Lamda not working

[Michael]

So after noticing a hard cold start(sometimes it would start right up cold, other times it would take about 5 or 6 seconds of cranking before it started), checked CSV operation. I went to look for any obvious vacuum leaks, none that I could see. I then attached my dwell(engine warm enough for closed loop) to see if it was rich and no activity(right on 50%) on the o2 sensor at all. I replaced the o2 sensor and still dead dwell needle although it is rich, barely fluctuating a little(probably because of the cam) and does change with throttle, but no real o2 signal that I can see. The car does start, idles, runs fine, but the o2 signal is just not there. However there has been once or twice in the past few days of me finding a few minutes to work on it that it does work and mixture is set perfect. I am thinking maybe my lambda computer is bad. The WOT switch works, the o2 sensor is new, idle circuit is working. FV is fairly new and seems like its working. Right now I am just waiting on the engine to cool and maybe dive deeper to find a vacuum leak. As far as I can tell the thing is stuck in open loop and has only went into closed loop like twice during my messing with it. This is a stage II engine (cams, headers, etc.)

Is the temp switch on the water pump housing the only thing that tells the lamda system to work?

[NckT]

The switch on the water pump is the thermo-time switch and this only operates the cold start valve etc.

I'll assume your frequency valve is buzzing and you've connected your dwell meter ok and set it to 4 cylinder seeing (not 6 cylinder) etc.

There is a switch on the V pipe under the Ubley manifold and this has the lambda switch, ie it is closed when the water temp is approx less than 80c and open when the engine is warm etc. This closed switch is an input to the fuel computer to operate a fixed dwell at 54 deg/about 60%. This is in parallel to the full load throttle switch so if you're getting this dwell value, try disconnecting the full loaf throttle switch as it's not unknown for them to stick in the closed position.

If your'e only getting a fixed value of 45deg /50% dwell then it could be likely that the oxygen sensor signal is not switching/ voltage input that would be enough for the ecu to operate. This would be the case if you do get a varying dwell if you run the engine at 3000-3500 rpm. In this condition it would clearly indicate a bad oxygen sensor.

If the above is not true, then connect an oscilloscope to the oxygen sensor signal wrote to the fuel ecu and if it gives a sinusoidal voltage change between 0.2v to 0.8v at 1hz idle but no varying dwell then I'd check the connection to the evu input or try another fuel computer

Good luck, I hope it's something simple.

[Bitsyncmaster]

I'm pretty sure what tells the lambda to go closed loop is voltage from the O2. There is no voltage when the O2 sensor is cold.

You can try unplugging the lambda connector and plug it in again just in case there is a marginal connection.

I have a few lambda ECUs that I have tested as working. If you want to borrow one to test.

[opethmike]

Have you considered giving the car a hug?

[David T]

To get into "Closed Loop" the temp senor must cycle and the O2 sensor must be putting out enough voltage. Of course that is assuming everything else is working. If you are getting a reading at all we can assume the Lambda ECU is getting power and if you hear the freq valve buzzing it is connected and working. There are tests you can do to the ECU to verify operation. See the notes in the Workshop Manual. The O2 sensor puts out a very low voltage so if you ANY bad connections between it and the ECU, or you do not have a GOOD ground the Lambda ECU will not get the signal from the O2 sensor.

[Bitsyncmaster]

Your blue bulkhead connector pins 22 and 23 have to go to engine block. Those are two grounds going to lambda ECU

22 B Engine block, from Lambda ECU pin 5
23 B Engine block, from Lambda ECU pin 16

[Jonathan]

Something free and easy to also check is the O2 sensor (reference) ground. It is (usually) attached to the block on the passenger/forward side, sort of near where the frequency valve is. The short length of wire that is screwed into the block is RED and then has a black quick connector on it that connects to the longer, regular black ground wire that eventually meets up and gives the O2 sensor it's ground reference. Check to see if that connection is good.

Shown in photo (I relocated mine slightly higher as I have a small chunk of broken off screw in the hole it is supposed to go in).

