i'm interested in hearing strategies any of you have applied to help drivability with high overlap cams. i have an LT4 stroker running an OBD-I ecm, and 70 degrees of overlap.

i rescaled the timing vs map tables, knowing the manifold vacuum for a given engine load would be a lot lower, and this helped tremendously. the drivability isn't too bad, it can pull 68-70mph in 6th (3.45 gears). what concerns me is how rich it runs below about 2200rpm, due to the overlap dumping raw fuel and oxygen into the exhaust and fooling the O2 sensors. it pulls about 20mpg at 70mph, then as the engine clears out and the rpm rises, it will read 25mpg at almost 80mph. (above that the gain in engine efficiency is offset by drag.)

i'd like a way to get the fueling right at low rpm in lieu of what the O2 sensors think. ideas?

thanks.

-michael

 

After a lot of idle tuning (idle speed, timing, MAF tables, O2 thresholds, etc) that still resulted in a very rich idle, I finally decided to run completely in open loop. That was the only thing that really did the trick. "Reported" idle A/F is in the 16.5 - 17.0 to 1 range, but the plugs indicate that it's not too far off from a fairly clean idle.

Drivability still isn't ideal under 2000 rpms, but it is definitely better than what it was in closed loop.

 

i have contemplated this. in open loop the PE fuel vs rpm is ignored and just the base open loof afr table is used. how do you tune the wot fueling if it varies per rpm? do you have to change the MAF curve? did you have any issues with that?

do you still have cats? if so, any issues?

thanks.

-michael

 

As I understand it, the problem with overlap and a common plenum or single plane intake is that on the intake stroke, you're drawing exhaust gases through the still open exhaust valve from the 90 degree adjacent (previous) cylinder that contaminates and mixes with the fresh intake air coming through the throttle.

The result is less oxygen for a given manifold pressure which results in excess fuel, unless the fuel is somehow reduced to compensate for this effect.

The trick is to figure out how much inert residual exhaust gas is present, so that the fueling can be reduced accordingly.

I don't think that the O2 sensors are being fooled due to overlap, unless you're so rich beyond the combustion limits that your getting complete misfires. I hope that is not the case.

 

that's not my understanding of what's happening. however, your bottom line comes to the same thing, that fueling needs to be reduced.

the way i see it is that there is indeed some mixing of exhaust, as you state. but, there's also bleeding in the opposite direction. unburnt air/fuel mixture is going out the still open exhaust valve. since an O2 sensor (incorrectly) corresponds the percentage of oxygen with a/f ratio, the sensor sees the abundance of oxygen as meaning the mixture lacks fuel. the ecm then compensates by adding fuel.

my experience with using a F.A.S.T. to tune racecars corresponds with what byebyeL98 said, that idle a/f mixture as reported by the wideband O2 sensor shows very lean when the trapped mixture in the combustion chamber is correct. i took that idea and assumed the MAF on my vette would produce a fairly correct trapped mixture, and that the closed loop feedback is what causes my car to run too rich. note, the wideband does indeed show the car idles and cruises at ~14.7:1, but all other evidence shows the trapped/burned mixture is too rich.

but, if i'm wrong and you're correct, what should the O2 sensor read in your scenario?

-michael

 

Hi, Mike

Couple of points -

1. If you look at a WB output from a fast data recorder (e.g. the one on my web page), at low
RPM there is in fact a spike of O2 that I attribute to cam overlap. The lambda control loop
sees that and adds fuel in closed loop. You can try to work around it with the O2 rich/lean
set points and proportional gains at low intake flows - but I've never been able to get rid of
the problem entirely through tuning. I can pass smog, but the O2 spike is still visible on the
recorder, just smaller...

Some folks with very radical cams have gone to running 100% open loop - see the discussion
here: http://www.thirdgen.org/techboard/di...ecm-large.html

It also depends somewhat on whether you need to pass smog check or not. When you say
70* overlap, is that the 0.006 number?

2. Dunno much detail on the '95 PCM, but on the older MAP designs, PE is invoked ONLY in
open loop. Dumping in ~25% excess fuel while trying to run closed loop wouldn't make much
sense. I'd be surprised if your PCM was set up differently, you might want to check that.

HTH

DrJ

 

The behavior is also going to depend upon the throttle. Since you were talking about low rpm, driveablity, I was thinking of part throttle operation.

If the engine is being throttled (high vacuum-low pressure), the pressure in the exhaust should generally be greater than the manifold pressure, tending to flow from the exhaust into the intake, through the previous cylinders valves when they are in overlap.

If the throttle is open (low vacuum-high pressure), exhaust scavenging can result in periods of lower pressure in the exhaust than the intake which can pull some fresh mixture into the exhaust during the intake stroke. This of course is a more dominant effect at mid-high rpm.

In realitiy, it is likely that both effects are occuring in various degrees depending upon the conditions.

Is closed loop actually adding more enrichment under these conditions, or is the open loop tune also too rich?

Does leaning out the open loop mixture result in more closed loop enrichment to compensate which would be visible in the blms?

It does sound like the wideband is reporting leaner values than actual based upon the other indications. Is this also reflected in the narrow band or blms?

I still wouldn't expect too much excess air in the exhaust under part throttle.

Retarding the ignition timing a bit should help clean up the exhaust if there are unburned combustibles, but may hurt low rpm performance, so it may not be good trade off. This may help consume some of the excess air before it is seen by the O2 sensors if that is the case.

 

Sorry to get off subject, but how do you force the car to stay in open loop at idle, and which table would control the open loop idle AFR?

