O2 ?

[C2500]

I’m new here and to TBI tuning so bare with me. After reading threw the O2 placement that EagleMark posted I think I need to move my O2.. I have a 94 c2500 pickup with a 383 in it that I just completed putting in the motor.
I have been dealing with ED for a chip and its on its way back to him for adjustments..I noticed that when you shut down and restart it goes in open loop and sits there till the O2 starts switching or if you hold the rpm up it goes in closed loop quicker..
I replaced the Y-pipe with a after market and now notice that the O2 bung seems to be longer and not much of the sensor is in the flow path so I was thinking if shaving down the bung to get it in there farther but now thinking of placing it closer to the collector.. I have a set of stainless shorties that bolt up to the original Y-pipe as the stock manifolds kept on cracking (Chinese junk aftermarket) so these have been good so far..But its only has a bung in the left collector is that an issue? I could tig in another in the right side and swap side to side any thoughts? Also I’m going to need a harness extension to relocate it..
After reading Eagle Marks post it should heat it up quicker and go in closed loop quicker right??
Here’s a pic of where the O2 is just before the cat..

[dave w]

I would agree, the current O2 bung should be shorter.

The basic theory of operation for an O2 sensor: needs to be hot to work. There were some factory 4.3 V6 applications that used a 3 wire heated O2, most likely because the exhaust did not get a single wire hot enough to work properly in all driving conditions.

Maybe convert you O2 to a 3 wire? https://www.ebay.com/itm/3-Wire-Heated-Oxygen-Sensor-O2-Wiring-Harness-Adapter-TPI-TBI-HAR-017-/291414899872?_trksid=p2385738.m4383.l4275.c10 and https://www.ebay.com/itm/ACDelco-AFS74-Oxygen-Sensor/292007294395?epid=114900594&hash=item43fcfdb5bb:g:J5wAAOSwD0lUkUyV&vxp=mtr

dave w

[C2500]

Dave it is a 3 wire I should have posted that sorry. But I would still think it should work better if its up stream closer to the engine.

[Terminal_Crazy]

The LT1 $EE PCM has 3 time delay setting and some temp settings that have to pass before going into CL.
As I understand it this is to stop pollution of the BLM tables with flakey data until idle settles down.
I presume your pcm does something similar.

HTH
Mitch

[lionelhutz]

That's the stock location, right? If yes, then it should be fine there. Just shorten that bung so the whole sensor nose is in the pipe. The bung should just be long enough to hold the threads.

Is ED from Harris Performance or TBI Chips? If yes, good luck with that one....

[C2500]

Yes that’s stock location...
No its Ed from fast chips...I have heard to stay away from Harris performance..

[C2500]

Yes that’s stock location...
No its Ed from fast chips...I have heard to stay away from Harris performance..

Well let me retract that this is an aftermarket front pipe with a converter and I’m not sure if its in the same location as original but I do know that the O2 wires are a little taught..
Any one have a pic of an original front pipe for a 94 with a 350 or even a 305 I would think be the same...

[lionelhutz]

It should be OK since the configuration is close to stock. I would check that the heater is working. Try cutting down the sensor bung and see what happens.

[C2500]

I cut a good 3/16 off of the bung and put another bung up from that one that is in more in the flow of the exhaust.. when I get my chip back Ill test it to see if it makes any difference when its hot to go into CL..

[57 Handyman]

Question: When using and tuning with a WB02, do you install the WB02 in place of the stock O2 sensor or is it necessary to install another bung in close proximity to the stock sensor location?

[dave w]

Question: When using and tuning with a WB02, do you install the WB02 in place of the stock O2 sensor or is it necessary to install another bung in close proximity to the stock sensor location?

Two Oxygen Sensor Bungs, one for the original factory O2 sensor and a second one for the WBO2.

dave w

[C2500]

Correct me if this is not right>TBI uses a NB O2 so its rich or lean so if a WB O2 is in the same area of the exhaust that the NB is with a separate controller like a LM 1 what would your advantage be doing that? Maybe answer my own question but a NB only switch’s from Lean to rich? And a WB reads true AFR..New to this so bare with me.

