Intermittent O2 reading

[Dont fake it break it]

87 Blazer 5.7, recent top end rebuild, ALL new sensors n pigtail connectors, Tuneup components, exhaust manifolds, etc.
So I finally purchased an Actron scanner to help figure out my lean and rich conditions, funky idle, running like garbage, etc. What I mean by lean and rich is that it historically runs lean but failed a smog due to excessive CO2 and the tech told me it was running rich. After reading a lot of posts and websites, I think I have an understanding of what I’m looking at. First it appears as though my INT and BLM are only normal in an open loop. When I enter a closed loop they both tend to rise. INT will hover around 145-ish but the BLM will hover around 185.

I know this means the computer is dumping excessive fuel to accommodate what a perceives as excessive air. I can find no vacuum leaks although while watching the scanner, I rev it up to 2000 for a minute and those numbers don’t come down at all. In fact the BLM will keep rising. I think that would indicate a vacuum leak. That being said, I also noticed intermittently that the O2 sensor reading drops to 4mv Which I take to be zero signal and sometimes stays there for a few seconds or in one case I recorded, a full minute and never came out of it. Of course I went back into an open loop at that point in the BLM and INT return to normal. So after all the time I’ve spent chasing this down over the past month or so, the one thing that seems consistent is the computer making adjustments based on information received that does not necessarily reflect a component malfunction.

Again this is based on all new parts. The only things not new or the throttlebody housing itself, intake manifold, block, pistons and rods, oil pump, and that secondary air unit Diverter valve thing which incidentally is always pumping air into the emission tube and air cleaner and never stops doing so.The first of my last two recent recordings went like this…

Idle around 1000 which seems normal-ish for this truck anyway,
02 sensor acting normally bouncing back-and-forth between 100 and 900.
02 status alternating between lean and rich.
INT rose to 145 and BLM up to 181 only stopping when the testing ended. This is the best it’s ran up to this point since the rebuild. Before the rebuild it actually ran better. LOL.
A minute later I restarted it for another test while changing absolutely nothing. The results were much different and went like this…

Fired up but died instantly which is very unusual after the rebuild. Usually two cranks and it’s on. Period. Each time it died it had opened in a closed loop. It’s plenty warm by now so I’m not sure if that’s relevant.
When it catches and holds an idle The O2 sensor steadily drops from 300 MV down to 4 over the next 30 seconds.

The BLM and INT stay around 130 for 20 of those seconds before starting to climb. By the time 02 drops to 4MV, INT is at 141 and BLM is at 139.
The last reading before it went into open loop was BLM 150 and INT 137. Now keep in mind those two numbers are normally there or higher at idle especially the BLM, when it’s idling in a closed loop as per the previous test. It continued to idle out 1000 in an open loop for another 30 seconds before the test ended.

So my primary question after this long diatribe which I am very sorry about, is why would the O2 sensor reading fade away into no signal at all and stay there for that length of time when nothing has changed between the prior start up and test only a minute before? My second question would be even when the O2 sensor is functioning normally which it does a lot, it’s still dumping plenty of fuel at idle and acceleration.

The test I mentioned previously indicated no vacuum leaks and my fuel pressure after the filter tests at 14 LBS. The rest of the fuel delivery system has been rebuilt last month with new injectors as well, I was about to rebuild the pressure regulator, which was rebuilt two years ago, until I got this lack of O2 reading which seem to take precedent. I may have a fuel delivery issue with a leaky regulator or something but the lack of consistency between these two and several other tests is baffling me.

Should I just buy another computer? Does it seem logical and reasonable after replacing every sensor to have a new computer that will read the information and respond correctly? Thanks so much in advance for any thoughts.

