Reinstated lambda system yesterday - strange behaviour, please help!

[Horsebox]

So the car has been open loop lambda since I bought it, and it's been fine. I wanted to reinstate the system so I bought a new lambda sensor as I knew the lambda ECU was operating (FV buzzes and duty cycle pegged at 50%). The car has always started on the first turn every time, hot or cold.

YESTERDAY
I take the car and the new DMC lambda probe to a mechanic to be fitted. He phones back a few hours later and says it's now running closed loop, been out with it and it drives fine, and the before-and-after CO measurements were 4% (open loop) which reduced to 1% (closed loop). Happy days. So it looks like its been running rich all the time. I drive home via a petrol station, refill, and hot starting is difficult, which I'd never seen before. Once it's running it's fine. I get it home and measure the duty cycle, and it's fluctuating around 25%, which I reckon is a lean correction to an otherwise slightly rich mixture. Hot start it again and needs a bit of cranking and pumping the accelerator to get it going, but it does go, and I put it back in the garage for the night to cool off.

TODAY
I want to see what happens with the reinstated lambda system from a cold start, as I haven't seen it yet. The car starts on the button with the FV initially 60% open loop, then drops to 50% open loop. Idlespeed is holding 775rpm. After a couple of mins it goes closed loop and the FV immediately climbs all the way from 50% to 96% and stays there. The rpm increases with the extra fuel going in and the idle motor corrects it. The car then hunts at idle with increasing oscillations until I shut it off five mins later. During the whole hunting episode the FV is pegged at 96% duty cycle, so the hunting is all due to the idle motor swinging back and forth.

I wanted to drive it and get it properly up to temperature to repeat the conditions that I picked the car up yesterday (it had just been out for a run). I don't want the hunting so I disable idlespeed. It hot starts on the button and I get back after a decent run (all fine) and measure FV duty cycle which is still pegged at 96%. I poked the idlespeed microswitch to activate the system and it idles correctly with a little bit of hunting.

I leave it for 15 mins then decide to put it back in the garage as it's going to rain soon. I reinstate the idlespeed. Car starts on the first turn. There is no hunting as the FV is still pegged at 50% open loop (not yet up to temperature). I faff about a bit but two mins later the car is in the garage. The car is an auto and it's downhill to the garage so idlespeed is in control and I'm riding the brakes. There is no hunting. I put the gearbox in to park, and with no gas pedal it hunts like crazy until I turn it off.

I'm completely confused. It's like someone has swapped the lambda sensor over night. I'd be very grateful for any help, and sorry for the length of the post!

Cheers,

Mark

[sdg3205]

Sounds like your mixture is still not set properly. A lot of mechanics simply don't set this properly and find the proper range of operation/fluctuation of the dwell cycle. The window is tiny! Took me a long time to figure it out myself.

[Horsebox]

The mixture adjustment hasn't been touched. It obviously isn't a million miles away as I've been running open loop for about a year with no hint of a running problem. (Lucky me!).

What is bonkers is how it goes from thinking its a bit rich yesterday, therefore 25% FV duty cycle in response to going closed loop for the first time in many years, to thinking it's very lean today, hence the 96% FV duty today (I guess it's run out of range).

Nothing changed overnight, apart from filling the car up with gas about 1 min drive to my house on the way back from having the sensor fitted.

Maybe that's it....the tank going empty to full....it's always run out of power as the tank has got emptied. I pushed it too far on the tank before last and the fuel light came on (yay it works!), but by the time I got to a gas station I was flat out just to hold 30mph.

This is driving me nuts as I can't see any sense to anything.

Mark

[sdg3205]

I believe if your average dwell is 25, that's leaner. Did the mixture get adjusted to run open loop? My understanding was the mixture had to be richer to run open loop, which would mean it would need to be readjusted to run closed loop again. You're dwell should be fluctuating around 40-45 IIRC.

[Bitsyncmaster]

I agree the mixture adjustment would be the first thing to do when your reading 96%. I have no clue why it read 25% the first day. Maybe you had the meter set to dwell with the wrong cylinders selected.

By the way, even if your duty cycle is 20% to 80% your still running the correct 14.7 AFR. You just try to center it about 35-40% so you hold closed loop with altitude and other variables.

[NightFlyer]

I agree the mixture adjustment would be the first thing to do when your reading 96%. I have no clue why it read 25% the first day. Maybe you had the meter set to dwell with the wrong cylinders selected.