Also how can I check the ground and power wire on the O2. It’s got 2 white wires and one black. Thanks

[lionelhutz]

A wideband measures lambda, not AFR. You use it to confirm that operation modes like WOT are correct since the NB can't read the fuel enrichment.

The 2 white wires should be the heater element and the black would be the sensor output.

[C2500]

Ok got it so one of the white wires be ground then?

[lionelhutz]

Yes, one wire will be power and the other will be ground.

[C2500]

Is the 4 wire O2 just have an extra chassis ground? So its not using the exhaust to ground the sensor?

[lionelhutz]

The 4-wire has a sensor element negative or return wire that runs directly to the ECU so that it doesn't rely on the exhaust as the signal path.

[C2500]

Ok so where does the 3 wire O2 ground? ECM??

[myburb]

So far as I know it grounds the same as a one wire. case,exhaust,motor and then where ever the ecm grounds, usually the block.

[lionelhutz]

Both have a connection between the negative of the element and the case of the sensor.

A 3-wire used the case for the signal path. The case connects to the exhaust, the exhaust connects to the head, the head connects to the intake and then the intake is connected to the ECM via a wire.

The 4-wire setup runs the wire that went to the intake directly to the sensor instead. It eliminates the possibility of a bad electrical connection in the exhaust pipes from causing false O2 readings.

[C2500]

What’s the best way to add a ground to the 3 wire O2?? Ground the exhaust somehow to the frame?

[lionelhutz]

Change it to a 4-wire sensor. Then, find the negative sensor input wire and remove it from the engine ground and take it to the 4-wire sensor plug instead.

It would be pointless connecting the exhaust to the frame, since the frame isn't directly connected to the ECU.

[daveosx]

Guys over thinking

one wire O2 goes to ECM

Three wire O2 one goes to ECM black or tan or purple
Two white wires one goes to hot in run 12V+ the other to engine ground

Four wire O2 one goes to ECM black or tan or purple or yellow
The other black or tan or purple or yellow goes to ground wire at back of head or on corvette the intake manifold. The ECM return path is a pass through that goes to the head on LT-1 and Corvette LT-1/lt-4 it goes to the intake next to distributor. On TBI the same circuit goes to back of head on Crossfire TBI it goes to the intake manifold.
The power wires one is grounded to engine block the other is hot in run 12V+

Besides I am thinking he needs a fuel pump.

The only reason for the difference on the Corvettes was the fear of electrolysis at the head gasket he first year we ran aluminum heads.
The corrosion engineer was a little paranoid about having constant current through the head to the block even though his fix did the same thing it was next to a water jacket and he thought it was better. On production the noise engineer moved the O2 return off the engine and ran it to the ECM 1992 in 1991 the loom goes to the head.

[C2500]

Thanks Dave!!

[lionelhutz]

Why would you take the 4-wire sensor element ground or negative wire to the engine block? Since It can be connected directly to the ECM O2 input circuit, it should be. This is a separate wire from any other ECM ground wires, typically tan colored. If you don't re-wire to connect the 2 together then you might as well just use a 3-wire sensor.

[daveosx]

Yep on older cars the provision for sensor return from the O2 is not present.
On the ECM circuit board all the grounds are tied together and tied to the case through the bolts.
On the next group of ECMS the grounds were separated to two groups sensor return and chassis.
These circuits are C429 C452 --- C429 is the TPS Coolant sensor Ground and C452 was all other sensors ground.
Starting in 1981 they moved the EST and HEI modules to C486.
In 1991 they started using the O2 sensor return.

On SBC GEN I in all GM vehicles there was one ground point on the rear of the cylinder head.
On the next series GEN 1.5 (one Piece Rear Main) GEN 2 there was the head ground and the Intake manifold ground.
On Later models there was the head, Intake and battery ground.

