FAQs about Fuel Injector O2 Sensor.

Part number for O2 sensor.

Well, I screwed up.

I think that I need a change in eye-glasses, 'cause I was reading DY-660 from the actual DY-606. So, the O2 sensor from Ford, that'll fit in a Volvo, the DY-606 is STILL a current part.

Sorry for any confusion... :>(I *did*, however, get the beck-Arnley part for $79... Part # 156-8000.

I'll be putting it on this weekend.

Nick

> From meaddata!meaddata.com!nick@uunet.uu.net Thu Mar 11 12:55:07 1993
> Date: Thu, 11 Mar 93 11:07:16 EST
> From: nick@meaddata.com (Nick Gough)
> To: swedishbricks@me.rochester.edu
> Subject: Re: O2 sensors
> Content-Length: 439

How can I inspect the O2 sensor?

Brian and fellow netters with Lambda Volvo's...

>What about the O2 sensor? [effecting cold start and running]

It doesn't come into the picture until the car is warm. The 3-wire O2 cars heat the O2 sensor to get it to the operating temp quickly, improving the emmisions sooner.

It is simple to check the O2 sensor..

0) inspect motor grounding strap and battery ground strap.

1) warm the car and let it idle.

2) disconnect the sensor lead from the fire-wall connector.

3) drop the firewall connector to ground and check the O2 sensor lead voltage. It should read about 1.2 Volts. An old sensor will only read about .9 volts. (this is how you force the motor to run rich)

Note: do not apply more than 2 volts to the firewall connector lead. This lead goes straight to the computer box and an overvoltage can dammage your FI brain.

>If this is a problem does anyone know the Domestic equivalent for the 3-wire sensor?

The domestic 3-wire sensors are all built on the same catalytic Oxygen sensing technique.. they should all read the same voltage for a given mixture. The Toyota/Nissan sensors may read differently...

herm

Herman L.N. Wiegman -> wiegman@orion.crd.ge.com
General Electric - Corporate R&D, Schenectady NY
- the Flying Dutchman in the DSP Swedish Brick -

Alan, [Oxygen sensors, summary]

Operation:

The Oxygen sensor is the key to a clean running automobile. If it should fail or age, then your brick will run less efficiently.

The O2 sensor measures the oxygen content in the exhaust. If you pass too much oxygen, then you know that teh motor is running lean and the FI brain compensates appropriately. This cycles every second or so to optimize the fuel mixture.

Locating:

The O2 is found in the exhaust manifold close to the motor. The operating temperature is high for the sensor so it needs to be withing 1 foot of the the motor. It is a silver like can which screws into the manifold. There is one to three wires running from it to the fire wall. The single lead sensors are simple to check... as described yesterday. (measure the O2 output voltage with the computer lead grounded. The three leaded sensors require the two sense leads to be shorted together and the voltage reading is between two of the O2 sensor leads. The heater winding lead is ignored.)

Diagnosing:

The car will tend to loose gas milage over time and will tend to run slightly lean (warmer combustion). The spark plugs may "glaze" more and the car may not pass inspection.

Maintenance:

The O2 sensors tend to last ~60kmiles but are on a 30kmile inspection cycle. They cost $30~$50 for the single lead unit and $79~$125 for teh three lead part...

>From Nick Gough nick@meaddata.com
>the actual DY-606. So, the O2 sensor from Ford, that'll fit in a Volvo, the
>DY-606 is STILL a current part.
>I *did*, however, get the Beck-Arnley part for $79... Part # 156-8000.

herm

Observations after replacing my lambda-sond computer and frequency valve.

I replaced the lambda-sond computer and frequency valve on my 80 240 yesterday, and have a few observations to pass along (I got a used computer for $50 from Strandbergs Salvage and a used freq valve for $60 from the RevolvStore, so it wasn't too expensive.)

