Dominic Sorresso

At the beginning of 2013 I had larger exhaust cams installed on my LT-5. Overall, this has turned out to be an great setup producing great power from a
5.7L. I'm averaging 24mpg at 75-80mph, up to 30mpg at a 60 cruise.
However, I have now gone thru 3 O2 sensors, all on the left side. When tuning, the BLMs on the left side are typically leaner than the right particularly at idle although an LT-5 has the split BLM issue and I am told it is not atypical to have a 20point difference left/right. In datalogs, you can see that the left side O2 is "lazy" and doesn't transition like the right side. However, when a new O2 is installed, the transitions are much more "normal".
So basically, my O2 is staying below .2v for longer than 50sec. to set the code.
The last time I went thru this, I swapped the O2s bank to bank, and the code followed the sensor so that it turned into a Code 64 Right side lean. So replaced the O2. Made a calibration change after running the car for the last six weeks w no issues. While idling in the garage yesterday, up pops the code on the left side. Now this only happens at idle, otherwise no problem. I can avoid the code by playing w the throttle before the 50sec. time limit.
I'm wondering if the O2 is being poisoned over time. Other causes of a lean condition, like vacuum leak or bad injector, would be much more consistent and wouldn't disappear w an O2 swap. Any insights are of course appreciated.

kevm14

My probably unhelpful suggestion is based on my LT1 experience. On those, exacerbated by a cam, you end up with split BLMs at idle. The fix is to close the throttle blades and force more idle air to go through the IAC circuit, which more evenly distributes the idle air to each cylinder. No idea if the LT5 intake has something similar or if this applies at all. But it would be fairly easy to try, I suppose.

Dominic Sorresso

kev,

Appreciate the reply. The LT-5 doesn't have anything like the IAC channels the LT-1 has if I understand u correctly. Although the IAC issue is another one I have been dealing w. Over time, the IAC count at idle continues to drop and it seems to be in line w the Coolant temps. Higher CTS, lower IAC steps at idle.
Sometimes it gets to 0 but the motor idles just fine w no sign of a vacuum leak.
Thinking of re-setting my Min Air.
Another piece of data on this. The 44 came on right after I loaded a new bin into the Ostrich. Now this has happened once before recently after installing new O2.
I cleared the code using my scan tool so as not to also clear the rest of the history. No problem since then until now.
The fact that this appears to happen over time and results in a bad O2 sensor takes me in the direction of a gradual degradation taking place and a poisoning of the sensor. And it's limited to the left side so nothing universal like fuel,vacuum leak etc. A bad or weak injector? Idle runs on primary injectors only. Secondary injectors come on line right after tip in and could make up the diff but why would that only show up over time each time. Why go away with change of O2 sensor?
I'll try re-tracing my steps today.

Jeffs82TA

iTrader: (1)

Had kind of the same problem on my 97 3800 f-body. Lean bank one, on mine turned out to be a bad exhaust manifold gasket. The I had both banks lean but that was a cracked egr pipe.

If you are having single bank issues you have to look after the runner gaskets, injectors, stuff like that.

also on my 97 after you wipe codes obd2 (if that is what the lt5 is) it must be driven to perform it's self tests prior to next accurate scan.

good luck though lean bank issues are a PITA to solve sometimes.

kevm14

Does anyone have details on how that ECM performs the DTC 44 diagnostic? Does the BLM just peg to the max (150, 160)?

Dominic Sorresso

This is happening at idle. Not anywhere else. In fact, the last datalog I did while cruising was like 132/124 BLM which is fine by me. But here's another little piece of data. Just datalogged idle. Started up "cold". In O/L I could see the left side O2 registering rich. .73, .6, .8 etc. it went into C/L at about 65C w Learn OFF, ran about 2 min that way. Then fans kicked on at about 77C and SES came on w
Code 44. At that point O2 basically flat lined at idle. It would change when I hit the accelerator and register rich but otherwise it was at .04-.06v. I have WB on right side and it was still showing 14.5-15.3 AFR oscillating between those points. Learn was OFF throughout, so BLMs were at 128. I cleared the code without disconnecting battery, and ran it again. Code came back.

kevm14

What are the chances that something is amiss with the O2 heater circuit/ground and there isn't enough current to keep it properly hot? Of course there's probably a diagnostic for that, too.

Dominic Sorresso

I don't think that explains why a new O2 works fine for several weeks. Something over time degrades the performance of the O2. I have had O2s last for years. Going to pull the plugs on the drivers side. May use laser thermo to look at exhaust gas temps on that bank also.

Dominic Sorresso

Pulled plugs on left side. They go 1-7 left to right. #1 looks a bit leaner but nothing overt IMO.

kevm14

My thought was that a fresh O2 sensor works OK for a little while with a weak heater but quickly degrades with use. Could help explain why it's only at idle. I could be way out in left field, I dunno.

