Help...O2 and FV problem!

[997at]

1982 SC. I'm tearing my hair out trying to get my cold start to operate properly. Car starts right up, dle goes to about 1000 and then drops super low (600 ish) on start up. Runs beautifully once warm, pulls hard from idle with no hesitation at all. I have no vacuum leaks and also have a calibrated WUR from Tony. AAR works fine and timing is good.
I've set the warm mixture using the FV test port and recently passed smog in CA with flying colors. I've noticed that if I unplug the O2 sensor once the engine is warm, the duty cycle goes to 65% and the idle picks up. According to everything I've read, the duty cycle with the sensor disconnected should be at 50%. Any ideas why it might be defaulting to the wrong value? Is there away to measure the disconnected O2 wire voltage at the ECU, and if so, what should it be?
Thanks for any help guys. I love my car, but I've been chasing this for so long that I'm about ready to give up and get something without CIS.

[gazzerr]

You can disconnect the O2 sensor wire and measure the voltage to ground on it. It's supposed to be in the 0.5 to 1 volt fluctuating range after it warms up I think - others will chime in. If there's little or no voltage it's bad. How old is it? They definitely go bad after 30 years. Mine was completely dead. You're supposed to change them out every 15,000 miles.

If it runs much better unplugged and it's flipping to 65% then I think your OXS electronics and relay are probably good. Do you have fresh proper spec spark plugs and clean injectors with a proper spray pattern too?

[Walt Fricke]

First off, other than the cold idle being about 200 rpm lower than you might like, what's wrong? Does it stall at the low idle?

Have you fiddled with the idle adjustment screw - the really big fat one which is a bit hard to reach on the left side rear of the throttle assembly?

You aren't clear whether this low idle goes away when the car warms up, or stays that way and you worry about stalling, or don't like a lumpy low idle.

I'd have to brush up on what I have saved from dozens of posts here, and books like Probst on CIS, to sort through what should be what. 50% isn't default for everything, I think. There is the micro switch right there near the throttle shaft so something changes as soon as you crack the throttle. Then there is the three pole switch on the right side of the TB - it has idle, mid-throttle, and firm to wide open throttle positions. All these tell the mini brain (I hesitate to call it an ECU, though I suppose if E means electronic, and not engine, it is).

You can pick up the O2 sensor signal at the ECU, but not easily. You can identify the wire in the huge connector, open the connector up, add a wire to that pin, and run that wire back out of the connector. Or you can open up the big silver box, locate where the sensor input is, solder a wire to that, and lead it out through an existing hole in the silver box.

You can find here and by Googling enough information to figure this kind of thing out.

There is the smaller relay control box, and then there is the relay, which is what controls the frequency valve. I added a plastic box with a green LED which lights when there is power to the relay (it comes from a circuit you would not expect - the side dome lights are fused on that circuit, and if water gets into the relay the spring rusts and the relay fails. And I added two BNCs - one gets its signal from the same line which feeds the test plug you have used. The other looks directly at the signal the ECU sends to the relay. I can use a $100 hand held digital oscilloscope to see what is going on with the control of the FV.

But do you have the gauge system so you can see what your control pressures are? Those tell you a lot - cold control pressure, dictated by the cold control part of the WUR. As the heater on the WUR internal bimetal arm does its job, the cold control pressure system is no longer affecting control pressure. The right side chain box sensor is what tells the ECU when it is time to shift the FV into the hot control pressure mode. But seeing what the cold control pressure is, and what the warm control pressure is, tells you of those parameters -- which aren't changed by the FV -- are in spec as a base line.

The most comprehensive reference can be found by searching for Jim Williams CIS. Jim Williams' Garage website has a lot of stuff, including some tables showing what everything is doing when the engine is cold, warm, or hot, and I think with and without the O2 sensor hooked up.

But if the low idle persists as soon as the extra boost from the cold start valve has dissipated, my first thought is just set the idle higher. Backing out the big screw should allow more air to bypass the (closed) throttle butterfly, leaning the mixture and speeding up the idle.

Or, call or e-mail Tony. He has most of this stuff memorized.

