O2 gurus - need advice...

[ShelGame]

So, I finally had the chance to datalog a full pass (well, 3 actually) with the new engine this past weekend. And, I saw something in the datalog that I had seen before on my simple 2nd gear blasts on my test track (actaully, the nearly deserted road running paralell to US 23 near my house), but never understood.

See below.

My concern is the O2 sensor volts. It seems to take a long time (relatively speaking) to get above 0.8v when at WOT. The PW, though, jumps right up with MAP. Is this slow rise just a function of the O2 sensor? Or is something else going on? I've blown up 2 engines this summer, and I want to make sure I stay away from any kind of dangerous conditions...

BTW, there was a small amount of knock retard (not plotted) at the end of the 2nd and 3rd gears. Not correlated to the O2 volts at all...

[shadow88]

That's scary stuff. just above .6v at wot and a couple of psi boost.

I can tell you mine does not work like this. It's above .8v within a moment of hitting wot, even part throttle with a little boost. I'm not a guru, but this seems really bad for the life of your engine if it's correct.

[Frank]

So, I finally had the chance to datalog a full pass (well, 3 actually) with the new engine this past weekend. And, I saw something in the datalog that I had seen before on my simple 2nd gear blasts on my test track (actaully, the nearly deserted road running paralell to US 23 near my house), but never understood.

See below.

My concern is the O2 sensor volts. It seems to take a long time (relatively speaking) to get above 0.8v when at WOT. The PW, though, jumps right up with MAP. Is this slow rise just a function of the O2 sensor? Or is something else going on? I've blown up 2 engines this summer, and I want to make sure I stay away from any kind of dangerous conditions...

BTW, there was a small amount of knock retard (not plotted) at the end of the 2nd and 3rd gears. Not correlated to the O2 volts at all...

That is definently odd. basically it is being slow to react from 14.5a/f to 11.0a/f or something like that, but it has no problem reacting when you shutdown the run in comparison.

Normally a sluggish O2 is a result of carbon buildup or damage. Typically if you melt pistons or have a bad headgasket and ignore it, you have to replace the O2 sensor for this very reason.

ASE says that a reaction time of around 100ms for 200mv to 800mv is ok for a good sensor (new is around 25ms). They dont think that reaction times of 200ms is good even when looking at 100mv to 900mv. Even your deaccell is not showing good.


Frank

[ShelGame]

So, it's likely just a sensor issue then? The sensor was new, but that was 50(?) gallons of leaded race gas and 2 melted pistons ago. I'll get a new one and see if it's better.

It really doesn't make sense that it's tracking A/F. I mean, MAP goes to 15psi, and the PW follows it. Why would it still be lean?

[Frank]

Oh SNAP! Leaded race gas? That will do it! PM either Reeves or GLHSKEN for the sensor they use that works with leaded race gas.

[turbovanman²]

Do you have a labscope?

It would appear your ox is dead or weak. If you have some propane, you can see how fast it will go to pig rich. Have it hot and idling, inject propane into the intake-don't create a vacuum leak and see how fast it pegs, it should do it instantly.

I have been using a 4 wire NTK ox sensor for over 3 years, its survived numerous blown engines, tanks of leaded race fuel and it still switches under 100ms.

I think Denso makes a tougher one.

[ShelGame]

Do you have a labscope?

It would appear your ox is dead or weak. If you have some propane, you can see how fast it will go to pig rich. Have it hot and idling, inject propane into the intake-don't create a vacuum leak and see how fast it pegs, it should do it instantly.

I have been using a 4 wire NTK ox sensor for over 3 years, its survived numerous blown engines, tanks of leaded race fuel and it still switches under 100ms.

I think Denso makes a tougher one.

Yeah, but when my engine blew this last time, I had to scrape a thin coating of aluminum from the inside of the downpipe. There wasn't any visible buildup on the sensor, so I thought I'd give it a shot. When I saw it switching, I thought it was OK. But, to tell you the truth, the idle isn't quite right (rough, won't always hold speed), which could also be due the O2 sensor.

Good info, thanks.

[ShelGame]

Next question: why does the RPM data seem so "noisy"? This is real unfiltered data. The engine's rotational speed does actually change in the middle of a revolution as each piston fires and the opposite cylinder compresses. But, I'm sampling at much to low a rate to actually "see" any of that. I guess it's just sampling the measured RPM at that point in time (which could randomly be anywhere in the cycle).

I wouldn't be concerned except that the timing seems to follow the "noisy" RPM signal (~4deg change based on the "noise"). Anyone else ever notice this? Is some of the "noise" possibly due to slop in the distributor?

[turbovanman²]

Noise could be due to igniton maybe? crappy or leaking wires?

If you had to scrape molten metal out of the dp, I would say change the ox. I will say I've melted numerous pistons with this sensor and it still works??????? lol!

Frank, I have never seen an automotive Ox swith at 25 ms, lol! Even brand new.

[Frank]

Can you e-mail the logged file? RPM being outputed by the ECU is based on HEP triggers. If it is not missing, then it should be spot on.


Frank

[Frank]

Noise could be due to igniton maybe? crappy or leaking wires?

If you had to scrape molten metal out of the dp, I would say change the ox. I will say I've melted numerous pistons with this sensor and it still works??????? lol!

Frank, I have never seen an automotive Ox swith at 25 ms, lol! Even brand new.

I am just repeating what ASE says.

[turbovanman²]

I am just repeating what ASE says.

Thats fine, don't get snitty, Just telling you real world experience. I have to test them for emissions and I always test them in a tune-up, the 2-800mv is right as is under 100 ms. I usually recommend them if there around 80-100ms, as long as there not $400 each,

[ShelGame]

Can you e-mail the logged file? RPM being outputed by the ECU is based on HEP triggers. If it is not missing, then it should be spot on.


Frank

e-mail sent...

But, the period between HEP signals won't be constant even at a "constant" rpm. As each cylinder fires and/or compresses, the crankshaft will accelerate/decelerate. I think I may just be measuring the hi/lo. I probably need to filter the data in my chart. What bothers me, is that the SMEC apparently doesn't filter that signal when it comes to calculating the advance vs. RPM.

And, that makes me wonder if the noise is really mechanical in nature (ie, distributor or HEP slop).

EDIT: I just calculated it, and at max RPM, there are ~40 HEP pulses in between my RPM samples.