Contrary to what all the manuals say, the open-loop duty cycle for an 80 appears to be around 69 degrees - I've measured it on two different computers with two different dwell meters. The manuals (both the Volvo problem solver and the Bentley Bosch FI manual) indicate that it should be in the mid-50s.

The two tests for the rich and lean stop conditions that the Bentley manual suggests work pretty well. Just unplug the O2 sensor from the computer and work with the green wire that goes to the computer. First ground the wire against the engine (this simulates a lean condition) The duty cycle for the frequency valve should go up to about 90 degrees and the engine should speed up. Then touch the wire to the positive terminal of a 1.5V battery and ground the negative terminal of the battery to the engine. This simulates a rich condition - the duty cycle should drop into the teens and the engine should nearly stall. (The original computer in my car didn't respond to the 1.5V battery, and Herm suggested that maybe a 2V battery was really necessary. But the replacement computer did respond to 1.5V. So maybe the original computer was busted. On the other hand, it had the same open loop duty cycle as the replacement and responded correctly to the lean condition - it only failed the rich condition test.)

As Darryl pointed out to me, the freq valve is just a solenoid that is hooked up to the car's battery and the lambda-computer. The computer works by opening and closing a ground connection, allowing current to flow from the battery through the freq valve to the ground. (When the current flows through the freq valve soleniod, it pulls back a plunger and allows more fuel to flow back to the gas tank. The Bentley manual contains a good explanation of how this serves to richen the mixture.) The control current from the computer is about .5V (DC). A total of about 6V (DC) flows through the freq valve to the computer ground when the freq valve closes. So one way to test whether your freq valve is opening or not is to test the voltage at the duty cycle test point. If you get no voltage there, the computer is hosed (or there's a short, etc.) If you get .5V (DC) there, the computer is working, but the freq valve isn't opening. If both the computer and freq valve are working, you'll get about 6V(DC) there.

How to test your O2 sensor.

Bricksters,

I posted something a while back on O2 sensor testing, but since the issue has resurfaced I thought I'd repost some of the O2 sensor info from earlier.

1) You need a voltmeter with high imput impedance. 10 MOhm is a typical recommendation.

2) What you're looking for is the O2 voltage CHANGING.
The voltage will vary between 0.1V and 0.9V. When it's low the engine is running lean (remember Low-Lean) When it's high it's running rich. As the ECM receives the lean/rich info from the O2-sensor it will compensate for the lean/rich condition. As it compensates (typically by adjusting fuel injector puselwidth), the lean rich condition will eventually swing the other way and the entire procedure can start all over again.

With a WARM engine at 2500 rpm the number of cross counts (the number of times the voltage crosses the middle 0.5V) should be around 8-10. It it's any less than that the O2 is possibly dying. If the voltage is stuck the O2 sensor is dead.

3) Because you're looking for voltage swings the choice of voltmeter becomes critical. Your typical cheapo voltmeter heavily averages its display values. More sophisticated ones still average heavily, but also add a fast reacting "needle type" display. I'm specifically thinking Fluke 87 or 88. While it can be hard to manually figure the number of cross counts from such an electronic "needle", at least it's possible to see that the O2 voltage is not stuck.

A portable oscilloscope would be the ultimate way of looking at an O2 sensor. With a scope you can not only see all the cross counts but can also check the response time of the O2sensor as you temporarily make the engine run lean or rich. The response time can, if you have a baseline with which to compare, tell you if the O2 sensor is getting lazy.

claes

Excellent post Claes!! A few questions (as I may have a dying O2 sensor at the moment):

Could you please explain the need for a high impedance VOM?

What is the proper VOM lead hookup assuming:

        1-wire sensor or....
        3-wire sensor

I've read and re-read the FAQ articles on this issue but am still a bit unclear on the principle (and don't want to risk damaging the ECU (or ECM as some have called it).

Thanks for the informative post!