[pmax]

O2 sensor drives a small voltage to the lambda box. With it disconnected, who knows what voltage the box is receiving on that open circuit.

[tirwin]

Originally Posted by pmax

O2 sensor drives a small voltage to the lambda box. With it disconnected, who knows what voltage the box is receiving on that open circuit.

With the O2 sensor disconnected, the lambda box goes to open loop mode which is 50% duty cycle.

[tirwin]

Originally Posted by 997at

1982 SC. I'm tearing my hair out trying to get my cold start to operate properly. Car starts right up, dle goes to about 1000 and then drops super low (600 ish) on start up. Runs beautifully once warm, pulls hard from idle with no hesitation at all. I have no vacuum leaks and also have a calibrated WUR from Tony. AAR works fine and timing is good.
I've set the warm mixture using the FV test port and recently passed smog in CA with flying colors. I've noticed that if I unplug the O2 sensor once the engine is warm, the duty cycle goes to 65% and the idle picks up. According to everything I've read, the duty cycle with the sensor disconnected should be at 50%. Any ideas why it might be defaulting to the wrong value? Is there away to measure the disconnected O2 wire voltage at the ECU, and if so, what should it be?
Thanks for any help guys. I love my car, but I've been chasing this for so long that I'm about ready to give up and get something without CIS.

The O2 sensor and FV have nothing to do with the cold start process.

It sounds like you’re lean on cold start.

First check your CCP. Then do WCP. If CCP is not in spec, you have to adjust the pin in the WUR.

Once CCP and WCP are verified, then check idle AFR. The best way for me is to use a wideband O2. You may have to adjust timing, idle speed and 3mm screw to get it dialed in. It may take some patience if it’s been messed with.

As far as 65% duty cycle, there is an old post from javadog that may help. Seems like the open loop duty cycle was different from some K-Jet lambda years.

[tirwin]

One other thing... the O2 sensor produces voltage after it is heated by exhaust gas. So you can read the voltage directly from the sensor. It wil be in .5 - 1V range like gazzerr said.

The lambda box is pretty simple. It is a series of conditional logic statement. If THIS then THAT.

O2 and FV don’t come into the picture until the engine is warm (15° C microswitch) AND O2 sensor is warm (producing voltage) AND throttle position is not WOT. I’m going from memory without enough coffee this morning so I may have left something out, but you can get the idea.

Point is once the criteria is met, then the lambda box starts looking at the O2 sensor input. It varies the duty cycle of the FV in response to lean or rich conditions trying to keep the engine at stoichiometric 14.7 AFR.

At WOT, the lambda O2 input is ignored and the duty cycle is fixed because you want to be a little richer 12-13 AFR for max power.

On cold start, the 15°C microswitch is not tripped and O2 is not warm so lambda is ignored. The only things involved in the cold start process are the WUR and AAR. If they are functioning correctly, they get out of the picture after 2-3 minutes of warm up time.

[997at]

Originally Posted by gazzerr

You can disconnect the O2 sensor wire and measure the voltage to ground on it. It's supposed to be in the 0.5 to 1 volt fluctuating range after it warms up I think - others will chime in. If there's little or no voltage it's bad. How old is it? They definitely go bad after 30 years. Mine was completely dead. You're supposed to change them out every 15,000 miles.

If it runs much better unplugged and it's flipping to 65% then I think your OXS electronics and relay are probably good. Do you have fresh proper spec spark plugs and clean injectors with a proper spray pattern too?

Thanks gazzer. I think you're right in that the relay etc, are working okay. I have a brand new O2 sensor in as well and given that my smog passed so easily when I tuned the mixture using the FV all must be okay. I get a very good dither and can see it compensate with the slightest turn of the screw. I have brand new injectors and while plugs are not new they're in good shape.

[997at]

Originally Posted by Walt Fricke

First off, other than the cold idle being about 200 rpm lower than you might like, what's wrong? Does it stall at the low idle?

Have you fiddled with the idle adjustment screw - the really big fat one which is a bit hard to reach on the left side rear of the throttle assembly?

You aren't clear whether this low idle goes away when the car warms up, or stays that way and you worry about stalling, or don't like a lumpy low idle.

