Suddenly pig rich

So there I was, happily tuning along, in pursuit of the ideal PID settings for my IAC in order to get a good smooth idle and making actual progress, when after a rather minor amendment when I loaded the tune and restarted, the engine had lost it's mind.

All sense of reason completely thrown out the window, it began behaving like it had buckets for injectors and very quickly would not start at all.

Why would it do this to me? Anybody have the remotest clue what single failure could cause it? 'Cause I'm all ears.

I blamed the ignition switch and went on a 2 month long campaign which eventually got me right back to the same point of failure. Naturally I first tried earlier and earlier backups all to no avail. I've forgotten which all of the parameters I checked when it first happened, and so I'm starting back over with injector size, but I don't think that's it since I tried 3 different controllers with 2 different OS's and half a dozen or more different tunes, all with the same result.

It has to be a failure in the vehicle wiring or sensors, right? But why rich?

Your suggestions most gratefully received,

Jim

'411 P-01

I will suspect some mechanical failure of injector or fuel pressure. Leaking injector can lead to pig rich fast. Higher than expected fuel pressure can also make it pig rich if you have oversized injectors. The injector failure can be intermitten, making it really hard to pick.

Look at Temp and BARO pids, make sure they aren't crazy numbers.

Biggest influences of mass fuel pulse width change

Thanks. I think I'll try to run a datalog tomorrow and see how it looks. I've been thinking about what kind of failure would do this and it seems like the most likely suspect could be the MAP sensor if it's possible for a failure to look like full pressure. With a 2 bar sensor that should be able to cause a massively rich condition. But first I should probably look at the injector times. That should show something is up. Or at least divide the troubleshooting between mechanical and electrical.

Jim

I think the datalog is going to be usable but the extension is not an acceptable file type here. I seem to recall there was a way to still post it but I forget, something about changing the extension I think. Wish we could fix that.

So I took a screen shot. But of course I can't just paste that here. Something else I wish we could fix. Pasted and saved it in Paint. I think that will have to do.

Now, I'm not positive that My scanner is set up right but there are some things about this MAP trace that certainly are NOT right. Like an average kPa of 200 when cranking. Note the engine fired off only after I had unplugged the fuel pump relay. Then the MAP pulled down as far as 53 before heading back up to 131 after the engine had stopped. So why didn't it at least drop to 131 while I was pulling the fuel pump relay? The engine was not cranking then.

Now to start with, since it is a 2 bar MAP sensor the range of 0-210 should be correct. However, MAP at cranking speed should be around 90, not twice that. So it sounds like something in the MAP parameters are not right. This confuses me further since it had been working correctly to all appearances. I do not see a clear way that a wiring or sensor failure could cause these errors. I wonder what would happen if I just unplugged the MAP sensor?

Jim

OK that seems to confirm the MAP as the problem area, with the sensor unplugged the engine fires right up but then dies. However it sounds like it might continue to run if it was warm. So, it looks like I need to sort out why the MAP has those readings... not really sure about that one. I played with MAP settings early on when I was trying to find a way to deal with the blower but haven't looked at it since I loaded the 2bar OS. Apparently something there isn't quite right. Back to my studies I guess but at least I have some idea of where to concentrate.

Look at MAP key on engine off and compare it to BARO. Hard to read values in screen shot in respect to cursor placement.

That's part of the oddness. Baro should be somewhere around 99 at 600ft above sea level where I am, though I don't know where to look for that reading (any suggestion?) KOEO the MAP reads zero until the engine begins cranking and then jumps to 197.8 kPa and holds there until the engine fires off, at which point it drops as the RPMs come up, hitting 53kPa or about 13.6 in.Hg on that trace. Considering the throttle was very slightly cracked and the IAC may have been open some as well that could be pretty close to real values for about 2500 rpm. This is taking all readings from the "Manifold Absolute Pressure - 2 bar" PID (channel) in VCM scanner. So the running values seem in the ballpark but the cranking value is whacked. And then the value after it died (ran out of gas because the FP relay was pulled) came up to 131 kPa which again is whacked.

I went back through the receipts and swapped sensors, apparently I was sent a 2 bar instead of a 1 bar (no P/N or visible external difference). Doesn't explain why it worked and then it didn't but now it's at least starting and showing a lower MAP reading. Still not right but it looks like I may have a usable sensor now. With the right part number I'd willingly buy the one that gives the ideal output.

This sensor is either P/N: 12615136 (Camaro) or 16235939 (GMC). There could be pin connection differences. I'm wired for Signal on B, +5v on C, and ground on A. Photo below. Pin A is on the left of the connector with the wires up and facing the latch.

After swapping the sensor the engine starts easily and then dies. (I'll work on that once the MAP is sorted out) Map reading is roughly inverse to RPM however initial KOEO gives a zero reading. After the first start/die cycle the MAP reads a steady 65.8 kPa which I'm sure can't be right as I'm at 600 ft above sea level. That comes out to slightly less than 10 in of vacuum which is ludicrous with the engine stopped. It dropped to between 22.2 and 26.8 on the three start cycles I ran, or around 21-22 in, of vacuum, a reasonably good number. probably within range if the at-rest reading was 99.

So what I'm seeing here is that it looks as if the OS is calibrated for a MAP sensor that is not the one I am using. Do you guys concur?