I've got a couple questions for you, if you don't mind.

Is there any advantage to adjusting the mixture based off the frequency valve's duty cycle as opposed to using the O2 sensor's dwell readings? Is one way more accurate than the other, or will they essentially yield the same results?

If you're adjusting via the O2 sensor's dwell readings, should you have your meter set to 4 cylinder or 6 cylinder, as I use a digital meter that has position settings for 4, 5, 6, and 8 cylinders.

Thanks Dave - you da man :thumbup:

[Bitsyncmaster]

I've got a couple questions for you, if you don't mind.

Is there any advantage to adjusting the mixture based off the frequency valve's duty cycle as opposed to using the O2 sensor's dwell readings? Is one way more accurate than the other, or will they essentially yield the same results?

If you're adjusting via the O2 sensor's dwell readings, should you have your meter set to 4 cylinder or 6 cylinder, as I use a digital meter that has position settings for 4, 5, 6, and 8 cylinders.

Thanks Dave - you da man :thumbup:

When using dwell you have to set it to 8 cylinder (I think it was 8, maybe 4 but not 6).

That is the advantage to using duty cycle. There is no setting for cylinders. Accuracy is the same either way but duty cycle number is the book dwell number times 1.10 Since dwell is maxed at 90 deg. and duty is maxed at 100%

[NightFlyer]

When using dwell you have to set it to 8 cylinder (I think it was 8, maybe 4 but not 6).

That is the advantage to using duty cycle. There is no setting for cylinders. Accuracy is the same either way but duty cycle number is the book dwell number times 1.10 Since dwell is maxed at 90 deg. and duty is maxed at 100%

That's what I thought - I just wanted to confirm it.

Thanks Dave :thumbup:

[NightFlyer]

Dave S got my mixture dialed in pretty good at DCS, but I have something else going on with idle system that I'll attempt to trouble shoot and remedy when I get some time.

[Horsebox]

I agree the mixture adjustment would be the first thing to do when your reading 96%. I have no clue why it read 25% the first day. Maybe you had the meter set to dwell with the wrong cylinders selected.

By the way, even if your duty cycle is 20% to 80% your still running the correct 14.7 AFR. You just try to center it about 35-40% so you hold closed loop with altitude and other variables.

The meter has always been set to duty cycle percentage. I completely take your point though.

Also agree that it's running stochiometric if the FV isn't at either extremes of movement. At the end of the day it's a trim valve, so makes sense to set the mixture so it has equal movement either side of neutral I.e 50% duty cycle or 45 degrees dwell if you prefer basing the denominator on 90 as opposed to 100. For me duty cycle (I.e. Based on fraction of 100) is more intuitive than basing on 90 (dwell).

As you say, no idea why yesterday it was 25% yet today it is off scale at 96%, despite the only change being a fresh tank of fuel and a sunrise.

Mark

[Bitsyncmaster]

The meter has always been set to duty cycle percentage. I completely take your point though.

Also agree that it's running stochiometric if the FV isn't at either extremes of movement. At the end of the day it's a trim valve, so makes sense to set the mixture so it has equal movement either side of neutral I.e 50% duty cycle or 45 degrees dwell if you prefer basing the denominator on 90 as opposed to 100. For me duty cycle (I.e. Based on fraction of 100) is more intuitive than basing on 90 (dwell).

As you say, no idea why yesterday it was 25% yet today it is off scale at 96%, despite the only change being a fresh tank of fuel and a sunrise.

Mark

Duty cycle is what you are really measuring (time on compared to time off). Dwell was done because most auto shops had a dwell meter that was used to set the points in the distributor. 60% for WOT and 50% open loop during warm up is easy to remember. I think you should stick with setting about 40% so when you have a cold start or WOT your mixture goes rich. It also keeps the WUR operation working as designed.

I seem to remember the lambda ECU had a max duty of 95% and a min duty of 12% or thereabouts.

[Horsebox]

I don't understand dwell but I don't think it matters with Deloreans, since duty cycle % for the FV seems more relevant and intuitive, and I can measure that easily enough.

I need to get my head around why a nights sleep for the car and a full tank of fuel has seemingly sent it from being slightly rich (resulting in FV=25%) to very lean (resulting in FV=96% - presumably off scale).

Thanks for the help guys, it's frustrating knowing that the answer will be simple, but there's just a few too many variables to pin it down at the moment.