Mic

P.S. Now if I could only figure out a way to make an o-scope fit my gauge panel... :=)

The 1-wire and 3-wire sensors are really the same thing. The two extra leads on the 3-wire sensor are there to heat the sensor so it starts working sooner when the catalyst is cold. (Dr. Tim had a post pertaining to this issue a while back)

To monitor the O2 sensor, simply tap into the signal wire, or backprobe it somehow. On my 745 the O2 sensor wires (signal wire and the two for heating) are all available at the right strut tower. I don't know where they are on other Volvos.

> I've read and re-read the FAQ articles on this issue but am still a bit
> unclear on the principle (and don't want to risk damaging the ECU (or ECM
> as some have called it).

Using a high-impedance voltmeter won't damage the sensor as long as you don't put it in amps mode while connected to the sensor since that will effectively short it to ground through the meter.

The O2-sensor input to the ECM is by necessity also of a high-impedance type, so the chances of destroying the ECM, say by connecting the O2sensor input to 12V, are pretty slim. Bosch makes very good and reliable ECMs and I'm sure they have all sorts of input protection against statics, shorts etc etc etc.

I'm presuming the various special O2-testers on the market are all high-impedance. I own a sophisticated voltmeter and use that for everything on the car, so I have no personal experience with these special testers.

O2-sensors aren't really that tricky to diagnose. It's just that, as I mentioned earlier, you need the right tool.

claes

Claes mentioned using a good multimeter like a Fluke 87. I just brought my Fluke 87 home from work tonight and tested my O2 sensor. My manual says (Bentley) to first connect the positive lead to the sensor (don't unplug sensor wire, just pull rubber boot back) and the negative meter lead to ground. Look for fluctuating voltage above and below .5 volts. Then you can also unplug the sensor wire and connect the meter pos. lead to the sensor and you should see a steady .5 volts if the sensor is good.

The Fluke meter is great. It can also check duty cycle. I set my air/fuel mixture with it while monitoring the frequency valve. It also has a frequency mode which I'll bet can check engine r.p.m. Only drawback, it sells for about $280.

Jared Aos
(No longer rough running) 85 245TI

Absolutely NOT!

>to first connect the positive lead to the sensor (don't unplug sensor wire, just pull rubber boot back) and the negative meter lead to ground. Look for fluctuating voltage above and below .5 volts. Then you can also unplug the sensor wire and connect the meter pos. lead to the sensor and you should see a steady .5 volts if the sensor is good.

The oxy sensor, when disconnected, should let the engine return to the "base CO" level setting... which should be between 1% and 2% CO. This mixture is decidely rich, and will cause the Oxygen Sensor to read firmly at the top of its output range (approx. 1v).

The other thing I've noticed is that while the mixture oscillates, at least under the Lambda Control brains I've tried (four at last count), it tends to oscillate about a mean of more like 0.7 volts; though it has a noticable lean period under gentle acceleration.

I've listened to these manuals before, and never could get the car (at that time the v6) to run right adjusting the base mixture to 0.5v. It was much too lean.....

-tim

---------------------------------------------------------------------------
Timothy Takahashi                  EMAIL : tim@sr71.arc.nasa.gov
M/S 247-2 (AA/AAL)                 PHONE : 415-604-4976
NASA Ames Research Center          U. Rochester Alum
Moffett Field, Ca. 94035            & Swedishbrick Enthusiast

>Absolutely NOT!

After reading the fine print in my manual I see that the procedure I stated (steady .5v from O2 sensor with sensor disconnected) only pertains to a turbo with CI injection. The Lambda system is independent of the CI injection system whereas on Jetronic injection the Lambda and Jetronic are fully integrated. Apparently the jetronic systems duty cycle the frequency valve at a rich (70-77%?) rate with the O2 sensor disconnected. Hence the high sensor voltage. With the CI (continuous injection) on turbo's the Lambda control duty cycles the frequency valve at a baseline rate (47-53%) with the O2 sensor disconnected. Hence the steady .5v sensor output.

I guess the bottom line is, no matter what type of meter you use, consult a manual for your type of injection system when checking the O2 sensor or adjusting the base mixture.