ZZ28ZZ

Running a 87 TPI and $32b code. See sig for other goodies.
I disconnected the o2 sensor to force open loop and tuned the "BPW fuel VS load" table using a DIY Wb-02.
Abt 60 EEPROM burns later, I have the idle and cruise settings dialed in at 14.7 to 1.

Reconnected the the factory O2 sensor and the cruise BLMs were 129. However, at idle, the mixture goes way too lean (like 20 to 1), and the eng starts running real rough. Idle BLM is pegged at 108.


I'm using GMECMedit, and can't find a **** to turn for this.
Any ideas?

Morley

BLM of 108 is Waaaaayyyyy rich, maxed, in fact, odd that the WB is reporting lean....tried a new "stock" O2 sensor? Maybe a heated one?

ZZ28ZZ

The stock sensor is brand new. It's also mounted in the normal location since I'm still running stock exhaust manifolds. Shouldn't need a heated one I wouldn't think.

Apparently the ecm thinks I'm too rich ( at idle) and is trying to compensate.

Grumpy

Unplugging the sensor may set defaults.

Set the Closed Loop Enable temp to like 255 to make an open loop chip.

Also disable the 13, 44, and 45 error flags.

ZZ28ZZ

I'm using GMECmedit and haven't been able find a "min temp for closed loop" enable flag.
Also, can't find the flags to disable the prev mentioned codes.

I do have hot, warm, and cold O2 closed loop timers. They default to 30, 160, and 300 secs respectively.

Could these all be set to zero to prevent closed loop ops?

Grumpy

Maybe time to update to the 6E code.
I'm sure I've seen a commented hac on it on the net.
Might look at what WWW.Tunercat.com has for sale, and see if his stuff fills the bill.

ZZ28ZZ

I'd prefer to stay with a MAF system running in closed loop if possible.
Especially since I can't really afford to change it at the moment.

Tried setting the hot,warm and cold o2 loop timers to zero, no help. Still went into closed loop.
Disconnecting the o2 sensor is the only way I can find to prevent closed loop ops and still run halfway decent.

Tried raising the lower BLM limit from 108 up to 128, then 130, then 132. Didn't have any effect on the mixture according to the wb-02. (Still ran lean at idle in closed loop). I thought for sure that would prevent it from trimming out the fuel at idle. Guess I don't understand how the BLMs work after all.

Tried raising the idle mixture PW up 30% in an effort to make the mixture so rich the ECM would max itself out trying to lean it and hopefully wind up at 14.7 to 1. That didn't work either. It still ran at ~20 to 1.

Played with the ign timing (in the s/w), still no help.

Any ideas what I do with this thing?

Morley

The $6E code Grumpy mentioned is MAF based, it is the 1989 ECM code, the last and best GM had to offer for MAF.

ZZ28ZZ

Oops, I thought $6E was for SD.


What's involved in switching from $32b to $6e?

I didn't see a BPW fuel Vs load table in the $6e.
That may be a problem.

ZZ28ZZ

I now have the ARAP ($6e) bin loaded.
Set the "min temp for closed loop" to 255*, and it stays in open loop. Idle mixture is perfect.
Switch back to closed loop and the same exact thing happens (goes waaay lean on the WB-O2 abt 1 minuite after start-up on warm eng).

Found a couple of new items in the $6e ecu that looked promising.


Adjusted the values for "closed loop rich/lean threshold VS airflow". Used various values from 100 to 900mv. No change seen.

Tried another item called "min LV8 to update BLM".
Found it at zero. Adjusted it to it's max setting of 12. No help.

What am I missing??

Morley

I'm wondering if maybe your WB O2 might not be the problem??

Ragtop89

I going thru the same problems.

Lean at idle.
Rich under load.
I replaced my O2 today and it helped a LITTLE.

As soon as I come out of Open Loop she pegs 160 and then comes down to about 120 when I give it gas.

I need to add fuel at idle somehow.

ZZ28ZZ

Thx for the replys.

Morley>> I considered that, but since my eng goes from sewing machine smooth to running really rough at the exact instant the WB shows a shift from 14.7-to-1 up to >20-to-1, it makes me feel the WB-o2 is probably accurate.

I'm now considering moving the factory O2 sensor a little further downstream. Closer to the WB-O2 sensor. May need a heated unit if I do.
Maybe it has something to do with the extensive induction mods and still running L98 exhaust manifolds.?.?..
Excessive heat, back press,??????
Somehow, the ECM thinks I running richer than I really am.

Anyone ever hear of a factory-type O2 sensor reporting false rich readings??

Ragtop89>> Our issues are not really the same.
I'm having 108 BLMs at idle. You stated your idle BLMs were 160.
We're at opposite ends of the spectrum.
My first guess on yours would be a vac leak, but that's only a guess since I don't know anything abt your set-up. Are you burning chips yet?

funstick

yep i had this problem with factory honda exhuast manifold on my accord 02 sensor was 4 inches from the exhuast port. apperantly over heating it will cuase it to read rich.

RBob

iTrader: (1)

Fun brings up a good point, ZZ28ZZ, have you looked at the O2 readings (actual value sensor returns)? Does it change at all as the BLM drops? Is the O2 reporting high voltage readings?