Mark

[David T]

For the car to run acceptably in open loop without the frequency valve the mixture needs to be richened. The mixture adjustment is the "base" fuel and the Lambda system adds just enough to that so you are closer to "stoch" or ideal A/F. Without the Lambda adding in additional fuel the base mixture is tweaked a bit richer. Once you get the Lambda system running again you are now putting it too much fuel and the mixture screw must now be adjusted a little bit leaner. The mixture screw adjustment is VERY sensitive, a tiny bit has a large effect. You need to adjust it so you get into the middle of the window so the Lambda can adjust both up and down and that is a hard spot to get because that mixture screw is so sensitive and the motor is hunting a little because of the idle system. The motor needs to be fully warmed up too before trying to adjust. If the motor is not in good "tune", that is the plugs are old, etc, it is not possible to properly adjust the Lambda system.

[Ron]

Mark,

Anytime you change ANYTHING in the fuel system (lambda, or even basic tune up parts, etc) the dwell needs to be checked out, LAST.
Flip your meter to dwell and check out all the specs at D:04:05 in the Workshop Manual.
Make sure the engine is warm, you use 4 cylinder setting (or double 8 cyl readings), and take each reading after letting it run a few seconds with the the CO adjusting hole plugged! I bet you'll find the problem (if your injectors etc are all good).

[Horsebox]

I should have been clearer. The FV has been running at 50% since I got the car, and it's been running fine. All I've done is replace the oxygen sensor to send it closed loop.

With the new oxygen sensor fitted I immediately measured 25% duty on the FV, confirming what I suspected was a rich base fuel mixture.

24 hours later, after filling the car up with fuel, and following a cold start, the FV duty went to 96% and stayed there.

I don't get how a full tank of fuel (the only change between the two readings) could send a rich base mixture to a lean one. This is the puzzle.

As an aside, I checked all the vacuum hoses. All were in the correct place except the two on the front of the pipe of agony, which were reversed. Would this make much of a difference do we know?

Cheers guys.

[David T]

The meter has always been set to duty cycle percentage. I completely take your point though.

Also agree that it's running stochiometric if the FV isn't at either extremes of movement. At the end of the day it's a trim valve, so makes sense to set the mixture so it has equal movement either side of neutral I.e 50% duty cycle or 45 degrees dwell if you prefer basing the denominator on 90 as opposed to 100. For me duty cycle (I.e. Based on fraction of 100) is more intuitive than basing on 90 (dwell).

As you say, no idea why yesterday it was 25% yet today it is off scale at 96%, despite the only change being a fresh tank of fuel and a sunrise.

Mark

I don't care where you set your meter to. The way you know you have it set correctly is:

O2 sensor disconnected-40-50 steady
O2 sensor lead grounded-87 min steady
O2 sensor lead 1.5 volts DC applied-20 max steady
cold motor or WOT-50-60 steady

This also tells you the Lambda ECU is alive and functioning properly.

When you can get these readings you know the meter is connected properly and on the right scales. Now warm up the motor and hook up the O2 sensor. Adjust the mixture screw to get 35-45 aiming to get 40. You have to remove the wrench and plug the hole and wait till things settle down. A tiny movement of the mixture screw makes a BIG change. Refer to D:04:15. I find the ground connection in the diagnostic plug is not always very good. You need a good ground connection for the dwell meter to get good readings. Use fresh gasoline and this adjustment is done after ALL other work. Do not attempt to set this if you have old, bad, worn ignition parts or dirty/bad patterns on the fuel injectors. Or vacuum leaks all over. Don't be tempted to try to "adjust" your troubles away. It never happens. If the motor was running very rich for a while the O2 sensor is contaminated and will not give accurate readings anymore and should be replaced before trying to set the mixture. If the mixture was never adjusted it is probably very close to where it should be and should NOT be adjusted.

[Horsebox]

I just wanted to tie this thread up rather than leave it without a conclusion - I know it's old :-)

As of two weeks ago it's now running closed-loop properly.

I was going about everything back to front. It turned out the CPR was way out of spec, the fuel pump wasn't getting a proper voltage due to bad connectors, there were vacuum leaks through the evap system and the injectors sprayed lines of fuel.

After fixing all of that, replacing the ignition consumables and yet another oxygen sensor all is well now.

Thanks for your help!

Mark