How does the diagnostic work? You said it's lean but it seems like it reads rich right before the DTC sets, then it reads very lean. It is interesting, however, that one time you swapped the suspect left sensor to the right and the code followed.

Dominic Sorresso

It was ok in O/L. O2 showed voltages that were rich. Not until it went C/L did O2 drop to .04-.06v.

kevm14

What was the BLM/INT on that bank while in CL but before the DTC, or were you not able to catch that? It's weird. Sort of like the ECM is pulling its own fuel, then complaining it's too lean.

Dominic Sorresso

This is getting a bit more curious. Ran the car yesterday, went C/L but no Learn,
O2 v dropped once it went C/L finally resulting in Code 44. Then it went back
to O/L. BOTH the left and right O2s basically flatlined at that point. I have a datalog of it and will try posting it here.

kevm14

What about letting it transition to learn while driving, then idle it and log behavior? It doesn't seem to be in learn mode long enough to justify triggering the code...

Ronny

iTrader: (1)

I suspect you are using Delco sensors. Hot rod magazine just ran an article on cheap Chinese knockoffs. Are you buying from a reputable source?

Dominic Sorresso

I am cautiously optimistic right now but don't want to jinx things. However, I made a change to the bin I have been using. Some of you who are running larger cams now know about how you can manipulate the O2 voltage values
to control that over-rich condition brought about by increased overlap of bigger cams. Traxion was one of the first I recall to point in that direction, and I began using that method with my 84 Xfire. It worked.
More recently I have uograded my LT-5 motor with porting and larger Intake/Exhaust cams. The LT-5 is an SD motor so there's quite a bit of calibration changes that translate from the lowly Xfire to a DOHC motor.
I think what I found was that the changes I had made to the calibration regarding the O2 window and the swing points may have been too low.
I have attached views of my calibration both prior to and after the change I made that appears to have eliminated my Code 44 at idle. Just as a reminder, Code 44 pops up after the voltage for the left side O2 sensor has been below .2v for longer than 50sec. In reviewing my calibration, I took a SWAG that I was forcing the O2 voltage to be below that .2v threshold. The LT-5 suffers from the same "split BLM" or LT motors have and the left side is the leaner of the 2. The disparity between banks is greatest at idle. Personally, I would like to do individual fuel feeds to each side of the fuel rails. My hypothesis is that the fuel pressure likely drops between right to left and that requires a bit larger PW for the left side. Its similar to the use of a larger fuel injector for the left side TBI unit on a Xfire induction system. Requiring larger injectors on an Xfire forces the use of parallel plumbing to use the same size injectors on both sides.
Since the left bank is typically leaner, it will be the left bank that will go lean sooner than the right. So as I continue to evolve a tune for the motor, I finally get it to the point where I trip a Code 44. Through the use of the O2 swing point and window, I was forcing the ECM to keep the O2 below the threshold for a Code 44. And the left bank is the first side to satisfy the requirements for setting a Code 44.
It appears that the solution is to raise the O2 window/swing values to at least equal to .2v, the Code 44 threshold. I made that change as you see in the O2 POST attachment and went out for a drive and logging session. Although the left side BLM was lean at idle (137-140), I could also see that overall range of movement was greater and that it appeared much more "alive". Once in my garage, I sat there for over 2 minutes at idle. No code. I could see the
left O2 moving above and below the .2v threshold w regularity. It seems to have eliminated my issue. I had not made any changes to the O2 setting on anything more than idle keeping stock O2 settings for all other airflow levels.
Please feel free to critique. Am I on the right track here?

Ronny

iTrader: (1)

When would the motor ever see 0 grams/sec? At a stall? Or is that to allow interpolation from 1-15 g/sec?

At 16 g/s I see values over 200mv so i presume the tables shown at right are corrected. Stock tables at left.

Why did you leave CL Low and upper window at 199 vs say 205?

Dominic Sorresso

Ronny,

The 0g/sec is idle. Anything below 16g/sec. The pre O2 are the settings I now believe were forcing the O2 to stay below the Code 44 threshold. The POST is what I changed them to.
The Low O2v is the stock value so I started from there and maintained the same interval.

kevm14

Yes, a method that GM has employed on certain vehicles for the AIR system.

So you previously compensated for the greater overlap, but postulate that you over-compensated? What was the INT doing during the 50 seconds of low O2 voltage? Or maybe the real problem is the O2 voltage criteria for ENTERING C/L is the real issue? I would expect the INT to be climbing under these conditions.

I didn't realize you modified your swing thresholds from stock. Because I WAS thinking about your more aggressive exhaust cam, but I expected the stock behavior to be an increase in INT/BLM (possibly to cause a DTC 45 instead), rather than a DTC 44. I also didn't know the DTC 44 criteria.

Certainly you could manipulate the switching voltages to remedy this.