A silicon poisoned sensor will also report rich. Pull the sensor out and check it for a white possibly fluffy coating.

Not enough timing at idle will cause high exhaust temperatures.

RBob.

Grumpy

The oem series of O2 sensors are back pressure and EGT sensitive.

Not to mention the Bosch replacements have been flakey in HiPo engines since like forever.

Are you setting the idle AFR with the MAF Tables and scalers?.

With the stock manifolds, and cam at idle you have ALOT of EGR, so you might try adding some timing at idle.
At idle you can get into odd areas of timing/fuel confusing an O2. Again related to lots of EGR.

And just to confuse the MAF issue a lil more remember the MAF is blind to reversion, so you might have to make some serious Scaler changes to get it right.

ZZ28ZZ

I recently rtv'ed the runners and throttle body in an attempt to prove to myself I don't have any vacuum leaks.
RTV (Permatex, gray stuff) said it was "sensor safe".
I removed the WB-O2 sensor since I was still a little paranoid abt it, but left the factory sensor installed.
Ran car abt 100 miles and re-installed the WB-O2 sensor.

This problem has been with me since I first fired-up the eng (didn't start with the RTV party).


After looking at the winaldl log file, the factory O2 sensor appears to be working normally.
The O2 voltage goes high initially, blms start dropping, O2 voltage starts cycling high/low/high/low.

Wish I could post the Winaldl data here. Maybe someone might see something I'm missing.
I can send you an email with the Winaldl data (.txt and/or .xls)capturing a warm-up if you're interested in viewing it.

ZZ28ZZ

Grumpy>
I put the MAF tables (scalers never touched) back to original abt 60 EEPROMs ago.

All the tuning, after the WB was up and going, was done by adjusting the BPW Fuel VS Load table.
That was with the $32b code.

I've only logged one warm-up with the newly acquired ARAP $6e code.
If forced into open loop (by raising the closed loop enable to 255*), the idle AFR just happens to be right on (14.7-to-1).
If closed loop is entered, it goes to ~20-to-1 just like it did with the $32b code.

Hadn't even considered reversion.
Reversion makes the stock sensor indicate rich??

funstick

reversion can make cylinders read rich by having the air blown out of them or sucked out of adjacent runners. i cant imagine its being particulalry problematic with a TPI intake with the runner length. id think that there something else going on. can you email me some scans ?

Grumpy

Your confusing air movement, with wave action.

Grumpy

Try going thru things using the 6E code and adding to the MAF tables in the lean areas.

Yes, reversion will cause a MAF to read excessively high air flow numbers. A MAF sensor reads airflow in either direction, that's why some have a cone shaped piece in them, to min the effect of reversion across the sensor. At low airflows, small reporting errors can be problematic. Even with a seemingly small cam with just the right small errors you can have headaches.

If an engine wasn't so prone to reversion, and the inertia of air, the MAF tables would be just about linear. Heck, all the fueling for steady state, would be a piece of cake to tune.

funstick

got some news for you. actually having worked for a company that made MAF's i can tell you for sure. reversion in the intake manifold will nto affect the maf readings except to lower them by drawing less air.if he had reversion that bad wed call that a backfire. i think this might have more to do with one rich cylinder. but hey WTF do i know

ZZ28ZZ

I'm a little confused by the MAF issue being introduced.

I can see how if reversion was triggering the MAF to over-estimate airflow, the mixture would be rich, but I don't see how it could make the factory o2 sensor indicate improperly.

My understanding was that exhaust in the intake (i.e. EGR) would only dilute the mixture and not change the air-to-fuel ratio.

I was thinking that if some of the fresh mixture was making it past the exhaust valve, as it was closing; it might make the factory o2 sensor read rich. Once past the factory O2 sensor, it would burn up and not have the same effect on the WB sensor which is down near the converter.

Any thoughts on that theory?

Morley

Try swapping locations on the sensors and see what happens.

dnult

And you see these as two different things? A pressure wave is moving air, is it not?

-dnult

ZZ28ZZ

Morley>> I'd like to swap the sensors, but there's a few problems involved with that.

I'm not sure if a heated sensor (WB unit) will survive in such a hot location.
The WB wiring would have to be extended.

I suspect the WB sensor location wouldn't be hot enough for the stock (1 wire type) sensor to work.

I'd like to install a factory heated type sensor down by the converter next to where I mounted the WB sensor.
I'll first have to locate a heated sensor that will work with my ECM. Hopefully, a search of this site will list an alternative sensor I can use.

32V_DOHC

And you see these as two different things?

Yes When the intake valve closes the air still has inertia. This causes the air to stack up as a high pressure zone. The pressure can not stay higher than its surroundings so the air starts to flow toward the lower pressure. When the air pressure equals the moving air still has inertia. This causes the air to rush past the equal pressure point and become lower than the air outside the pipe. Since it can not stay lower the air from the higher pressure zone pushes back in. This is wave action.

Reversion is where the pressure in the cylinder is higher than the pressure in the manifold. This causes air to push into the intake. This MAY happen twice per cycle. At the overlap period and just before the intake valve closes.

reversion in the intake manifold will nto affect the maf readings except to lower them by drawing less air.

