapid change in throttle plate position. A MAP
fuel table reacts more quickly to such transients. Remember the MAF is reading ambient air flow,
while the gas after the butterfly is following the equation V=nRT/P.


I made a data logger some years ago to look at just such transients - logging RPM and WB volts vs time,
each at 120 samples/sec - to look for fuel bogs. (Spark transients I get more by feel than by data
stream.) For example: http://www.corvetteforum.net/c4/doct...taq24apr03.jpg
This shows tip-in enrichment & engine speed vs time in graphic form - you can infer power output
from the slope of the curve, if you are working with WOT transitions. I found that form useful for
isolating transient fuel problems...


2. Re the TPS reading - dunno why you are seeing that. Is there a different scale in your data recording
prog to look at TPS volts (or A/D counts) directly, instead of TPS %? Are you using an aftermarket
TB? In engineering terms, you're probably aware that gas-flow vs position is VERY non-linear for butterfly
valves, with very large outputs for very small changes at the low end, trailing off more or less to an
asymptote once you reach ~75-80% open on the blade. A large-bore TB only amplifies that transfer
function - having way more opening than you need for flow modulation.

BTW, on the LT1 tunes I've looked at, most wanted to use a lot of IAC opening. That put air into the
idle circuit holes in the manifold, correcting what looked like big air-flow maldistribution in the base
intake shape at lower RPMs. Quite a bit was written some years ago on machining aftermarket TBs
to fit the LT1 idle air circuit...


3. Re off-throttle transitions, I agree with using a lot of throttle follower on deceleration. On the
acceleration side, I set the closed TPS spark table to be close to the main spark table values, to ease
any bump in the power curve. On mine, the change from closed-TPS to main-table is governed by
both RPM and VSS - so I have the transition set for ~5 mph. Works for me, anyway...

HTH
DrJ