PDA

View Full Version : 3-wire O2 innacurate



iTurbo
05-29-2006, 05:25 PM
I have a 3-wire Mopar O2 sensor on my Shelby Lancer. It seems to be very innacurate at WOT; which isn't very surprising considering it's narrow-band and doesn't have a sensor ground like the 4-wire O2 sensor.

About a year ago I had the car on a dyno with a wideband sensor up the tailpipe. The car does not have a catylitc convertor BTW. The wideband sensor showed the mixture was extremely rich as soon as the car spooled up. So rich it was that the A/F went off the scale. My autometer was showing 18 lights.

Seems like 18 lights is as rich as it will show, even though the wideband was showing it to be extremely rich. I'm thinking that the 3-wire sensor doesn't have a very good ground path which is pulling the output voltage down. Is there anyway I can make the ground better for the 3-wire sensor or is converting to 4-wire the best fix? If I convert to 4-wire O2, where should I attach the extra ground? Should I splice it into the sensor ground circuit that the other sensors use, or should I just run it to the cylinder head or (-) battery terminal?

iTurbo
06-02-2006, 03:02 AM
Narrow band O2 sucks!

Here's why:

With my Lancer at 14 psi (stock TII setup w/3" exhaust and +20s), my scanner reads about .90 from the O2 sensor at WOT. I turned the AFPR in one turn: no difference. Raised the boost to 16 psi, and then 18 psi. Still .90 at WOT. I'm betting it's actually very rich still. Wideband is the only way to go for tuning. I can't wait to get an Innovate Motorsports LM-1 and compare it to what the scanner reads.

86Shelby
06-04-2006, 01:06 PM
I did a few different things on my GLHS to try to improve the ground signal for the o2 sensor since it's a 3 wire sensor; it ended up being an extra ground for the LM and all 5v sensors. I cleaned off the swingvalve where it mounts so it was bare metal there; did the same to the sensor's gasket area. I also removed one exhaust manifold nut, cleaned the crap out of that area and added a ground wire from there to the battery. And also a second ground wire from that point to the LM's ground wires inside the car.

Here's a rough diagram of how the ground path should work. Note there's no actual wire from the O2 sensor, it's path is through the turbo and exhaust manifold.
Battery---exhaust manifold stud---O2 sensor
..........................|
.........................LM

Chris Faulk
06-04-2006, 01:23 PM
You can install a 4-wire 02 and make sure the grounds are CLEAN. Many many problems have been attributed to poor grounds whether it be dirty or not soldered, or whatever. Some will ground the 4-wire close to the sensor, but I always go the extra mile and run it over to the battery, or one of the ground straps that come off of the battery...no guessing there!

Your best bet is to install a wideband meter...the cost is certainly justifyable.

Chris Faulk
06-04-2006, 01:25 PM
Narrow band O2 sucks!

Here's why:

With my Lancer at 14 psi (stock TII setup w/3" exhaust and +20s), my scanner reads about .90 from the O2 sensor at WOT. I turned the AFPR in one turn: no difference. Raised the boost to 16 psi, and then 18 psi. Still .90 at WOT. I'm betting it's actually very rich still. Wideband is the only way to go for tuning. I can't wait to get an Innovate Motorsports LM-1 and compare it to what the scanner reads.

You won't be happy with what you find...as in it's likely further off then you think. I bought the LC-1 'cause of the price. I'll get the AuxBox later for expansion and logging. :thumb:

I've gotten to a point where I un-installed the narrow-band voltage meter, AND I rarely look at my EGT gauge now. :(

http://www.turbo-mopar.com/forums/showthread.php?t=4102

Read post no.18 for extra info.

MiniMopar
06-05-2006, 02:15 PM
Not only is the narrow band, well narrow, it is very non-linear. Once you get away from stoich, the response slows dramatically. So, the change in ratio between 0.8V and 0.9V is several times that of the change between 0.7V and 0.8V. Here is one picture showing the curve. I had a better link at one point, but you get the idea:

http://www.aa1car.com/library/o2volts.gif

Add to that the grounding errors and the fact that these sensors not in any way calibrated when operated far away from stoich, you begin to realize how unreliable they are for direct tuning.

Don't get me wrong...I still have one because it is great for quick diagnosis of issues while driving and for reassurance that you aren't way lean due to some mechanical problem...like an oil pressure guage. But when you are making major changes and are tuning your A/F, you really need a wide-band to figure out what is really going on.

Tony Hanna
06-05-2006, 03:38 PM
I had an experience the exact opposite. I was short on cash and installed a 1 wire sensor on the daytona (supposed to be 3 wire). It really surprised me that the 1 wire sensor would light the 10 light at full rich. Other than taking a little longer to warm up, it worked flawlessly.

MiniMopar
06-05-2006, 04:42 PM
It probably lit the 10th light because the reference ground for the sensor (the swingvalve) was floating a tenth or two above the reference ground for your meter (the chassis or the battery). That makes the output of the sensor look a tenth or two higher, from the meter's perspective.

Tony Hanna
06-06-2006, 01:01 AM
It probably lit the 10th light because the reference ground for the sensor (the swingvalve) was floating a tenth or two above the reference ground for your meter (the chassis or the battery). That makes the output of the sensor look a tenth or two higher, from the meter's perspective.

That makes sense. That sensor is still on the car (John's now). Next time I'm up there, I may move the meter ground from the battery to a minifold bolt just for kicks to see how the meter acts. I'm sure he'd rather know if it's reading higher than the actual o2 output.

MiniMopar
06-06-2006, 01:49 AM
Try moving it to the fuel rail bolt. That's where the LM gets its ground reference from.

Tony Hanna
06-06-2006, 03:17 AM
I'll give it a shot.

iTurbo
06-06-2006, 03:21 AM
Thanks for all the replies and ground ideas. This turned into a really good thread.

boost geek
06-11-2006, 12:20 AM
Would a bad ground (fpr nut) cause the 3 wire sensor to make a Dawes read red? I put the fuel pressure to 62 psi (with vac) and still reads intermitent red. I have a 2.2 with ported everything usually running 6-7 psi on a GLHS Stage 2 module and T2 injectors, huge intercooler, I should be way rich.
This o2 sensor has lived through a blown headgasket 2 years ago, drove it home for about 50 kms leaving a horrendous trail of white smoke. If it was shot wouldnt go into closed loop and run rich?:confused: