Mixture Adjustment Confirmation

1982911SCTarga · 06:17 PM

I'm looking for some advice from the old hands who have been through this exercise to help me confirm some things. I'm trying to set my mixture on my cab's rebuilt engine by using a dwell meter and tapping the green/white wire in the lambda test port in the engine bay.

The background: As part of the rebuild, I went through all the CIS componentry cleaning, spec'ing, etc. The intake runner boots are all new, as well as the injector O-rings. I also checked the injectors for spray pattern. Among other new items, the oxygen sensor is a new Bosch single wire unit.

The dwell meter is an old '80s-vintage -- which fits in with the subject car -- Actron Model 612 dwell/tach unit. This unit has 8- and 6-cylinder scales with instructions to double the 8-cylinder numbers to derive a 4-cylinder result. Then, to compare the readings with the Bentley specs you have to calculate the ratio of degrees to duty cycle (45 degrees equals 50 percent).

More background: On the first startup, the engine ran but was rich, as exhibited by a hunting idle and a hesitance in throttling up (the engine was loading up). I figured I would be rich on the first spool up because of renewing many of the old wear items. Now, leaning the mixture would smooth that hunting. However, when going too lean, the hesitancy on throttle-up would return.

The ignition timing has been set. Regarding vacuum leaks, if I remove the oil tank cap, the idle roughens and drops down, so I'm feeling good about where I'm at on that front.

OK, so now to the duty-cycle testing. On cold start up, the meter (I'm quoting 4-cylinder percentages here, which I understand is what the lambda system standard is based on) is at 83 percent, which is outside the bounds of what Bentley says I should be seeing for an '81-'83 SC. Bentley says it should be at 65 percent. For an '80 SC, the spec cold start percentage is 85.

Once warm, the best setting is right around 63 percent. This is with the engine idling at 950 RPMs. At this position, there's no hesitation on throttle-up. Bentley says all lambda-equipped models should be in the 45 to 65 percent range. If I go anywhere (lean or rich) from that position, the idle worsens and the throttle-up hesitancy returns, and if you compensate through the idle screw (air bypass), the dwell needle doesn't move much if at all.

The curious thing is at warm idle, there's no "dithering" of the dwell needle. It stays pretty much exactly in place. It does change if you throttle up (it goes up and then comes back down as the engine RPMs return to idle speed).

This engine has new rings, so I would expect to be seeing some slightly rich numbers to ensure proper running until things seal up. What I'm looking for is confirmation that I'm on the right track with this, and that my lambda system is working (and not running in default mode).

What say everyone? Thanks!

Brian

1982911SCTarga · 08-28-2008, 04:01 AM

The more I think about this, the more I believe that my oxygen sensor isn't doing anything. That is, the system is operating in default mode.

Brian

1982911SCTarga · 08-28-2008, 05:04 AM

However, if it were operating in default mode, I should be seeing a 50 percent duty cycle on the frequency valve. That would be 45 degrees dwell, or 22.5 on the 8-cylinder scale of my dwell meter (to double and translate to 4 cylinders). So what's going on?

Brian

Jim Williams · 08-28-2008, 06:42 AM

I couldn't find it in your post, but are you disconnecting the O2 sensor before you make your measurements?

1982911SCTarga · 08-28-2008, 06:51 AM

Jim, I haven't disconnected the oxygen sensor yet. Doing so would confirm whether or not the system is in default mode, right? Thanks!

Brian

1982911SCTarga · 08-28-2008, 04:48 PM

OK, I'm feeling a lot better about this.

First off, doubling the dwell angle number for 8 cylinders to produce a 4 cylinder result doesn't -- ahem, to borrow an MFI word -- correlate. At least with my dwell meter -- even though on the meter's face it says to do so, but maybe that's only for RPM readings? It's been probably 15 years since I've played with this dwell meter.

I figured this out by doing as Jim suggested and warming the engine up and then unplugging the oxygen sensor. I could see the dwell angle swing into default mode, which is a duty cycle of 50 percent. Given that, I could then plug it back in and watch the oxygen sensor do its thing and bring the duty cycle down to 40 percent or so.

Looking good! I've learned what that test port can do! Hey, and I got to use my crappy dwell meter! Here's to stoichiometry!

Brian

Jim Williams · 08-28-2008, 04:58 PM

Brian,

I guess you could say there are three modes for the lambda system. Open loop, with the engine "cold"; closed loop, with the mixture controlled by the O2 sensor; and closed loop (sort of), with the O2 sensor unplugged or plugged in but inoperable.

In this last mode, the mixture can be adjusted to a value different than the stoichiometric value ot 14.7:1 air to fuel ratio, and the O2 sensor will not be in control.