Internals of the sheet metal GM ECMs all had a common chassis buss and intake buss some had a ground lead for HUD dash.
The cast aluminum ECM cased boards actually had better noise control on the electronics.
Some ECM circuits used a difference amplifier (http://www.electronics-tutorials.ws/opamp/opamp_5.html)to increase O2 sensor signal reliability it was only then in 1996 that the GM 4 wire sensor could not be substituted with a three wire.
pre 1996 four wire sensors can be substituted with a three wire.

But as far as sensor return lines on pre 1996 GM vehicles the return wires are all tied to ground just some do not share the same physical ground path.
There is an effect called loop induced current that will induce a voltage on a intended shield line or a secondary ground loop.
Electrical engineers at GM had seniority and the electronic engineers worked under them.
The radio engineers worked for AC Delco not GM proper.
So where I or any modern electronics engineer or technician would use shields and ground distribution blocks in a car they did not.

Two areas that will benefit from using shielded cable
The Reference line from a HEI7-8 to the ECM with the distributor side ground only to the shield
Knock sensor with the EST module side of the shield grounded.

pre 1996 The rest of the returns and other grounds will function fine if grounded to the same point.

[lionelhutz]

Yep on older cars the provision for sensor return from the O2 is not present.
On the ECM circuit board all the grounds are tied together and tied to the case through the bolts.
On the next group of ECMS the grounds were separated to two groups sensor return and chassis.


Every ECM I have ever looked at (for example '7749 or '7747 from 1987 or newer) has an O2 return terminal on the PCM and a separate dedicated wire from this terminal to the block. If I'm recalling right, it is typically a tan wire. Every one of these ECUs has had a special O2 sensor IC for the O2 sensor. To switch to a 4-wire sensor you disconnect this wire from the block and wire it to the O2 sensor instead.

I have never seen a ECU that uses a common O2 negative combined with the other sensor negative wires. I have never seen one with combined ECU grounds and sensor grounds. The ground circuits are typically black with black/white for sensor negative. Sensor negatives are never grounded to the engine or chassis.

Maybe something quite old isn't this way, but it must only be with a bit of the early stuff.

[daveosx]

The harness for the 7747 I am familiar with.
The black and white wire can be either circuit 451 or 450 on all GM up to 95 on cars 99 on trucks.
circuit 451 is serial data Assembly line link data second pin top row from left of OBDI connector.
The tan wire is circuit 413 called O2 sensor low it is just another ground at G161 it is combined with 450 there. On the back of drivers side head top bolt all GM.
Disconnecting this and connecting to the 4th wire will make the O2 sense extremely noisy the output from the Operational Amplifier attached at 412 before the ADC amplifies the signal from a scale of .1 to .9 mapped to 0 to 5V this 5 volt signal is referenced to that ground at 413 and the common 450 ground level the following stage is the ADC that provides a 7 bit resolution.
Circuit 450 wires are tied together at splice S450 in the harness behind the distributor the ground wire is attached on non corvettes at the passenger side head rear top bolt on corvettes this is attached at the intake manifold. This is the location of G160.
Circuit 450 is a ground the ECM MOTHER BOARD ties all of the returns together internally there is NO differential amplifier on the O2 sensor on any of the GM sheet metal ECMs.

With a meter you can verify this for yourself unplug the ECM touch the negative lead to pin 13 of the 24 pin connector and continuity test pins 3, 6, 11, 12 on the 24 pin and
3,11 and 15 on the 22 pin connector.

Checking a GM harness at the same plugs and pins by the Service Bulletin the OHM reading was less than 1 ohm or repair harness at G160 and G161 if the repair did not provide result than the splice at S450 was inspected and resoldered. If that provided no result than the whole harness was replaced.
I replaced 4 corvette harnesses in 1991 due to corroded 450 wires.

There is also circuit 452 black wire at coolant sensor and TPS that is grounded in the ECM.

So on cars or truck with a single wire purple circuit 412 you can use a three wire black to circuit 412 or you can use a four wire one side of sensor to 412 the other to ground.