I think idle conditions would be the most sympathetic to any fuel rail flow bias. Like an electronic circuit, if you have a resistance, but low current flow, the voltage drop will still be low. Any rail issue that causes a discrepancy at idle I would expect to be a BIG problem under high load. If you were really curious, you could potentially drill and tap a test port on each rail and see for yourself if there is any pressure difference.

That's why I asked before about the criteria for DTC 44...

I think a better solution would be to make the swing thresholds whatever you think they should be, then modify the DTC 44 criteria, but that may require code hacking.

How split are they under load?

Dominic Sorresso

Ronny,

Bosch from AZ.

Dominic Sorresso

"Yes, a method that GM has employed on certain vehicles for the AIR system."

You mean EGR?

"So you previously compensated for the greater overlap, but postulate that you over-compensated? What was the INT doing during the 50 seconds of low O2 voltage? Or maybe the real problem is the O2 voltage criteria for ENTERING C/L is the real issue? I would expect the INT to be climbing under these conditions."

LEARN was not active at this point. Code 44 is set regardless of O2 parameter settings. As I stated previously, a DTC 44 is set when the ECM sees O2 voltage of <.2v for a minimum of 50sec. This is not modifiable or at least I would not be interested in doing so.

"The LT-5 suffers from the same "split BLM" or LT motors have and the left side is the leaner of the 2. The disparity between banks is greatest at idle."

"Certainly you could manipulate the switching voltages to remedy this."

Switching voltage is universal not by bank.

'How split are they under load?"

A disparity of 20points between left/right bank BLMs is expected at idle on an LT-5.
In my last datalog I had a 140/120 BLM split. Also, I expected the BLM to go higher given that I had raised the O2 voltages for idle. At cruise the spread is 133/124.

Dominic Sorresso

So far my hypothesis seems to be bearing the result I'm looking for. No more DTC 44 at idle. With a re-done VE, the split BLM at idle is now at 132/118. Idle is much smoother. O2 transitions more active than previous.

Dominic Sorresso

RBob,

Given that you haven't commented so far, can I assume I'm on the right track?

Thx,

RBob

iTrader: (1)

Being that the issue is at idle, I don't believe that the fuel rail feed/return has much to do with it. If it was at higher engine speed and load, that can have an affect on fuel distribution.

From what I can tell reading through and looking at the original O2 low/high switch points, it was too low. Below the trip point of code 44. If those low swing points were the ones to use, lowering the code 44 threshold should have taken care of code 44.

But the O2 swing needing to be that low has been dis-proven. As you raised them the idle became better.

RBob.

Dominic Sorresso

RBob,

Thanks. Exactly the kind of info I was hoping to get. The fuel rail thing was somewhat extraneous. I only brought it into the discussion because of the split BLM issue and that I was drawing an analogy to the Xfire TBI setup. Given that those TBI were fed in series, GM needed to install a somewhat larger injector in the rear tbi in order to equalize fuel flow. I was suggesting something similar was happening w the L/R fuel rails on the LT-5. But my main point was that would help explain why the left side is usually the leaner of the two, since both sides use the same size injectors.
But you are correct that the fuel rail did not have anything to do w the issue I was dealing w.

If the Code 44 threshold (.2v) is modifiable either w a time or voltage parameter, I am not aware of that and since I don't do code, didn't see a way to change it. As it is, the motor idles better by raising the O2 swing points at idle air flow.

So RBob, just to clarify your statement re: O2 swing points and larger cams/greater overlap.

Are you saying its been "disproven" that they need to be that low (ie as low as I originally had them)

OR

Its been disproven that the swing points don't need to be lowered at all?

Dominic Sorresso

" If those low swing points were the ones to use, lowering the code 44 threshold should have taken care of code 44."
-RBob

Ok, went back to the cal and looked for this. What I do see is the ability to uncheck the bit for setting the Code 44 MALF. However, I think more critical than idle is keeping it for operation in "higher" rpm zones.

RBob

iTrader: (1)

This one.

Not this one.

RBob.

RBob

iTrader: (1)

Some parameters so rarely need to be changed that they don't end up in the ECU/XDF files.

RBob.

Dominic Sorresso

Again, thank you for the clarification. So the original premise of manipulating the O2 swing points is valid for accomodating larger cams, particularly at idle. Apparently I went a bit overboard but hey I learned something from the experience.

So, has there been any determination as to:

a. how much the swing points should be lowered?

b. Is it independent of the cam and motor?

c. How do you know how far to go?

BTW, along w raising the O2 idle voltage window, I also increased the Prop Gain, again at the 0-15g/sec range. I'm thinking this also contributed to "taming" the idle. I recall something along those lines in your "white paper" on PID controls.

Cheers and thx

Dominic Sorresso

I'm thinking this is one of them.

Dominic Sorresso

Anyone?