 

[Bent77]

GM ECMs are fairly stable, unless something has been done to damage them

1) Was the engine rebuilt due to some issue consistent with what's happening now (ie did this problem exist before rebuild)

2) If no to #1, first presume the engine is correct. From that look for bad grounds, pinched wires, or even an exhaust leak or open pipe near the O2 sensor. If the O2 sensors get some smoke off new gasket sealant (or even oil haze burn off), they can also be damaged

If you confirmation of the vacuum is correct, it would lead me to believe its not in the engine part, but rather in one of the 2 above details

 

[Dont fake it break it]

Thanks for the reply Bent77. The reason I rebuilt it was because it failed smog. I figured it had an intake leak possibly and then one thing led to another. There was no failure that led to it though. The truck ran fine and I’ve driven it hundreds of miles at a time without issue other than it would run a little hot if I had to climb hills. Big hills would overheat it. That being said yes there was a condition that existed before, the lean condition. I would get that idiot light now and again while driving and it would go away if I stomped on it for a few seconds. It happens much more frequently now though which to me would sound Like a result of something I did such as manifold or TBI leak, but I think I’ve been through everything.
When replacing sensor pigtails, I did a brief look at the wires and most seemed OK. Anything crusty or damaged got replaced completely or with a splice.That’s definitely not conclusive out there not being an issue there.I’ve been meaning to recheck the ECM ground but it’s where it was before, attached to that little bracket on the back of the passenger side cylinder head. Both exhaust manifolds are new. I have found a crack in the passenger side and replaced it for that but with a continuing lean problem, replace the driver side as well. New gaskets of course and torqued down several times over the last few weeks. I don’t believe there’s any leak there. I took a torch to the O2 sensor and checked the voltage and it seem to be working OK but good information about oils and sealants affecting it. It would bounce between 1 and 9 as I applied heat and took it away.
Coolant sensor, map, EGR and solenoid, IAC, PCV, TPS, all check out OK. The IAC pintle seems to fully seat as well. I was trying to research how to check I vacuum at idle but not sure where to check it orWhat it should be at. Not even sure if that’s a factor but it was on the list.

 

[Bent77]

I use the volt meter, checking how much resistance I have on those ground points to the battery. I actually added several wires, essentially making each point of ground connected to at leas one other point of ground, and an extra to the throttle body

one of the vacuum ports on the front of the TBI should be what you want. Add a T fitting so you can start and run with the gage inline

 

[Dont fake it break it]

I use the volt meter, checking how much resistance I have on those ground points to the battery. I actually added several wires, essentially making each point of ground connected to at leas one other point of ground, and an extra to the throttle body

one of the vacuum ports on the front of the TBI should be what you want. Add a T fitting so you can start and run with the gage inline

Understood on all that. So what amount of vacuum would be a normal reading at idle?

 

[Bent77]

Depends on what you did to the cam.

I’ve never had a stock TBI, so I don’t have a reference for what that engine should do

Most 350s with a mild engine should see 17” or more

 

[Dont fake it break it]

Depends on what you did to the cam.

I’ve never had a stock TBI, so I don’t have a reference for what that engine should do

Most 350s with a mild engine should see 17” or more

The cam and crank are both stock. I didn’t pull the motor when I did all this work so they’re original. I’ll check it out. Thanks.

 

[bp71k5]

I might borrow or get another O2 sensor as well to confirm it’s good. If it does the same thing, you have a good known spare. Would also check it’s wire to make sure it’s not half melted to something or pinched off.

 

[Dont fake it break it]

I might borrow or get another O2 sensor as well to confirm it’s good. If it does the same thing, you have a good known spare. Would also check it’s wire to make sure it’s not half melted to something or pinched off.

So the sensor of course comes with its own wire lead and connector. I tried to find a replacement for the connector that receives it but couldn’t. It’s cracked and broken but the contact inside seems fully intact. It fits together and makes contact. I actually removed both connectors one point and just twisted the wires together to eliminate the possibility of a bad contact and that didn’t exactly have good results so I undid it. Maybe that connector is a resistor of some sort? I don’t remember if I tracked that wire all the way back through to where it enters the firewall but I will now.

