Rough Cold Idle After O2 Sensor Replacement??
Thread Starter
Man of many SIGs
Rennlist Member
Rennlist Member
Joined: May 2005
Posts: 8,724
Likes: 12
From: Florida, USA
Rough Cold Idle After O2 Sensor Replacement??
I changed my O2 sensor recently. And ever since my car has had a varying idle when cold. It never had this issue before. The O2 sensor I used was a Bosche unit for an 89 Ford Taurus. Before you laugh, that was actually recomended here on Rennlist by Mark Robinson. After my car warms up well, there are no more idle issues. Basically what it does is vary the idle about 50 rpm up and down constantly. Its easy to hear but barely visible on the tach. Could this be somehow be caused by hooking up the two heater wires backwards. I just clipped and spliced with no regard as to which heater wire was supposd to go where. This really shouldn't matter if it's just some sort of coil resistor. Does anyone know how the "heater" works on an oxygen sensor? I was thinking about unplugging the heater wires and seeing if it made a difference in the cold idle. I just thought I would post while I was thinking about it. Has anyone had this issue before?
1984 928s w/ 5 speed
1984 928s w/ 5 speed
Rennlist Member
Joined: Oct 2003
Posts: 7,210
Likes: 17
From: Nova Scotia
The heater is resistive as you said, so the two white wires have no specific connection - just complete the loop, and connect the black signal wire to the LH ( what model year car?).
If you put a new sensor on a x-y plotting meter, some take a while to begin generating output, others swing wildly until the voltage output begins to move in a slow sine wave form, assuming the fuel injection is OK. You may see some initial variance due to this, which disappears in 30 min running time. edit ... this is a first time thing only with an out of the box sensor.
The Bosch unit mentioned should pose no problem .... any 3-wire sensor of that generation will do the job.
If you put a new sensor on a x-y plotting meter, some take a while to begin generating output, others swing wildly until the voltage output begins to move in a slow sine wave form, assuming the fuel injection is OK. You may see some initial variance due to this, which disappears in 30 min running time. edit ... this is a first time thing only with an out of the box sensor.
The Bosch unit mentioned should pose no problem .... any 3-wire sensor of that generation will do the job.
Last edited by Garth S; Dec 11, 2005 at 09:08 PM.
Burning Brakes
Joined: Oct 2004
Posts: 877
Likes: 1
From: Suburban St. Louis in Illinois.
Don't know, but 50 RPM,... up or down,... sounds like you are listening for a problem that may or may not be a problem. You put this "cheaper" alternative O2 sensor in there, and now you "notice" an unusual fluctuation...Believe me...if all is well "off-idle"...check out vacuum leak... bad plug wires...or whateve elser.. you might suspect. The O2 sensor (if I recall correctly...) is only in control of the fuel managementl, after coolant warmup, and in part-throttle, warm, cruise type conditions. Idle...and WOT are strictly open loop. Somebody correct me if I'm wrong here... 
Nordschleife Master
Joined: Sep 2004
Posts: 9,429
Likes: 6
From: Not close enough to VIR.
Unplug the sensor and see if the idle hunt changes.
The heater should heat to operating temp in under 30 seconds and polarity on the white wires does not matter. The variance Garth mentioned should disappear in under 2 minutes or whenever the engine reaches operating temp. With an an heated sensor it could take longer.
Was the old sensor rusted in place? 3-wire sensors ground the sensor to the exhaust. If the threads were rusted or you used non-conductive sealant/antiseize the sensor will not ground properly resulting in a bad signal.
The heater should heat to operating temp in under 30 seconds and polarity on the white wires does not matter. The variance Garth mentioned should disappear in under 2 minutes or whenever the engine reaches operating temp. With an an heated sensor it could take longer.
Was the old sensor rusted in place? 3-wire sensors ground the sensor to the exhaust. If the threads were rusted or you used non-conductive sealant/antiseize the sensor will not ground properly resulting in a bad signal.
Thread Starter
Man of many SIGs
Rennlist Member
Rennlist Member
Joined: May 2005
Posts: 8,724
Likes: 12
From: Florida, USA
Garth, you mention that it may take 30 min to plot itself. I have driven the car at least 5 times since this repair. I'm not sure if any trip was 30 min long or not. Would the car need to be running continously for 30 min. to allow for this? Thanks
Jeff, it is deff. wandering a little. The plugs, wires, rotor, and cap were all replaced a couple of weeks prior to the O2 so I feel confident in ruling them out. As far as vacume lines, I didn't have my hands anywhere near any vacume lines during that maintenance. But I will try to check for that just in case. As far as idle being open loop. Are you reffering to the micro switch on the low side of the throttle? If so, mine doesn't work and didn't work prior to the O2 sensor. According to another thread here on Rennlist, adjusting so that micro switch is innoperative is actually supposed to be a positive performance enhancement. I'll try to locate that thread. I think it was M. Kibort who brought it up in a discussion about L-Jet troubleshooting.
