If one wanted to lose the stock 2-into-1-into-2 exhaust system on the 2.8 in favor of a true dual, how should the O2 sensor be handled?
Would it be a simple matter of running two separate sensors, and wiring them in parallel to the existing O2 line to the ECM, or would an intermediate circuit be required to average the two voltages?
------------------
his: 1985 2M6 SE ½ poly'd, ½ Koni'd, formula nose & factory sub
hers: 1984 2M4 all stock... so far
[This message has been edited by dguy (edited 01-27-2004).]
IP: Logged
10:27 AM
PFF
System Bot
dguy Member
Posts: 2416 From: Beckwith Township, ON, Canada Registered: Jan 2003
I'd put a cross-over between the two down-pipes and stick the O2 sensor in there and get an average reading from both cylinder banks. If the engine's older it might help the computer with the A/F ratio if the two cylinder banks wear differently.
A cross-over almost always helps with low-end torque production too.
YMMV
------------------ Smokin Joe Fission 1986 Fiero SE 2.8 Perhaps I shouldn't have taken on a Fiero that the original owner said he liked to work on, especially after it spent some time 'submerged'. At least the car was free.
If you put the O2 sensor in only one side then the engine mixture will follow that side only. For instance, say a plug misfires on that side, since the O2 sensor will see a greater percentace of richness ( 1 of 3 cylinders ) instead of one of six cylinders. It may exagerate minor tuning problems, and it may be more susceptable to going into open loop with only three cylinders to keep the sensor heated up.
JazzMan
IP: Logged
11:23 PM
Jan 28th, 2004
dguy Member
Posts: 2416 From: Beckwith Township, ON, Canada Registered: Jan 2003
Well JazzMan & SmokinJoeFission hit my concern right on the head.
In the mythical perfect world, each cylinder bank (not to mention each cylinder) would perform identically. But we're not living in a myth...
Placing a single O2 sensor in a cross-over pipe is likely the easy way out with respect to the electronics; it instead shifts the challenge over to how to fit a x-over in there. Also given the distance between the O2 sensor and the manifolds in the stock system, I expect if a similar distance is maintained at the x-over, there should be few concerns about getting stuck in open-loop. Theoretically these concerns could be minimized even further with the judicial application of heat wrap & its variants.
In a dual system with no x-over, I imagine the best placement for the O2 sensor(s) would be as close to the manifolds as possible. Again, with heat wrap to minimize any concerns about them not heating to their working temperature. Dusty old high school electronics knowledge tells me that if the output from dual O2 sensors was simply tied together and fed to the ECM, the ECM would see the higher of the two voltages (assuming that the readings from each bank were different). Any knowledgeable people care to confirm or at least set me straight here?
Assuming that the above is true, I suspect that in this scenario some sort of intermediate circuit should be used to provide the ECM with the average value of the two sensors. After all the original single-sensor is picking up the combined exhaust gasses from both banks, and feeding a voltage to the ECM based on the combination of all. Hmmm...
Don't mind me, I'm thinking out loud and using everyone who's reading this to keep my twisted little brain in line.
Does anyone know of any engines which (in the stock configuration) have dual O2 sensors? I might be able to get a better idea of where to go with this if I could look up a few wiring diagrams.
-d.
IP: Logged
09:03 AM
HitesFiero Member
Posts: 401 From: St. Charles, MO, USA Registered: Sep 2003
since it's OBD1 and using a throttle body and not tuned port the yes just pick a side.
The Fiero V6 IS a port injected engine, not throttle body injected. Throttle body injection places the injectors above the throttle body valves. A port injected engine has an injector at the base of each intake runner. I'm not trying to be harsh, I just wanted to clairify.
------------------ Don Hites 88 GT, Getrag 5spd, 4.9 v8 with a Delta cam and other heavy mods. 88 Coupe 5spd Duke (now under restoration)
[This message has been edited by HitesFiero (edited 01-28-2004).]