 

[Bent77]

That connection is not a resistor, but that system is sensitive, and the wiring being undone or wonky may be something you want to address

 

[Dont fake it break it]

That connection is not a resistor, but that system is sensitive, and the wiring being undone or wonky may be something you want to address

Here’s something funny that happened. Hooked up the scanner and started it up and it idled fine. Decided to take advantage and go take a quick drive. When I put it in reverse the idle drop down to 600, but then started to steadily climb. Now I’ve seen this happen before on my scanner but what I noticed is that it directly coincides with no reading from the O2 sensor. I mean instantly. I put it in park giving it a few seconds in the idle and O2 sensor return to normal. Put it in reverse and the O2 sensor immediately went to zero, it went into open loop, and the idle steadily climbed up to over 900… in gear!

 

[Dont fake it break it]

That connection is not a resistor, but that system is sensitive, and the wiring being undone or wonky may be something you want to address

so here is the latest information. I began to suspect the brand new O2 sensor so I went and replaced it with a new Bosch part. Everything seem fine and thought I had it solved. Put it in reverse to take her for a drive and the idle started climbing again up to almost 1000. I shut it down and did an IAC reset when I noticed the pintle was fully open at idle allowing a huge amount of air past the throttlebody. The reset seem to work and everything is fine at idle so I took it for a drive. Got 30 yards and it went into open loop receiving no signal from the O2 sensor. No I suspected the wire so I jumped a new wire from the new sensor to where the wire comes out of the firewall. It worked great and thought I solved my problem until I put it into gear again. I actually had to disconnect the O2 sensor and reconnected again to get it back into a closed loop And I did so while it was idling at over 1500. Every single time I put it into gear the O2 sensor instantly drops to zero and it goes into an open loop. Sometimes it’ll register a reading for a split second but always goes back to zero and certainly not for long enough to enter closed loop. Well I thought that was true until I put it back into gear again, watched it go into open loop but then fight it’s way back into a closed loop.. first time ever. Then when I put it back into park the idle continue to rise until I had to shut it off at 3000 RPM.Now my IAC is completely open with the engine off and here we go again. How’s that for a vicious cycle?

 

[Bent77]

That’s a new one on me

 

[Dont fake it break it]

That’s a new one on me

So the way I have to think about it is this… What can I have done to create this in the last few weeks? The only thing I can think of that I’ve done is replace the IAC valve and numerous resets, and now the O2 sensor. I reset the timing a couple times using the proper method, after accidentally bumping into distributor. I did a TPS relearn as well thinking that might be the problem but it’s voltage is fine after bending the Taub this way and that way to get the voltage I wanted. I think it’s a .7 if I remember right. At one point I arced the coolant temperature sensor when I dropped a wrench on it because I have crimp connectors on top of it all Mickey Mouse style, but I tested it with resistance and voltage last week and it seem to function according to the temperature gauge in the truck. I may have to check it again but still can’t imagine how that would create a cycle like this as opposed to a constant issue. So scratching my brain to think what wires may have been crossed to create this scenario. Could arcing the coolant temperature sensor have damaged the computer? I’m really starting to think I might just buy one for the 85 bucks.

 

[dyeager535]

What does the scanner say for coolant temps? If it reads right, then you haven't screwed anything up. ECM failures are incredibly unlikely, when you start leaning that way you lose focus of other potential causes. It gets easier and easier to blame the ECM.

I try not to swap out known good components, although I do have a stockpile (my own parts store) and when it's quick and easy, will toss new or known good parts at a problem just to simplify the process. 99 out of 100 times, that hasn't worked. It's the one-offs...loose terminal connector, intermittent component failure, etc.

 

[Dont fake it break it]

What does the scanner say for coolant temps? If it reads right, then you haven't screwed anything up. ECM failures are incredibly unlikely, when you start leaning that way you lose focus of other potential causes. It gets easier and easier to blame the ECM.