53726

[Michael]

Thanks for all the advice, I definitely have plenty to check. FWIW I do think it is going into closed loop now. I started it fairly cold and hooked my meter up and watched until it went closed loop. As soon as it did the engine wants to die, so I think any signal I should see gets lost with the engine starving:


https://www.youtube.com/watch?v=62M_OE7j4WQ

I am leaning heavily towards a vacuum leak after seeing that, but then about a minute later (see next post, only one vid allowed per post these days I guess)

[Michael]

https://www.youtube.com/watch?v=aHRPAFl-kmo

About the :25 mark you can see how the engine shudders at the starve point. It has always had a bit of a lope to it because of the cams, but this was a bit more pronounced. I think I need a few cans of carb cleaner and start looking under the intake for a leak. I'm just not seeing enough action from the O2 (which is brand new BTW) and I did take it out and drive it to burn off any impurities before this.

[David T]

I have found that the cars with the "hotter" cams need a little bit higher idle speed to be happy. Try bumping it up by opening the brass air screws. You may not get the dwell to read 35-45 for a good idle either. It may take a bit richer mixture to get it to run smooth at idle too.

[DMCVegas]

Workshop Manual Section D:04:14 will lay everything out and give you the exact specs you should be seeing.

LAMBDA has 3 modes of operation for target duty cycles it applies to the Frequency Valve:


The first is Open Loop for when the O2 sensor is offline because it is not yet warmed up. 40-50° Steady Reading
The Second is for when Enrichment is triggered. This occurs when the LG, or Light Green wire is grounded. Both the WOT switch and the Thermal Switch in the Y-pipe are spliced into this same wire and can trigger this by WOT or coolant temp of 15°C/59°F or less. 50-60° Steady Reading
The third and final is normal operation. O2 Sensor is reporting back, and neither the WOT or Thermal Switch is triggered. 35-45° Pulsating


If you want to test the LAMBDA ECU, you can do it easily once it's warmed up. Ground the O2 lead and it'll hit 87° minimum. Apply 1.5V and it'll hit 20° maximum. That'll let you know if it's bad or not. Just in case too, disconnect the WOT Switch, and then check the LG wire to see if it's grounded. That'll tell you if the Thermal Switch is bad and screwing with things by constantly running too rich when not needed. Of corse, check all grounds as well. For the hell of it, also gently snug up the 3 brass screws to ensure there are no leaks. Probably not related, but it's good to ensure you're not running unbalanced cylinder banks.



I have found that the cars with the "hotter" cams need a little bit higher idle speed to be happy. Try bumping it up by opening the brass air screws.

How does that work exactly? Not trolling here at all, but genuinely curious. Any amount of additional metered airflow you supply would increase idle speeds, yes. But once the RPMs increase, the Idle Speed Circuit is going to see that and adjust the Idle Speed Motor accordingly to drop the speeds back down to defeat said modification to the RPMs.

[Bitsyncmaster]

The problem with our stock O2 sensor is the fact it's not heated. That's mostly why all the newer cars have heated O2 sensors now. They found the lower RPM at idle that sometimes the sensor is just to cool to provide a good output. Maybe the higher lift duration cams run cooler (don't know). Maybe some of our idle hunting problems are caused by that.

You can run a higher closed loop idle RPM with my idle ECU or a modified stock ECU. Any other adjustments will make the idle RPM go open loop.

[NckT]

I have found that the cars with the "hotter" cams need a little bit higher idle speed to be happy. Try bumping it up by opening the brass air screws. You may not get the dwell to read 35-45 for a good idle either. It may take a bit richer mixture to get it to run smooth at idle too.

How many have you done exactly? I've set up 3 with the stage 1 dmch and 1 with just the performance cams and have not had to increase the rpms from stock all with the brass screw modification I outlined in another post.

[David T]

How many have you done exactly? I've set up 3 with the stage 1 dmch and 1 with just the performance cams and have not had to increase the rpms from stock all with the brass screw modification I outlined in another post.

Exactly 2. Opening the brass screws bypasses control of the idle motor. The more air you allow around it (through the air screws) the less control it has and the less ability it has to reduce the idle speed. Cars went to the heated O2 sensors mostly to reduce the time of the warm-up cycle to control the emissions quicker. The other reason is to prolong the life of the sensor so it doesn't get contaminated from raw fuel as easily. What base timing were the cars running when you got the idle set?