I just found this adjustment constant "Closed Loop Idle Enable Vehicle Speed". Would this be it?

 

Here's an interesting article describing the effects of overlap. It doesn't provide any tuning answers but it is still interesting nonetheless.

http://compcams.com/Community/Articl...ID=-2026144213

 

good discussion, thanks!

DrJ:

1. i've thought about trying to tighten the learn limits to clamp down on how much it corrects, but i'm concerned the ecm is smart enough to figure it out. i guess you're basically saying the same thing, there's just limited success in that approach.

the overlap is at .006", correct. the lobes are pretty radical, there's 22 degrees overlap at .050". i don't think there's any way this engine would pass a sniffer test.

2. the terminology i used is confusing, but i'm essentially correct. the '95 ecm ignores any wot enrichment when in open loop. it drives purely off the afr table that is MAP vs ECT. once it goes into closed loop, it stays in closed loop even when at WOT. it ignores O2 sensor feedback at WOT, of course, but it's still technically in closed loop mode, as confirmed by datamaster.

once in closed loop mode, it uses a base of 14.7:1 for everything, then adds PE tables when at WOT. anything in the base afr table is ignored. there are 2 PE tables, one is vs ECT, the other is vs RPM. this allows one to tune a perfect straight line at WOT.

i hope i explained that sufficiently well. it's mostly confusing because open-loop and closed loop in this ecm don't follow traditional definitions, and because each mode uses mutually exclusive tables, even at WOT.

my concern is that open loop only has no way to change vs rpm, only vs MAP and ECT. if the afr varies at WOT with rpm, how do i tune that? i'm very averse to changing the MAF curve.

tequilaboy:

i don't really know how it's behaving in open loop. my wideband didn't warm up fast enough to see that. in closed loop it is adding a little fuel to correct at idle and low rpm, and about spot on cruising at 2500+rpm. i think i need to force open loop and do some tuning/datalogging to see if i can get a better handle on this.

96lt4c4:

if you mean stay open loop only at idle, it's impossible. the ecm will allow you to force open loop all the time, but you can't pick and choose via rpm or load.

this is actually a very funky attribute of how GM implemented the design. you could force the ecm to think idle is WOT, so that it uses the PE tables and ignores O2 sensor feedback, but the ecm still treats that as closed loop. i suspect this means it's doing some other things and will probably figure out you have tricked it, eventually. i never tried this myself, though, it might work. the one thing i don't quite understand about how it works is whether or not any learning happens or is applied to PE mode.

if you try it, let me know how it works. find the table where you can choose percent TPS for WOT, and put it to 0 for 800-1200rpm. then go to the 2 PE tables and 0 zero out the same rpm range so that the ecm isn't trying to add fuel.

-michael

 

i've never noticed that! i have no idea if it can be used to our advantage. what's it set to, stock?

-michael

 

ya, vizard wrote that. following those principles consistently means +20hp over what almost everyone buys/recommends. i went from a custom grind cam motion cam (with 112lsa) to the current (same duration but slightly more radical lobe ramp) cam on a 108lsa, and went from 403rw to 441rw. it makes more from 2500rpm upward, so really no loss anywhere.

actually, i think it's easily over 450rw, now. my laptop fritzed and i was unable to change the base tune i had. it was 11.5:1 afr when it made 441rw, and i let off before the curve flattened off. i've since tuned the afr but haven't made it back to the dyno. also, it's now maxing the injector duty cycle by 6200rpm.

-michael

 

"Closed Loop Idle Enable Vehicle Speed"

I suspect this is the vehicle speed threshold to enable closed loop PID idle speed control.

On the L98s this value is typically 1.9 mph, which means the car won't enter closed loop PID control ablove 1.9 mph. It has nothing to do with closed loop fuel control.

On the L98s its easy to stay in open loop control at idle. Just use a regular unheated O2 sensor with your long tubes. It will drop out of closed loop at idle every time.

 

My stock setting is .94 MPH

 

i still don't quite understand that. what's PID? does that setting have to do with when the ecm opens the IAC a bit to hang the idle when moving?

-michael

 

How about increasing the min load limit to permit BLM update?
Do you have a parameter like this available in your ecms?

If you set this to a load value somewhere above idle, you will essentially lock in the blms for the lower loads.

PID is just short for proportional/integral/differential control. It is the method used to control the idle rpm to the target rpm.

There are other idle control modes active for cruise and to provide dash-pot function other than jsut strict rpm control. The IAC needs to transition between the different control modes, conditionally.

 

i can't seem to find anything like that. maybe i'm missing it, anyone else seen this in the $EE ecm?

thanks.

-michael

 

Sorry for the delayed response :o . I haven't spent much time tuning WOT just yet - still trying to get idle and drivability down .

I didn't realize that the "PE vs RPM" table would be inactive for open loop. Hmmmm..... - do you know if this is true for an '87 165 ECM car?

The MAF is limited to 254 gms / sec at the higher volt ranges, so that wouldn't work. I could tune using "open loop AFR vs Load", but like you said - not ideal for varying afr within an rpm range.

 

To byebyeL98:

The PE vs rpm table and PE vs coolant temperature tables are definately active in both open and closed loop to provide PE enrichement.

The only difference is that the starting point (target AFR in open loop vs closed loop) may be slightly different depending upon the coolant temperature.

The Closed loop AFR PE correction will be based off of a constant 14.73:1 AFR.

In open loop, this target AFR prior to PE correction will be varied according to coolant temperature, but should be close to 14.73:1 for a warm engine depending of course on tuning.

 

Great to hear, Patrick! Thanks so much for confirming!!