On Cars and trucks with sheet metal cased ECMs there is no differential amplifier on the 02 sensor the input circuit to the controller is just a comparator set at .7 V and an ADC that provides basic diagnostics by giving a 7bit digital approximation of the analog signal,

On 91 Lumina Corvette Cadilac and Some models of Fbodies there was a change to the GM66ECU with the cast aluminum case up to 1995 they were called OBD1.5 because they are not compliant with OBDII connectors or function.
The same ECMS are used in 1996 with different software to enable OBDII.
On these ECMS there is a differential amplifier connected to the O2 sensor and the ADC counts are used for closed loop operation.
So on cars with the cast aluminum ECM there are two four wire O2 sensor that has 412 413 terminated at the front left connector 414 415 at the front right
In 1996 the numbering scheme also changed and to post cat O2 sensors were added.

On 1995 Bodies and Corvette the was a optional performance switch that changed the spark and fuel maps on the fly some ECMs have the switch pined out on Caprice Impala ECMs the 3 gear selector switch triggers the second map. There is also a highway mode stored in the State trooper cars providing a third fuel and spark map location if I think of it later today I will post the part number for the three way ECM.

Thought you might like this
12488

[C2500]

Dave do you have a pin out pic like above for a 6395 eco that’s in my truck? Or where can I find one..
Thanks

[lionelhutz]

The tan wire is circuit 413 called O2 sensor low it is just another ground at G161 it is combined with 450 there. On the back of drivers side head top bolt all GM.
Disconnecting this and connecting to the 4th wire will make the O2 sense extremely noisy

The tan is a completely different circuit than the other grounds. On the GM's I have worked on it's a separate wire to a separate ring terminal attached to the engine. The ECM terminal connects to a lambda sensor interface amplifier chip. Here is an example of this chip, different pinout than the '7747 ECM but the same pins and same functionality.

http://www.ti.com/lit/ds/symlink/lm9044.pdf

4-wire O2 sensors for use in GM's have the negative of the sensor element grounded to the case. I just double checked this and confirmed it is true due to another thread here. So, the wire is not left floating when it runs directly to the sensor.

I have no idea what pins you're describing on the '7747 ECM??? The pins go A1-A12, B1-B12, C1-C16 and D1-D16.



On Cars and trucks with sheet metal cased ECMs there is no differential amplifier on the 02 sensor the input circuit to the controller is just a comparator set at .7 V and an ADC that provides basic diagnostics by giving a 7bit digital approximation of the analog signal,

It definitely is not a hard wired comparator in the '7747. You can most definitely set the O2 switching voltage threshold limits.

[lionelhutz]

Dave do you have a pin out pic like above for a 6395 eco that’s in my truck? Or where can I find one..
Thanks

It should be $0E.

http://www.gearhead-efi.com/Fuel-Injection/showthread.php?337-16196395-PCM-Information-OE

[daveosx]

Yes the Tan wire circuit 413 is a separate wire that is GROUNDED at G160 along with Circuit 450 Sensor return ECM ground
In the ECM the pin that 413 is attached too is almost always grounded at the ECM buss as well.

The lambda amplifier is an Operational amplifier and or comparator the main difference between the differential amplifier and the Operational amplifier is the ground reference of the OP amp.

Operational Amplifer (http://www.electronics-tutorials.ws/opamp/opamp_1.html)
Comparator (http://www.electronics-tutorials.ws/opamp/op-amp-comparator.html)
Differential Amplifier (http://www.electronics-tutorials.ws/opamp/opamp_5.html)

Prior models use the same circuit as the National instruments matlab adapter for O2 sensors.
O2_Sensor_UM_RevD.pdf (http://www.ni.com/pdf/manuals/O2_Sensor_UM_RevD.pdf)
As you can see the manual for the engineering and design tool used for the creation of the ECM shows how the one lead of the sensor is tied to ground through the EGO- buss. This is the same setup used up to 1993 on the sheet metal cased ECMs.
The only reason I say 1993 is thats when I retired from GM.

Here is the Chip or similar used most commonly from 1995 on (1991 on some models)
LM9044 (http://www.ti.com/lit/ds/symlink/lm9044.pdf)

On page one under Typical Application you see the diagram of the chip internals it is a differential amplifier followed by ground referenced OP amp.
This is the stye used from 1995 on when the chips where first made available.