[NckT]

Exactly 2. Opening the brass screws bypasses control of the idle motor. The more air you allow around it (through the air screws) the less control it has and the less ability it has to reduce the idle speed. .........

No it doesn't, as it is dependent on how far you open the brass screws. I explained this in post no.19 in
http://dmctalk.org/showthread.php?15633-Attempting-to-destroy-my-idle-lumpiness/page2

The advance for all were set to 14-14.5 deg btdc vacuum isolated, apart from one car that I set up at the last euro fest (in a car park) where the timing was 13btdc.

I don't want this to get off topic from the original post.

[MrChocky]

Reopening old threads seems to be current de jure, so here we go. Michael's descriptions and at least the diagnostic steps here lead me to choose this particular thread.

What I'm actually chasing is high idle (about 975) but the car based upon different attempts to measure it, is not going into closed loop (or at least, very rarely).

Last year, I replaced all the vacuum lines, O2 sensor, and have also put additional clamps on suspect vacuum leak spots, as well as replacing injector
seals, etc. This spring, I pulled my manifold for painting and expressly to check the vacuum routing in the valley - all fine.

I can't hear it, but I'm assured my FV is working. I've checked physical and electrical operation of idle and WOT switches, and recently also checked
the butterfly valves are closing fully. I do have DM's idle ECU (set to 775) which has the LED read out for closed loop mode which is obviously not coming on, and
for completeness, I've hooked up the "A/C bump" (as an aside, this has largely eliminated my stalls when coming to a stop), but obviously for testing no A/C is running.

The car has no cold start, hot start or other issues, and general seems to run exceptionally well given everything. What I will do shortly is pull the vacuum canister;
I know its hoses aren't always in good shape, and easy to check.

In any case, suggestions on things to things to measure, check (or hell, triple check). Nothing immediately jumps out at me from the thread, apart from perhaps
measuring the O2 sensor reading.

Thanks!

[Bitsyncmaster]

I do have DM's idle ECU (set to 775) which has the LED read out for closed loop mode which is obviously not coming on, and
for completeness, I've hooked up the "A/C bump" (as an aside, this has largely eliminated my stalls when coming to a stop), but obviously for testing no A/C is running.

Thanks!

The LED turning on indicates the idle motor is fully closed so it can't reduce idle any more. With the LED off it indicates the idle motor still has range to control idle speed. If it is still off and your still indicating high RPM then it may be your ignition is missing making the ECU think it's running a lower RPM. Verify the idle motor is buzzing when engine is at idle (you can feel the idle motor and feel vibration).

[Helirich]

Subscribed for future. Good info here!

[Ron]

Reopening old threads seems to be current de jure, so here we go. Michael's descriptions and at least the diagnostic steps here lead me to choose this particular thread.

What I'm actually chasing is high idle (about 975) but the car based upon different attempts to measure it, is not going into closed loop (or at least, very rarely).

Last year, I replaced all the vacuum lines, O2 sensor, and have also put additional clamps on suspect vacuum leak spots, as well as replacing injector
seals, etc. This spring, I pulled my manifold for painting and expressly to check the vacuum routing in the valley - all fine.

I can't hear it, but I'm assured my FV is working. I've checked physical and electrical operation of idle and WOT switches, and recently also checked
the butterfly valves are closing fully. I do have DM's idle ECU (set to 775) which has the LED read out for closed loop mode which is obviously not coming on, and
for completeness, I've hooked up the "A/C bump" (as an aside, this has largely eliminated my stalls when coming to a stop), but obviously for testing no A/C is running.

The car has no cold start, hot start or other issues, and general seems to run exceptionally well given everything. What I will do shortly is pull the vacuum canister;
I know its hoses aren't always in good shape, and easy to check.

In any case, suggestions on things to things to measure, check (or hell, triple check). Nothing immediately jumps out at me from the thread, apart from perhaps
measuring the O2 sensor reading.

Thanks!Didn't re-read it all, so stabs in the dark:
Ensure the three brass screws are shut.

Perform all of the procedures (record, and post if off) at D:04:15/5.

Record where the CO screw is currently at. Then nudge it in steps, as small as you can one way, until it about dies or starts swinging. If no luck, try the other direction.