The issue is not whether the engine will draw more or less air than some other hypothetical engine with less reversion. The issue is whether on not reversion will cause the MAF to indicate that the engine is consuming more air than it actually is consuming. The MAF cannot possibly indicate less than the used air unless you have a leak after the maf.(Disclaimer: For the purpose of this discussion I am assuming that flow is uniform across the section and that there are no slow spots around the wire or air tumble that would allow the MAF to read low.)

If you create a higher pressure zone in the manifold through reversion it will try to equalize by pushing air backwards through the MAF. The MAF will count the air as it goes in, again as it is pushed back and a third time as it goes back in. It does not happen with all of the air but if 2% of your charge is reversed then your MAF will indicate %104 of the actual air.

got some news for you. actually having worked for a company that made MAF's i can tell you for sure.

got some news for you. If it is wrong someone will tell you regardless of who you have worked for.

HTH

John

funstick

first off thats called a back fire. again your making the assumption. if this statement were based in facutality the you would understand that every holley carberator would deliver 104% of the actual fuel needed becuase you think you idea holds water it doenst. for the last time the air stream is not going to reverse out the intake manifold on a running engine. the TB blades are in the way, the incomming air is in the way, the stack of air in the runner is in the way. now its completely possiable that cylinders are richeing leaning becuase they are swapping air. but with a TPI intake and its long runners i find this to be very unlikely unless hes got a fanatastically large cam.


yes this is reversion. but again not going to affect the maf with the velocity of air comming into the engine. measure the air speed behind the maf next to the TB on a running engine at idle. itll surpise you just how fast its moving.


ye again unless the engine is backfiring the worst reversion will do is slow the incoming airspeed. it will actually indicate less air. also as long as the maf sensor hasnt been follwed with and the screens are intact then it should be pretty resistant to changes ducting and position. the worst moving a screened maf will do is slow down its respone time.

yet again put a vacum guage between a maf and a TB you will see that there are to many things in the path of that high pressure air to actually let any manifold pressure out. also there are low pressure zones in the intake IE other cylinder opening that woud tend to invite cross filling. even though on papaer your idea sound fanastic its not the problem with this vehicle. its most likely the exhuast valve being open to long and allowing a chunk of fresh mixture into the exhuast stream.

a good sound thoery but in practice just doenst work out like youd think it would. already done alot of testing. now there is a thing with reversion in the maf sensor itself where the air will actually turn around and pass the wire twice. ive seen it happen with a few of the after market MAF sensors. however the GM TPI maf in stock unported shape with screens is not susceptiable to this problem.

dnult

32V_DOHC and Grumpy: I guess I see the point your making regarding "wave action" and reversion. I'll give thought to how big a factor this will play on the MAF reading, doesn't seem like the effect would be very pronouced. In fact it seems like wave action would impact the MAF at multiple RPMs as the intake charge passed through the acoustic harmonics of the intake. But anyway, I get the point.

-dnult

funstick

well the wave action will be least pronounced at the actual MAF sensor. the wave most pronounced heading into the cylinder along the runner. trust me reversion into maf sensor is not cuasing this problem. it would be like next to impossiable to have reversion from the engie make it into the maf ( i feel like broken record) you know if this theory was so sound then craig moates would be fighting like ganbusters with his 4hole TB single plane large plenum big cam 383 motor and hes not. Ahh proof in the pudding the single plan high rise intake would exhibit this problem like 10 fold over a TPI intake. unless reffering to reversion specific to the actuall maf sensor itself ( which can happen put a descreend maf next to a 90* elbow ) then i think grumpy and 32v dohc are blowing smoke up everybodys ***.

ZZ28ZZ

Good news to report:
Disconnected the factory O2 sensor. Installed a heated sensor on the drivers side down-pipe just before the "Y".

Idle mixture, as reported by the WB-O2, went from >20-to-1 to abt 15.5-to-1. Idle is much smoother. No other changes were made.
This reinforces my theory abt fresh mixture making it past the exhaust valve and triggering a false rich signal.
Since the issue was only at idle, I tend to discount the heat and system back-press being the main problems.

I'm thinking that moving the new sensor a couple of more feet down the exhaust pipe would bring me even closer to 14.7-to-1, but I'm unsure what the effects of the associated reporting delay would be.

While the idle mixture still isn't quite where I wanted it, it should be OK till I can the get headers and cat-back installed.
In the meantime, I'll explore some of the new tables and constants that the $6e code offers, to see if any further progress can be realized.
I tried adjusting the "Closed Loop Rich/Lean threshold VS airflow" table for the airflow rates associated with idle, but the effects were very small.

Lesson learned: The WB-O2 is an awsome tool, and not just good for WOT tuning. Without it, I would have been chasing this issue for who knows how long.

Thx for all the replies, and a special thx to funstick for reviewing my log data and replying with his recommendation.

funstick

glad i could provided assitance. although i think you figured out the problem before you asked me. dont bother with those airflow tables and go back the $32b. i think youll find it works much better now with accurate reporting of the A/F ratio.