I'd have to brush up on what I have saved from dozens of posts here, and books like Probst on CIS, to sort through what should be what. 50% isn't default for everything, I think. There is the micro switch right there near the throttle shaft so something changes as soon as you crack the throttle. Then there is the three pole switch on the right side of the TB - it has idle, mid-throttle, and firm to wide open throttle positions. All these tell the mini brain (I hesitate to call it an ECU, though I suppose if E means electronic, and not engine, it is).

You can pick up the O2 sensor signal at the ECU, but not easily. You can identify the wire in the huge connector, open the connector up, add a wire to that pin, and run that wire back out of the connector. Or you can open up the big silver box, locate where the sensor input is, solder a wire to that, and lead it out through an existing hole in the silver box.

You can find here and by Googling enough information to figure this kind of thing out.

There is the smaller relay control box, and then there is the relay, which is what controls the frequency valve. I added a plastic box with a green LED which lights when there is power to the relay (it comes from a circuit you would not expect - the side dome lights are fused on that circuit, and if water gets into the relay the spring rusts and the relay fails. And I added two BNCs - one gets its signal from the same line which feeds the test plug you have used. The other looks directly at the signal the ECU sends to the relay. I can use a $100 hand held digital oscilloscope to see what is going on with the control of the FV.

But do you have the gauge system so you can see what your control pressures are? Those tell you a lot - cold control pressure, dictated by the cold control part of the WUR. As the heater on the WUR internal bimetal arm does its job, the cold control pressure system is no longer affecting control pressure. The right side chain box sensor is what tells the ECU when it is time to shift the FV into the hot control pressure mode. But seeing what the cold control pressure is, and what the warm control pressure is, tells you of those parameters -- which aren't changed by the FV -- are in spec as a base line.

The most comprehensive reference can be found by searching for Jim Williams CIS. Jim Williams' Garage website has a lot of stuff, including some tables showing what everything is doing when the engine is cold, warm, or hot, and I think with and without the O2 sensor hooked up.

But if the low idle persists as soon as the extra boost from the cold start valve has dissipated, my first thought is just set the idle higher. Backing out the big screw should allow more air to bypass the (closed) throttle butterfly, leaning the mixture and speeding up the idle.

Or, call or e-mail Tony. He has most of this stuff memorized.

Hi Walt. Thanks for the input. While the idle is lowish for about 5 min, it doesn't hesitate, stall, or run lumpy. As the car warms up, the idle slowly increases up to 950 where it sits at operating temp. I do notice another interesting trend with my car. For every 10 deg C above operating temp of 190 (in hot weather idle conditions) the rpm will creep up 50 or so up to a max of about 1100.
My WUR control pressures are spot on as measured with the gauge set. It was Jim's CIS primer that told me the default with the O2 sensor disconnected should be at 50% dwell once the engine is warmed up. I'm beginning to wonder if my throttle body valve setting is a tad too far open and triggering the 2 degree enrichment switch. I might close it a bit more and see what happens.

[997at]

Originally Posted by tirwin

The O2 sensor and FV have nothing to do with the cold start process.

It sounds like you’re lean on cold start.

First check your CCP. Then do WCP. If CCP is not in spec, you have to adjust the pin in the WUR.

Once CCP and WCP are verified, then check idle AFR. The best way for me is to use a wideband O2. You may have to adjust timing, idle speed and 3mm screw to get it dialed in. It may take some patience if it’s been messed with.

As far as 65% duty cycle, there is an old post from javadog that may help. Seems like the open loop duty cycle was different from some K-Jet lambda years.

Thanks Tim. I'm also suspecting that I'm lean on cold start. WUR pressures are in spec but I think. that the WUR is transitioning to WCP much quicker than the AAR is closing creating a lean condition until the O2 sensor heats up goes into closed loop.

[997at]

Originally Posted by tirwin

One other thing... the O2 sensor produces voltage after it is heated by exhaust gas. So you can read the voltage directly from the sensor. It wil be in .5 - 1V range like gazzerr said.

The lambda box is pretty simple. It is a series of conditional logic statement. If THIS then THAT.