The whole thread was about how to wire the different sensors to a system with a one wire sensor it is simple you ground the 4th wire.

[daveosx]

It should be $0E.

http://www.gearhead-efi.com/Fuel-Injection/showthread.php?337-16196395-PCM-Information-OE

Cool thanks I did not have these
Added to library of old stuff.

Does anyone have the factory diagnostic wiring diagrams with the connector and splice locations for the C1500 Suburban?

[lionelhutz]

Yes the Tan wire circuit 413 is a separate wire that is GROUNDED at G160 along with Circuit 450 Sensor return ECM ground
In the ECM the pin that 413 is attached too is almost always grounded at the ECM buss as well.

Why is a separate wire needed if it's just grounded to the other ground pins?



As you can see the manual for the engineering and design tool used for the creation of the ECM...

Where can I see the original design documents used for the '7747 ECM?



Here is the Chip or similar used most commonly from 1995 on (1991 on some models)
LM9044 (http://www.ti.com/lit/ds/symlink/lm9044.pdf)

On page one under Typical Application you see the diagram of the chip internals it is a differential amplifier followed by ground referenced OP amp.
This is the stye used from 1995 on when the chips where first made available.

That chip still has the same pin connections as the chip for the O2 in the '7747. The '7747 chip has different pin numbers on the package, because it is DIP, not PLCC.

What are you meaning with the internal description? A differential amplifier allows both input pins to float from the ground and it will still output the difference.The stage after is internal in the IC so it doesn't need to be differential. Right from the datasheet the IC will work with up to -3V on the O2 input pins.

Did that style of chip become available in 91 or 95? You're contradicting yourself.

[daveosx]

Why is a separate wire needed if it's just grounded to the other ground pins?




Where can I see the original design documents used for the '7747 ECM?




That chip still has the same pin connections as the chip for the O2 in the '7747. The '7747 chip has different pin numbers on the package, because it is DIP, not PLCC.

What are you meaning with the internal description? A differential amplifier allows both input pins to float from the ground and it will still output the difference.The stage after is internal in the IC so it doesn't need to be differential. Right from the datasheet the IC will work with up to -3V on the O2 input pins.

Did that style of chip become available in 91 or 95? You're contradicting yourself.

The tan wire was a design intent to use a differential amp on the earlier sensors but the costs associated with O2 sensors that far back and the need to economize on wiring and parts kind of won out. On the early 1970s and 1980s controllers noise was a huge issue. Lambda feed back was kind of a novel way to accomplish feedback. But putting shielded cable in a harness was too costly. Later top end cars did get shielded cable when needed but the lower end cars had different solutions. Hense the single wire O2 sensors from Bosch.

I have no idea where to get design notes. maybe one of the other engineers still has some stuff left over.
The guy who did most of the digital dash for vettes has shared the original dash schematics.
I have a copy if you would like them.

If it has a lambda amplifier in a different package (DIP vs PLCC) then thats the reason but I do not believe that the O2 amp is a lambda differential amp on the sheet metal ECMs.
On the cast aluminum ECMs there is a differential amplifier.
On early models the amp is discrete on later it is integrated.

The early ones 91 to 95 are discrete the Lambda chip came out in 1995 and was used from 1996 model year on.

No contradiction

When thinking about this and understanding how a differential sensor and amplifier works the two input lines are held equally between ground and VCC+.
The output of an old style Operational amp has the ability to change from about .7V above ground level to about .7 v below VCC+ So in typical use they where run with -15V VSS and +15V VCC. In a car you have start sag voltage of 9V and full charge voltage of 14.8V no -Vss available. SO you use the Operational Amps that function on single supply and confine the voltage to a working VCC usually 5V but I have seen Nissan and BMW controllers at 6 and 9 V VCC.

O2 sensor voltage varies from .35V to .9V a differential amplifier built from a single supply operational amp will have an output proportional to the difference between the two wires of the sensor. An operational amplifier with a ground reference will have an output proportional to the difference between the ground and the input signal.
A lambda amplifier uses a differential amp feed into a Operational amp providing output proportional to the difference between the ground and the differential amp signal.