Flying Dog, I didn't use any sealant on the sensor and the anti-sieze was already on the sensor when it came to me. I would assume it would be OK. There wasn't that much of it anywway.
Oh yeah. One more thing. I also noticed that the idle speed is 200 RPM higher upon startup/ warm-up. It used to be 1000 an dnow it is 1200. I was planning on adjusting it down to where it is supposed to be but wanted to do that after the O2 install, just in case it was causing it. I guess it wsn't. The WS manuals say that a O2 sensing tool is required to set the idle adjustment properly. Is this really necessary?
1984 928s 5-speed
Jeff, it is deff. wandering a little. The plugs, wires, rotor, and cap were all replaced a couple of weeks prior to the O2 so I feel confident in ruling them out. As far as vacume lines, I didn't have my hands anywhere near any vacume lines during that maintenance. But I will try to check for that just in case. As far as idle being open loop. Are you reffering to the micro switch on the low side of the throttle? If so, mine doesn't work and didn't work prior to the O2 sensor. According to another thread here on Rennlist, adjusting so that micro switch is innoperative is actually supposed to be a positive performance enhancement. I'll try to locate that thread. I think it was M. Kibort who brought it up in a discussion about L-Jet troubleshooting.
Flying Dog, I didn't use any sealant on the sensor and the anti-sieze was already on the sensor when it came to me. I would assume it would be OK. There wasn't that much of it anywway.
Oh yeah. One more thing. I also noticed that the idle speed is 200 RPM higher upon startup/ warm-up. It used to be 1000 an dnow it is 1200. I was planning on adjusting it down to where it is supposed to be but wanted to do that after the O2 install, just in case it was causing it. I guess it wsn't. The WS manuals say that a O2 sensing tool is required to set the idle adjustment properly. Is this really necessary?
1984 928s 5-speed
Nordschleife Master
Joined: Sep 2004
Posts: 9,429
Likes: 6
From: Not close enough to VIR.
There are two idle adjustments, one for air and one for mixture. It sounds like you are getting too much idle air. There is a large screw on the front of the throttlebody to adjust idle. All you need is a screw driver for that adjustment.
Rennlist Member
Joined: Oct 2003
Posts: 7,210
Likes: 17
From: Nova Scotia
The sensor would be fully operational after the first few min - you're well past that. The anti-seize on the new sensor is the correct stuff: it is applied sparingly to avoid contact with the sensor head - for it could poison the sensor. It is a good idea to run a thread chase ( or tap) through the sensor well to clean it for an improved ground; however, unless the threads were really bad, you should be OK.
IMHO, it is a good idea to have the idle micro switch on the throttle body (TB) functional - without it, what do you have for engine braking? It reduces fuel at idle, and if it is not functioning, it may contribute to a strange idle: what was your initial problem ??
It can't hurt to reduce the idle by turning the huge sloted set screw (~1/2" head) on the right of the TB - which controls bypass air around the throttle plate for idle. When that is in the ball park, the CO adjust is actually the 3mm allan key head on the lower left of the AFM when facing the engine ( you need a 10-12" extension to reach it). It gets to be a long story how to adjust it, but this was previously posted: in short, it can be set much like a carburator mixture screw, but it controls unmetered bypass air around the vane that measures total air flow ... as few of us have CO sensors ( unless you care to hook a multimeter up to that new Bosch sensor - which does the same thing).
IMHO, it is a good idea to have the idle micro switch on the throttle body (TB) functional - without it, what do you have for engine braking? It reduces fuel at idle, and if it is not functioning, it may contribute to a strange idle: what was your initial problem ??
It can't hurt to reduce the idle by turning the huge sloted set screw (~1/2" head) on the right of the TB - which controls bypass air around the throttle plate for idle. When that is in the ball park, the CO adjust is actually the 3mm allan key head on the lower left of the AFM when facing the engine ( you need a 10-12" extension to reach it). It gets to be a long story how to adjust it, but this was previously posted: in short, it can be set much like a carburator mixture screw, but it controls unmetered bypass air around the vane that measures total air flow ... as few of us have CO sensors ( unless you care to hook a multimeter up to that new Bosch sensor - which does the same thing).
Last edited by Garth S; Dec 12, 2005 at 05:03 AM.