Using the dwell meter with the engine warm (closed loop), and with the O2 sensor unplugged, the electronics (ECU) should be sending a 50% duty cycle to the frequency valve. In this mode, adjustments could be to the idle mixture to produce a desired value other than the 14.7 base line. Regardless of the mixture, the meter is going to read the 50% number.

With an operable O2 sensor, the mixture is adjusted at warm idle to this same number (50%). The duty cycle will swing around the 50% value as the system tries to maintain the 14.7:1 ratio, so the meter has to be watched so that the average reading is 50%.

A further system check can be made of the ECU electronics by disconnecting the O2 sensor and monitoring the dwell meter. Grounding the O2 sensor lead to the ECU should drive the system toward a higher duty cycle number. This is equivalent to a lean condition telling the Frequency Valve to richen the mixture. Conversely, a 1.5 volt battery, (+ side to the O2 sensor lead,minus side to ground) should drive the system lean.

1982911SCTarga · 08-28-2008, 05:10 PM

OK.

Quote:

Using the dwell meter with the engine warm (closed loop), and with the O2 sensor unplugged, the electronics (ECU) should be sending a 50% duty cycle to the frequency valve.

This is what I see in closed loop operation. When I plug the oxygen sensor in, the needle drops back to 40 percent.

Quote:

With an operable O2 sensor, the mixture is adjusted at warm idle to this same number (50%).

Given this, I'm too lean then.

Quote:

The duty cycle will swing around the 50% value as the system tries to maintain the 14.7:1 ratio, so the meter has to be watched so that the average reading is 50%.

I don't see the needle swinging rapidly. You can change the mix and the needle will move around, but then settle in one place. Maybe it's my meter?

Thanks for your help!

Brian

1982911SCTarga · 08-28-2008, 05:54 PM

I just got done trying to bump the closed-loop, oxygen-sensor-plugged-in duty cycle to a 50 percent duty cycle to match the unplugged-oxygen-sensor default. I can't do it -- you have to increase the air through the idle bypass screw and you still quickly get wildly oscillating throttle.

The optimal seems to me to be where I was in my fifth post.

Brian

Jim Williams · 08-28-2008, 06:01 PM

Quote:

Originally Posted by 1982911SCTarga

OK.

This is what I see in closed loop operation. When I plug the oxygen sensor in, the needle drops back to 40 percent.

Given this, I'm too lean then.

I don't see the needle swinging rapidly. You can change the mix and the needle will move around, but then settle in one place. Maybe it's my meter?

Thanks for your help!

Brian

Then you're too rich...... 40% would indicate what the O2 sensor is telling the frequency valve to do -- "Go Lean" -- meaning the mixture is running too rich.

There wont be a rapid swing... the system can't react this fast. There being no rapid swing doesn't indicate a problem. Your adjustment and the system's reaction seem to indicate the system is doing what it should.... Adjust the duty cycle to 50% and the system should be adjusted properly.

1982911SCTarga · 06:36 PM

OK, then I was going the wrong way (richening from already rich) from the 40 percent closed-loop, plugged-in duty cycle. It seemed counterintuitive, but I get it.

Thanks, Jim. I'll report back.

Brian

1982911SCTarga · 03:46 AM

Here's the latest. Adjusting from rich to lean (Jim's advice) in closed-loop, oxygen-sensor-plugged-in mode, I can get needle to move slightly higher off the 40 percent duty cycle.

But then, I hit this point, where just the slightest turn to lean (almost a bump, actually) of my allen wrench swings the dwell needle slowly past the 50 percent point to damn near 90 percent. To get it back down, I have to go rich past the swing point (where it then returns to something like 43 percent duty or so). Then, I sneak back up on leaning and, bam, the needle does another wild swing past the 50 percent point.

What am I seeing happening here?

Brian

Jim Williams · 08-29-2008, 08:23 AM

Just to verify what I think I'm hearing --

As you are turning the mixture adjustment screw, going from rich to richer -- You're turning the screw clockwise. As you do this, the duty cycle is decreasing from 40% to a lower number.
As you go lean (counterclockwise) the duty cycle reading increases back towards 50% and keeps going on it's own up to 90% or so (this is the rich stop, where the circuit hits a limit and can't make it any richer via the Frequency Valve).

Right off, I'm not clear on what could be causing this. But to help rule out what could be a mechanical issue, that is an issue not caused by the Lambda system, you could try the following:
With the engine warm and the O2 sensor unplugged, try to get the mixture in the ball-park of stoichiometric by using the method outlined here on my website:
http://members.rennlist.com/jimwms/CIS/CIShome.html
Testing>•Air/fuel ratio

It would be easier of course if you had access to a CO meter, and could set the mixture to a CO of maybe 1% or so and see how the car drives, or to see if you could set it there and if it would stay there.

