LM-1 telling me something strange...

sammyg2 · 03:08 PM

I just hooked up my new LM-1 a/f gauge with wideband O2 sensor, it it telling me something i halfway expected but don't quite understand yet. with lambda brain not yet connected:

At idle it is 14.8 which is about what i expected and where i figured I had it set. But as soon as I get going, it goes rich. At anything above 20 mph on level ground it reads from 12.5 to 13.
That explains the failed smog test.

Now here is what i have already checked:
1) vacuum leaks, none that I can find.
2) WUR, recently checked control and system pressures, within specifications.
3) decell valve, checked it and it opens at 15" mg, i disconnected it to see if it made any difference, it doesn't.
4) sensor plate checked, looks like what is described on the factory manuals as far as rest position.

I have some serious diagnosing to do. I'm getting ready to go check some other things but I want to open it up to the forum (everyone except Loren), any suggestions, something i might not think of right away?

I have a spare WUR which I may install just for giggles to confirm my previous diagnosis. I also plan to disconnect the temp sensor in the chain cover to make sure the fuel enrichment isn't kicking on for some unexplained reason.

Eventually I will hook up the LM-1 to provide an accurate signal to the lambda brain but I need to get it running right before I do that. I don't want to try and use the lambda brain to cover up or compensate for a problem.
Thanks in advance.

PS, I have my oil tank vented to the air filter box instead of the rubber boot between the sensor and throttle body. i did it that way because of previous turbo installation. I don't see how that could make a difference, but maybe someone else can.

aigel · aigel's Garage

Tell us how you hooked up the LM1 together with the stock O2 sensor. Did you add a bung?

George

sammyg2 · 01-29-2006, 02:02 PM

I removed the stock O2 sensor and installed the wide band sensor in it's place, after going through all the calibrations.
I intend to use the #1 analog out signal from the LM-1 to send a signal to the lambda brain after i get everything else sorted out.

911pcars · 01-29-2006, 02:18 PM

"I intend to use the #1 analog out signal from the LM-1 to send a signal to the lambda brain after i get everything else sorted out."

When you disconnect the factory O2 harness, the DME brain goes into a default mode (aka rich). After you connect the analog-out signal to the DME ECU, you should be back to more normal A/R settings. Try that.

Sherwood

sammyg2 · 02:48 PM

This is on a 1980 CIS equipped engine. No DME brain, just a primitive lambda brain.

I just found some bad news.
I taped a large hose to the air filter snorkel so that it would only pull in fresh outside air, then started checking for vacuum leaks again. So far so good, until i squirted under the air box on the passenger side between the intake runners and the fiberglass cooling shroud. A/F went from 15.2 to 14.5 and idle increased 200 rpm. The only possible explanation would be that the vapors leaked under the filter cover, or it's a vacuum leak.

Looks like it's engine pull time.
i did the same test after I went through the engine two years ago and it passed with flying colors, but somehow it has developed a leak since then. I haven't had a back fire so i'm hoping the air box is cool, I replaced all the rubber intake sleeves when I did the top end so they should be good. No leaks around the injectors, I did the sleeves and o-rings too.
Hopefully it is something simple but I'm not going to find it with the engine in the car. I'll just pull it and go through everything.

What's worse is my schedule will not permit more than a couple of hours a week to work on the car right now so it could take a month to get this sorted out. I might as well non-op the car until I have time to find and fix the leak.
Plus the fridge in the garage does not have any beer in it.

89turbocabmike · 01-29-2006, 03:21 PM

Sammy, so the factory O2 is disconnected? If so, then Sherwood is right and the lambda control would go into default mode lowering control pressure and giving your rich mixture at speed. The vacuum leak you may have found could account for the higher A/F at idle and low throttle plate openings where a vacuum leak would have more effect. Just thinking out loud here

aigel · aigel's Garage

Quote:

Originally posted by 911pcars
"I intend to use the #1 analog out signal from the LM-1 to send a signal to the lambda brain after i get everything else sorted out."

When you disconnect the factory O2 harness, the DME brain goes into a default mode (aka rich). After you connect the analog-out signal to the DME ECU, you should be back to more normal A/R settings. Try that.

Sherwood

Yes, that's exactly why I asked the question, how you hooked the two 02 sensors (stock and LM1) up. Unless you hook them BOTH up, your readings on the LM1 will be meaningless.

Glad you found something that may pinpoint to a problem, but I still recommend that you put the stock 02 sensor back, put an extra bung (in front of the cat.) for the LM1 and retest.

George

sammyg2 · 01-29-2006, 04:02 PM

Thanks for your help guys, I think I have it figured out now.

With the O2 sensor disconnected on a CIS system the frequency valve goes to a flat default setting, something like a 40% duty cycle IIRC. Just disconnect it and re-set the CO to the desired number. That's how so many guys can run with them disconnected. In fact I have done that in the past.

That is what I was trying to do, dial in the A/F with the O2 sensor disconnected and then hook it back up for fine tune control.

Unfortunately the amount of un-metered air made a much bigger difference at idle than it did when the throttle was open, hense the engine going richer under load. The more air going through the throttle body, the richer ther mixture got because the un-metered air became a smaller percentage when compared to the air being metered by the air sensor plate.
I set the idle at around stoichiometric but that was with the vacuum leak. The CIS was actually adjusted very rich, it just ran lean because of the un-metered air.
As the amount of air going through the intake increased under load, the vacuum leak became a much smaller percentage of the total air passing though, so the CIS added more fuel than was necessary.

I tried to lean it out more to get it closer to what it should be but the A/F at idle got so lean that it would stumble.
If I set the A/F ratio to 14.7 under load, the car would be too lean to idle because of the air leak.
I need to find and eliminate the un-metered air source so that the CIS can meter the fuel the same at idle as under load.
Then I can set the CO, and hook up the O2 sensor.

I'm trying to talk myself into a partial drop instead of a complete engine drop but I have large hands and I hate fighting to reach something where my hands don't fit.
It's sooo much easier with the engine on a stand.