So if you use a four wire sensor on a system with a grounded TAN 413 wire just ground two of the O2 sensors wires run the heater wire to 12V in run and the other wire to 412 purple wire.

The reason you would not just ground one input of a differential amp is that there is a bias circuit used to float the two inputs equal between ground and VCC this will become a current path if one input is grounded. (Dead Battery)

if you wanted to make a differential lambda circuit you can use an LM358 single supply OP amp 4 20K resistors 2 1k 2 10k and have a four wire O2 sensor amplifier that will give you 3.5V to 9V output when used with 12V VCC and ground on VSS.


12492

[lionelhutz]

I would bet money the PCM in this thread uses a lambda IC and can be hooked to a 4-wire sensor as I suggested.

The negative side of the sensor element in a 4-wire sensor is still tied to the case of the O2 sensor. I just measured this with a meter and the negative sensor pin IS grounded to the case. Look at the typical application diagram of the TI datasheet and it shows this ground connection at the sensor end too. So, no need for any other ground on the sensor circuit. No need to worry about noise from an ungrounded sensor either.

You posted "Here is the Chip or similar used most commonly from 1995 on (1991 on some models)" and then posted that lambda chips came out in 95. Definitely contradicting.

You posted "as you can see" about some kind of engineering and design tool manual used to create the ECU, so I was assuming you missed a link to some documents because none of the links you posted had anything to do with designing the old ECU's.



The reason you would not just ground one input of a differential amp is that there is a bias circuit used to float the two inputs equal between ground and VCC this will become a current path if one input is grounded.


Sure, if that was the circuit. But those lambda IC's do not have such a bias circuit. The datasheet even recommends grounding the sensor negative lead at the O2 sensor. There are biasing current sources on the pins, but they are not doing as you suggest. They are to provide a default voltage output if the wiring to the sensor opens.

I have seen the schematics of the '7746 ECU, which is basically the same boards as the '7747. Just for you, I'll look it up and post the links.

Input schematic, O2 sensor is bottom right

https://pcmhacking.net/ludis/1228746sheet3.gif

The lambda IC.

https://pcmhacking.net/ludis/c3pinouts.html#16007757

It's the same IC circuit but in DIP case. The DIP version was probably long discontinued before TI bought National so TI did not bother with posting datasheets of it. Yes, TI bought National maybe 8 or 10 years ago.

The 1995 date on the TI datasheet is likely the date the PLCC version of that lambda IC was released.

If the '7747 era ECU's had the lambda IC then this later model PCM definitely has it.

[daveosx]

If an earlier model 1991 - 1995 uses a differential amplifier for the input of the O2 than it is a similar circuit to the lambda chip that came out in 1995 and was used from 1995 on. This is not a contradiction.

[daveosx]

Looking at your references I see that yes it is a similar chip.
What GM cars shipped in 1991-1995 with a four wire and sheet metal ECU cases
My knowledge was that only the aluminum cast cases had differential amps.

And as I stated grounding one side off a differential amplifier input would create a path to ground.
http://www.ti.com/lit/ds/symlink/lm9044.pdf
Page 7 block diagram shows
two current sources one 65nA the other 380nA at the -Vin on this specific chip when -Vin is grounded that would be a current path to ground.
Most likely the source is switched not battery.

[daveosx]

http://www.ti.com/lit/ds/symlink/lm311.pdf

http://www.ti.com/lit/ds/symlink/ina105.pdf

Are two that I know of being used

[daveosx]

Not seeing a four wire with any internal grounds.
12505

What O2 sensor are you using?
It might be a three or four wire replacement maybe IDK.

[daveosx]

Anyway the whole point to my posting was that if the 413 lead is grounded then you can just ground the other sensor wire.

I think the whole reason for going to the 4 wire sensor with isolated sensing element was to eliminate the galvanic and induced voltages that are present between ECM ground and the bung on the exhaust pipe. In a modern engine you have many different metals and induced currents and resistances.
.7 volt is a very small voltage that is easily lost in the other noise.