Another question comes to mind, how old is the O2 sensor?

1982911SCTarga · 08-29-2008, 09:34 AM

Jim, the oxygen sensor is new.

Quote:

As you are turning the mixture adjustment screw, going from rich to richer -- You're turning the screw clockwise. As you do this, the duty cycle is decreasing from 40% to a lower number.
As you go lean (counterclockwise) the duty cycle reading increases back towards 50% and keeps going on it's own up to 90% or so (this is the rich stop, where the circuit hits a limit and can't make it any richer via the Frequency Valve).

Correct, as I go from rich to lean, the needle stays fairly in place at something like 42 percent and then with just a tiny nudge of the wrench, the needle goes to full duty. There's no in-between.

I'll do the ballpark setting and see where that gets me on the dwell meter.

Brian

1982911SCTarga · 08-31-2008, 01:30 PM

OK, problem solved, I think.

I messed it with yesterday and got no joy at all after checking all my exhaust connections (looking for a leak upstream of the oxygen sensor). I also removed the new oxygen sensor, cleaned it and re-installed. Still, I could not adjust the dwell-duty cycle of the frequency valve. It would stay low until I hit a mixture point that would swing the dwell needle far left where it stayed until I richened the mixture back up.

This morning my wife and I took the targa when we went out for breakfast. When I got home I put the dwell meter on it using the test port connection. I've never messed with the mixture on it so I wanted to see what it would look like. Finally I got a good idea of what you're supposed to be seeing as the K-Jet computer, oxygen sensor and frequency valve talk amongst each other. In other words, I could see on the dwell meter the slow and very small "dithering" of the needle as the frequency valve corrects the mixture.

OK back to the cabriolet. Still nothing new there, so I pulled the passenger seat to get at the K-Jet box, the oxygen sensor relay and the enrichment circuit. Here, things get a little screwy. I had been using a new oxygen sensor relay.

Anyway after warming the engine, I unplugged the new oxygen sensor relay and noticed no difference in how the engine was running. I then unplugged the K-Jet box and again noticed no difference.

I then plugged the K-Jet box back in along with connecting the old oxygen sensor. I could tell immediately that something was going on.

Moving back to the engine bay I looked at the dwell meter and -- I'll be damned -- it's dithering! It was dithering wildly, but at least the system was doing something. As Jim said earlier, it was rich, so I began slowly leaning the mixture out.

As you do this, the dithering movement swings become smaller and smaller. I think I'm good now!

What's bizarre is a bad new oxygen sensor relay. Go figure.

Brian

Jim Williams · 08-31-2008, 02:53 PM

Congratulations on taking the time to do a little experimenting, and learning the system in the process. Problem solutions are not always straight forward with the CIS. Sometimes it just takes knowledge of the system and a little perseverance.

1982911SCTarga · 08-31-2008, 04:05 PM

Thanks, Jim. It's pretty neat how the dwell meter diagnostic goes when everything works as it should. That test port is a dead simple thing to get to check on things, too.

This is a definite keeper in the CIS bag-o-tricks.

Brian

GothingNC · 05-25-2010, 12:27 PM

I now this is an old thread but does the o2 sensor need to be connected to confirm the mixture?

Jim Williams · 05-25-2010, 02:20 PM

Quote:

Originally Posted by GothingNC

I now this is an old thread but does the o2 sensor need to be connected to confirm the mixture?

By "confirm" do you mean measure it? Without it connected, you can't use the duty cycle measurement method to adjust it to Stoichiometric, or 14.7:1 air to fuel.

But without it connected, you can get an idea, but not a definite number, by raising the airflow sensor plate at warm idle. This method has been described here on the forum, and other places, such as my CIS Primer website.

GothingNC · 05-25-2010, 04:54 PM

Jim,

I have had the Targa for about 4 1/2 year, PO set it up with the O2 sensor disconnected -SSI's
All intake seals, injector seals and vacuum lines were replaced about 20,000 miles ago long with new injectors

I followed your procedure with the sensor plate and the rpm rise slightly just before dropping. Pull down a hair and it drops.

Just trying to tracking down a random idle hunting issue that occurs when it is warm out.

After a valve adjustment (1 exhaust loose and 3 intake tight by a hair) new plugs rotor and cap I am still coming up with the issue.

I would be at a light, idle is fine at 900, rev the gas and it drops right back to 900 then a minute or two later it would suddenly start hunting from 750 to 900 for a a while and then stop a few miles later.

If I turn of the car for a few seconds and restart it stabilizes.

Removing the oil cap causes the idle to drop when it is idling at 900

I tried the same when it was hunting, remove the oil cap and the idle smooths out.

Oxygen sensor relay