IP: Logged
09:17 AM
dguy Member
Posts: 2416 From: Beckwith Township, ON, Canada Registered: Jan 2003
For those who are curious, I found the circuit I was looking for. It can't get much simpler.
The circuit below is an example used to average the voltages of three different batteries, but works just as well if you replace the batteries with O2 sensors (an O2 sensor is just another voltage source after all).
The closer R1 & R2 are in value to each other, the truer the voltage across the circuit will be to the true average. Personally, I'd replace the fixed resistors with two fine-wound POTs and adjust them until they are exactly the same rather than fart around with finding two 5% tolerance resistors which are "close". But that's just me.
[This message has been edited by dguy (edited 01-29-2004).]
IP: Logged
09:05 AM
PFF
System Bot
dguy Member
Posts: 2416 From: Beckwith Township, ON, Canada Registered: Jan 2003
You may run into a problem because the O2 sensors are actually voltage generators, with 500 mV being the point where the mixture is "perfect". Also, the ECM apparantly applies voltage to the sensors, but basically if you add resistors you shift the voltage the ECM sees and that will drive the mixture out. Not sure how to handle that one.
You may run into a problem because the O2 sensors are actually voltage generators, with 500 mV being the point where the mixture is "perfect".
The batteries used in the sample circuit above are also voltage generators.
If used as a 2-into-1 O2 sensor converter, I suspect it would work as:
one "side" of the O2 sensor is grounded, similar to the how the voltage sources in the sample are grounded.
the ECM will be reading the voltage potential across the circuit with respect to ground, much as the connection of the voltmeter is in the sample.
quote
Also, the ECM apparantly applies voltage to the sensors, but basically if you add resistors you shift the voltage the ECM sees and that will drive the mixture out. Not sure how to handle that one.
Ok this part I didn't know about. Interesting. The immediate comment which jumps to mind is "weird", because the O2 sensor works by generating a voltage. Applying a voltage to a sensor which is providing a voltage just sounds... goofy. But what do I know?
I think what I may do at this point, is a bit of an experiment. I have an old L6 in the garage which I've been meaning to place on a "runnable" stand for some time now. If I can get that taken care of some time over the summer, it would be intersting to place a couple of O2 sensor bungs in its exhaust, hook up a pair of DVMs, and take some readings both with and without the averaging circuit tied to them.
Things to ponder...
-d.
IP: Logged
08:33 AM
dguy Member
Posts: 2416 From: Beckwith Township, ON, Canada Registered: Jan 2003
you dont need true duals the fiero exhaust is free flowing as it is.
Well not to steer my own thread off-course, but have you looked at a) the manifolds, and b) the y-pipe? There are obvious restrictions in each, and you don't need an engineering degree to identify them.
Whether I need it or not is irrelevant. Many people on here have modded in their cars in ways which they really do not need, but have done it anyway. I'm simply exploring the feasability of a modification which I find interesting.
IP: Logged
08:41 AM
Paul Taylor Member
Posts: 383 From: Reading, Berkshire, ENGLAND Registered: Oct 2002
Dusty old high school electronics knowledge tells me that if the output from dual O2 sensors was simply tied together and fed to the ECM, the ECM would see the higher of the two voltages (assuming that the readings from each bank were different). Any knowledgeable people care to confirm or at least set me straight here?
When you put two things on a battery, things wired in series have the same current, but different voltages and things wired in parallel have the same voltage, but different currents.
If you are combining two sources of power together and you wire them in parallel you will increase the voltage (1v + 3v = 4v).
My question would be this: would you have to cut the total current in half to get the average? After all, if both banks are running perfectly, each O2 sensor will be putting out 500mv or whatever, making a combined total of 1v.
Also given the distance between the O2 sensor and the manifolds in the stock system, I expect if a similar distance is maintained at the x-over, there should be few concerns about getting stuck in open-loop. Theoretically these concerns could be minimized even further with the judicial application of heat wrap & its variants.
But what do I know? 
Well not to steer my own thread off-course, but have you looked at a) the manifolds, and b) the y-pipe? There are obvious restrictions in each, and you don't need an engineering degree to identify them.