O2 and FV don’t come into the picture until the engine is warm (15° C microswitch) AND O2 sensor is warm (producing voltage) AND throttle position is not WOT. I’m going from memory without enough coffee this morning so I may have left something out, but you can get the idea.

Point is once the criteria is met, then the lambda box starts looking at the O2 sensor input. It varies the duty cycle of the FV in response to lean or rich conditions trying to keep the engine at stoichiometric 14.7 AFR.

At WOT, the lambda O2 input is ignored and the duty cycle is fixed because you want to be a little richer 12-13 AFR for max power.

On cold start, the 15°C microswitch is not tripped and O2 is not warm so lambda is ignored. The only things involved in the cold start process are the WUR and AAR. If they are functioning correctly, they get out of the picture after 2-3 minutes of warm up time.

Hi Tim. I did check the O2 sensor signal directly and it is functioning well (new sensor). The only thing in the whole picture that is still a bit confusing for me is that once warm, if I disconnect the O2 sensor I get a 65%duty cycle instead of the correct 50%. I'm wondering if there's something up with the micro switches at the throttle body making the system think that it needs to be in rich (65%) mode.

[tirwin]

Originally Posted by 997at

Thanks Tim. I'm also suspecting that I'm lean on cold start. WUR pressures are in spec but I think. that the WUR is transitioning to WCP much quicker than the AAR is closing creating a lean condition until the O2 sensor heats up goes into closed loop.

Ok, to prove your theory, here what you need to do.

Jump pins 87a and 30 on the fuel pump relay. Start with WUR electrical plug disconnected. Check CCP. Plug in electrical to WUR. Measure time to max WCP.

Remove the AAR from the car. Put it in a plastic bag and put it in the refrigerator overnight. The next day put it in your workbench so you can see through the opening. Connect it to 12v power. Measure how long it takes to completely close. If it doesn’t completely close you have a problem.

Compare the two times. If they are drastically different then you’re right. If they are close to the same you’re barking up the wrong tree.

The cold start process shouldn’t be more than 3-4 minutes.

PS - the plastic bag is an anti-divorce device.

[tirwin]

Originally Posted by 997at

Hi Tim. I did check the O2 sensor signal directly and it is functioning well (new sensor). The only thing in the whole picture that is still a bit confusing for me is that once warm, if I disconnect the O2 sensor I get a 65%duty cycle instead of the correct 50%. I'm wondering if there's something up with the micro switches at the throttle body making the system think that it needs to be in rich (65%) mode.

How are you measuring the duty cycle? It is important to be sure you know the 65% is OPEN or CLOSED. One is going to be causing a richer mixture and the other is going to be causing a leaner mixture. Easy to overlook. At the risk of being pedantic, it’s sometimes good to point out the obvious (or not so obvious).

Cold start should be richer mixture. After 3-4 minutes, the engine should be warm and idling. With O2 sensor connected, lambda box should be trying to keep mixture to 14.7.

If AAR wasn’t fully closing, then you’d have less fuel after cold start process is done and more air, which would cause a lean condition. Lambda would try to correct but it can only do so much.

[Walt Fricke]

1) I don't know at what voltage the ECU system would recognize a signal from the O2 sensor or would perceive it as an open. An open circuit like this isn't going to have more than a couple of microvolts on it, and that has to be below the threshold. If there was a higher stray voltage somehow imposed by frayed wires in contact with some other electrical circuit, the closed loop system wouldn't work.

2) As to the switches on the throttle body, the micro switch for fully closed/hair open butterfly is easy to check, which I am sure at997 has done. The other switch, the three position one, is a very simple mechanical device. I took mine off once and opened it up. Very unlikely to go bad.

3) A lean idle usually would be faster than it would be at stoich, other factors being equal? Wouldn't matter if it came from the AAR or the idle bleed air screw, would it?

4) The spark advance can affect idle. The centrifugal advance must be working or the engine would not be running on the highway as well as it does. Don't know if a small adjustment would take care of the little but not debilitating glitch, but checking distributor timing is often the first place to look - though after points went away, maybe not so much any more.

[gazzerr]

tirwin's point about the duty cycle measurement being open or closed is important. This had me chasing my tail for a while. Devices trigger differently on the square waveform being output from the test pin. You might find you are getting the right numbers when you take the number you are getting from 100%.

Note my car is also always 200 or so rpm lower than it should be on the warm up cycle. My base timing seems right but I suspect my distributor needs a refresh. The vacuum control after all these years has got to be sub par. I've replaced or refreshed everything else in my CIS. AAR's don't seem to work particularly well in summer weather on a first cold start regardless.

[997at]

Originally Posted by tirwin

How are you measuring the duty cycle? It is important to be sure you know the 65% is OPEN or CLOSED. One is going to be causing a richer mixture and the other is going to be causing a leaner mixture. Easy to overlook. At the risk of being pedantic, it’s sometimes good to point out the obvious (or not so obvious).

Cold start should be richer mixture. After 3-4 minutes, the engine should be warm and idling. With O2 sensor connected, lambda box should be trying to keep mixture to 14.7.

If AAR wasn’t fully closing, then you’d have less fuel after cold start process is done and more air, which would cause a lean condition. Lambda would try to correct but it can only do so much.

Hi Tim. I'm measuring the duty cycle with a digital dwell meter. I'm only getting 65% readings when the loop is open, but in two different states. One is during cold start before the 15 deg temp switch and the O2 sensor have heated up, and the other is when I disconnect the O2 sensor after the car is warmed up.
It's the 65% reading with the O2 sensor disconnected at warm idle that seems weird to me.

[wreckah]

Hi 997at, has your car been doing this all the time? or is it a recent evolution?

there has been an issue with cold start flanges: they provide the path to the AAR air passthrough at cold start. Some of these flanges have been delivered with a closed off port, even though the part nr. was correct. This of course makes the AAR not work at all, even though it opens and closes fine.

this flange sits under the CSV, at the back of the throttle body. Very annoying and difficult to get to...but maybe worth the check. (after you've done the basics: testing fuel pressures to start)

[wreckah]

at the bottom of this page, you can check all the duty cycle states for different engine temp and throttle conditions...
911 CIS Primer - CIS Lambda

[AndrewCologne]

I've noticed that if I unplug the O2 sensor once the engine is warm, the duty cycle goes to 65% and the idle picks up. According to everything I've read, the duty cycle with the sensor disconnected should be at 50%.

Did not read all answers before but regarding the issue you mentioned above:

If the O2 Sensor is disconnected and the engine is warm, means above 15 °C the dutycycle of the frequency valve at idle must be static at 50%! Not 65%.
Here I guess the 15°C Switch on the right chain housing could be flawed, means keeps always leading to ground which makes the ECU thinking its still below 15°C.
This also affects the CO-Measurement as it screwes it up as initial CO at idle MUST be measured and set at 50% dutycycle!
Without that the initial CO-setting results too lean and your cold start is messed up.

So...

1. Put your 15°C switch in a freezer and then check if it opens the circuit correctly at room temperature – use a DMM for measuring the open circuid

2. If the switch above is ok and with a warm engine the duty cycle then is at 50% @ idle, do a new initial CO-Measurement and -adjusting.

3. Check your fuel pressures. The WUR of the US SC comes with a slightly more mild curve as one part of the enrichment at cold start is also provided by a duty cycle of 65%.

Car starts right up, dle goes to about 1000 and then drops super low (600 ish) on start up.

This seems to me a problem related to a wrong adjusted bypass screw.
Because if you open the bypass screw and rise the idle by 300 rpm, then this should/could proportionally end in a perfect 1300 rpm cold idle and ... 900 rpm warm idle.

[Dave Kost]

If it were my car, I would your CO to factory specs- .4 to .8% at 900 RPM. In fact, if you are running Ethanol blended gas I would set it at 1 - 2% at 900 RPM. This is set at operating temperatures and before the cat, O2 sensor disconnected.

%CO A/F
.1 14.71
.5 14.27
1.0 14.10
1.5 13.93
2.0 13.76
2.5 13.55
3.0 13.37
3.5 13.19
4.0 12.99
5.0 12.63
6.0 12.24

Make sure your ignition system is working 100%.


Dave