I try not to swap out known good components, although I do have a stockpile (my own parts store) and when it's quick and easy, will toss new or known good parts at a problem just to simplify the process. 99 out of 100 times, that hasn't worked. It's the one-offs...loose terminal connector, intermittent component failure, etc.

Yeah the coolant temp is good. I wish it wasn’t so at least it would give me something to go on. LOL Believe me if I thought I was going to end up pulling the computer when I first started the rebuild, I might not have stuck with the stock motor. However I think I may be found something else. So putting it in gear puts a load on the engine. I’ve tested that enough times with the same results, IAC goes up, O2 goes down, RPMs go up, goes into open loop.So I decided to mash the break and noticed a slight change in RPM. Nothing real measurable because it’s always bouncing between 925 and 1025, but I could hear it.I turned on the heater full blast and it created the same result as putting it into gear. Great news right? Then I shut it off and started it up again to test it a second time and the computer fought back. IAC went up a little, O2 went down a little, RPMs went up a little, but it didn’t go into open loop. I don’t know much about the vacuum or the load on the engine but I have not found a vacuum leak anywhere and a test at the throttlebody was 17 to 20 psi. I hear that’s normal. Any thoughts on why an increased load would have this effect on an engine??

 

[bp71k5]

Yeah the coolant temp is good. I wish it wasn’t so at least it would give me something to go on. LOL Believe me if I thought I was going to end up pulling the computer when I first started the rebuild, I might not have stuck with the stock motor. However I think I may be found something else. So putting it in gear puts a load on the engine. I’ve tested that enough times with the same results, IAC goes up, O2 goes down, RPMs go up, goes into open loop.So I decided to mash the break and noticed a slight change in RPM. Nothing real measurable because it’s always bouncing between 925 and 1025, but I could hear it.I turned on the heater full blast and it created the same result as putting it into gear. Great news right? Then I shut it off and started it up again to test it a second time and the computer fought back. IAC went up a little, O2 went down a little, RPMs went up a little, but it didn’t go into open loop. I don’t know much about the vacuum or the load on the engine but I have not found a vacuum leak anywhere and a test at the throttlebody was 17 to 20 psi. I hear that’s normal. Any thoughts on why an increased load would have this effect on an engine??

you know there’s a switch that tells the computer when the trans is in gear and when it’s in park. Might check the wiring on that.

if your O2 sensor is loosing connection, you should get a check engine light turning on. Have you checked for codes on the computer (and not the scanner)?

 

[dyeager535]

you know there’s a switch that tells the computer when the trans is in gear and when it’s in park. Might check the wiring on that.

Pretty sure the scanner should also tell you if it's in park or drive.

The scanner kicks the idle up to ~1000RPM as well, correct? I really haven't dealt with TBI a ton, I can't remember if it does that when you have something hooked up to the ALDL or not. I know TPI does.

 

[Dont fake it break it]

you know there’s a switch that tells the computer when the trans is in gear and when it’s in park. Might check the wiring on that.

if your O2 sensor is loosing connection, you should get a check engine light turning on. Have you checked for codes on the computer (and not the scanner)?

Thanks for the input. I will check the wiring on that switch but I thought it was only to prevent a start in the wrong gear. Yes I will definitely check the codes on the ECM. At this point they are all repeated lean condition code 44. So my understanding is that and open loop is created by the O2 sensor being below 600° or coolant temperature sensor not reaching whatever temperature that parameter is. At this point it seems to me the O2 sensor is not failing to transmit data, is that the ECM has already chosen to ignore it because something else has occurred to put it into open loop. I have no fluctuations in my coolant temperature on the scanner nor on my dash gauge. So what else would do that? Misfire? I do have a bit of a choppy exhaust from time to time but usually when it’s warming up.

 

[dyeager535]

Some of the GM TBI setups were programmed to go open loop at idle.

The park neutral switch input to the ECM is partly to account for the torque converter load. But it also indicates to the ECM that it needs to watch the VSS, it will want to drastically cut fuel if it thinks it's out of gear and you are decelerating, which can lead to stalling.