Using a 2 wire shielded cable grounded at the ECM would have been the best solution.

[lionelhutz]

The sensing element in the O2 sensor is NOT isolated!!!

http://www.ti.com/lit/ds/symlink/lm9044.pdf

Page 1 clearly shows grounding the negative side at the sensor. The sensors have the connection. I have tested it. Do I need to take a picture of the ohm meter reading to prove the connection exists in a sensor?

Of course grounding the negative lead creates a path to ground. The lead is connected to ground for crying out loud.

But that should not be confused with claiming it screws up the chip, because the chip is designed with the expectation that the negative lead will be grounded.

The use of a 1, 3 or 4 wire O2 is rather irrelevant to the IC used inside the PCM. How many cast case PCM's with the lambda IC didn't use a 4-wire O2?

If you are going to use a 4-wire O2 then why leave the wiring with 2 separate connections to the engine or transmission or exhaust? Just avoid any possible connection issues and connect the O2 sensor wires right to the ECU/PCM and be done with it.

If you are going to leave connections to the engine block then you might as well stick with a 3-wire sensor.

[daveosx]

Sure, if that was the circuit. But those lambda IC's do not have such a bias circuit. The datasheet even recommends grounding the sensor negative lead at the O2 sensor. There are biasing current sources on the pins, but they are not doing as you suggest. They are to provide a default voltage output if the wiring to the sensor opens.

Actually some Lambda chips have another op amp that monitors for open.

I think that it is possible that the internals of a OP amp in discreet components and the tri angle symbol are hiding the current path I am speaking of.

On page 10 of the LM9044 data sheet there is a simplified internal diagram.
The V-in terminal is biased through a 10Kohm resistor through three transistors to VCC the ground side bias path is 10K through collector to base (weird) through 7.6K and 3.4k adjustable.

An open circuit would provide Vref through 400 200 Vout

[daveosx]

The sensing element in the O2 sensor is NOT isolated!!!

http://www.ti.com/lit/ds/symlink/lm9044.pdf

Page 1 clearly shows grounding the negative side at the sensor. The sensors have the connection. I have tested it. Do I need to take a picture of the ohm meter reading to prove the connection exists in a sensor?

Of course grounding the negative lead creates a path to ground. The lead is connected to ground for crying out loud.

But that should not be confused with claiming it screws up the chip, because the chip is designed with the expectation that the negative lead will be grounded.

The use of a 1, 3 or 4 wire O2 is rather irrelevant to the IC used inside the PCM. How many cast case PCM's with the lambda IC didn't use a 4-wire O2?

If you are going to use a 4-wire O2 then why leave the wiring with 2 separate connections to the engine or transmission or exhaust? Just avoid any possible connection issues and connect the O2 sensor wires right to the ECU/PCM and be done with it.

If you are going to leave connections to the engine block then you might as well stick with a 3-wire sensor.

Great discussion by the way making me pull out old books and links to chips.

Yes I see that the application note shows a ground in the sensor.

I disagree about the grounding is intended by the chip design not likely.

Thats interesting that your four wire could be used in a three wire setup could be a manufacturer cost savings thing.

Here is the diagram for three and four wire narrowband sensors.
12506

and another authoritative source on wideband sensors
https://emsd.weebly.com/wide-band-o2-sensor.html


I do not see the indication of a ground internal on the four wire of either.

As far as why wire to the block instead of the ECM ?
The cylinder head of either bank is typically less than 2 feet from the O2 bung and the ECM is typically on the inside of the car.

[daveosx]

To answer why the notes for electronic fuel injection controllers are not easily referenced.
Take a look at the D-Jetronic prior to O2 sensor inclusion.
125071250812509

[lionelhutz]

As far as why wire to the block instead of the ECM ?
The cylinder head of either bank is typically less than 2 feet from the O2 bung and the ECM is typically on the inside of the car.


You're hardly doing any extra work for the better wiring. The wire already comes from the ECU to the block. So, just disconnect it from the block and extend it to the O2 sensor.