Jim

Key On Engine Off (KOEO) MAP will read BARO. So if your local BARO is 99Kpa the MAP should read 99Kpa. BARO is generally calculated in the PCM from MAP at KOEO. Now you could be looking at a "corrected" MAP pid that would read 0Kpa KOEO, those are used for reading actual boost. So try looking for more PIDS that may contain MAP or BARO and try to eliminate something that is "corrected".

Good call. Not real sure what exactly that PID was, it was the only MAP channel labeled as 2 bar. However once I activated the SAE and voltage channels I got a better picture of what was going on. I have it reading correctly now and I have a 2 bar sensor on the engine with a genuine Delco one on the way. I will test it again once that gets here and probably use a pressure rig for the boost part of the range.

The only way I can make sense of the sequence of failure is that possibly there is a failure mode in the Chinese 2 bar sensor that makes it revert to 1 bar? That would explain the failure and I can think of nothing else that would, because I was able to tune the engine and then I wasn't. The sensor I removed is outputting correctly as a 1 bar, though it may be a bit over sensitive on the low end. Probably should go in the round bin, in fact I just did that. It's been more trouble than it was worth.

Looks like I'm back to regular tuning anyway. Concentrating on startup, warmup, and idle for now. That and learning how to properly set up the traces for the AFR. I'm using the Innovate MTX-AL with the EGR input on C1-55 so it's not just plug-n-play, and so far I've struggled with it. But I hope with experience I'll get better at setting up the scanner. That would need to be another thread though.

Jim

Hey guys. Thought you'd want to know the outcome. To make it short, a self inflicted wound of course. Be careful and methodical when turning off error messages. Turn off the wrong ones and (so I've heard) the ECU won't exit closed loop or something, which makes everything act very strange if you are tuning in speed density. So I was able to track down a tune from way back on the 9th of June and that one worked fine. Started good, idled down, and no surge. I'm down to 800rpm on the idle now, something this engine has never done before. And I'm pretty sure it can go slower, but with an aluminum flywheel it probably shouldn't.

Anyway, I got my scanner pretty well sorted out at last and have been running some histograms to edit the VE table (copy/paste, now below 3.0 and dividing by 2) and my AFR is getting pretty close to commanded afr. This is after warmup and at very light throttle settings. And this brings up a question. Probably a stupid one, but if I don't ask I won't learn so here goes. I'm sure I'm not the first.

About that commanded afr. That's pretty close to 14.7, actually a bit richer than that. I get that PE takes care of heavier throttle applications and I really haven't been playing with that yet. But if the commanded afr is trying to hit 14.7 and the engine needs more fuel because it's starting to take on a load, and the PE hasn't kicked in yet, what happens? I mean, at light throttle I can see it all working rather nicely and I suppose the EQ Ratio Subtractor settings take care of lean cruise once those are set up (mine are all zeros), but I'm a little worried about adding power without more fuel.

Any recommendations?

Also I'm still predominantly rich, black soot in the tailpipes, under 12 when I hit the gas so I should probably lean out the PE once I figure out how but presently the car is running quite good.

Jim

Quote:

---

Originally Posted by Jim Blackwood

About that commanded afr. That's pretty close to 14.7, actually a bit richer than that. I get that PE takes care of heavier throttle applications and I really haven't been playing with that yet. But if the commanded afr is trying to hit 14.7 and the engine needs more fuel because it's starting to take on a load, and the PE hasn't kicked in yet, what happens? I mean, at light throttle I can see it all working rather nicely and I suppose the EQ Ratio Subtractor settings take care of lean cruise once those are set up (mine are all zeros), but I'm a little worried about adding power without more fuel.

Jim

---

If I understand what you're asking, you wont be able to tune LOAD areas of the VE table until you actually introduce load via a dyno or the street. In either case there is no reason why you cant just sneak up on those areas of the map with out damage to the engine. You'll have to have the VE table dialed before you can use the PE table to command a proper AFR.

OK thanks. So do I need to zero out the PE tables to work on the rest of the VE table?

Jim

You could, or set the PE delay timer to some un-reasonable value. They're generally set pretty high anyway.

OK I'll do that.

What I'm seeing for AFR Commanded ranges between 13.2 and 13.7 which seems rich to me. My measured AFR is running 12.2 to 13.5 mostly, with an offset generally under 1.5 and this is at very light throttle application, generally around 20-30 kPa.

What would be the best way to bring these numbers up closer to stoich?

Jim

Slowly creeping up on it. It's got this thing right now where it idles down to 800 but then when I start driving it the idle stays up around 1300. I need to sort out what is causing that. I think it could be the IAC since it is open more (~30 vs 20) but I don't know why. I seem to be closing in on stoich, though commanded remains about 14.0, I guess I can pad the numbers for commanded AFR by maybe .7 to 1.0 in the formula to get the correction once it has settled down a bit more if it's needed. No real hurry now. It's about close enough now to be able to make a fairly accurate comparison between the trace and the AFR meter.

Starting is pretty good but warmup seems to be richer than it should be so that will be next. Transitions seem pretty good although initial off idle is a lot less crisp than it should be, makes me want to blip the throttle too much and that shouldn't be necessary at all. Should be adjustments for that too.

You guys are the best btw, in case I haven't said that lately.

jim