This time I'm asking about the shielded wiring for the oxygen sensors.

The O2 sensors have a shielding that goes to an engine ground. However, the ground branches off and goes to ECU pin 23, named "O2 Sensor Shield".

What does the ECU do with this ground signal?

Here's why I ask:

I am getting a high, fixed reading at the feedback monitor sockets, 3.8 volts. Same voltage with/without the disable plug installed, in "P" or in "D", with engine at idle or at 2000 rpm. Fixed reading; never changes. Swapped my spare ECU....same thing, only 4.0 volts.

The O2 sensors are new and I know they are doing 'something'...because the engine immediately becomes a bit rough when I disconnect either one of them....and becomes smooth again as soon as they a re-connected. The engine runs grrrreat, by the way. I'm am very pleased.

Meanwhile, while mulling and musing and smoking my pipe, I suddenly remembered that I have a stray, disconnected black wire with an eyelet at the end. It was a loose end from the conversion process that I "meant to get back to" but totally forgot.....until 20 minutes ago. I feel sure this must be a ground wire....perhaps related to the O2 sensor shield ground? If so, perhaps a solution to the mystery of fixed feedback readings?

We'll find out tomorrow.

Still, though, the original question remains: what does the ECU do with the ground on pin 23?

Cheers and thanks
Doug

 

I'm assuming this is a V12? If so, Lucas ignition?

Assuming so, the ground for the shield is simply to make sure that ground is ground. What can happen is that if things are grounded at different points, it is possible for there to be small voltage differences between the different ground points, either due to voltages being induced in the wiring by other components, or by less than ideal grounds. Therefore if things reference off "ground" that can be a different voltage at different places, and thus give different readings.

The idea of grounding the shield both to the ECU and the Engine (as shown in the S57 guide) is that both reference the same ground.

On aftermarket EFI systems like Megasquirt it's emphasised that all the sensors, such as O2, coolant temp, air temp and throttle position must all be grounded together at a common ground point to prevent false readings. They make a distinction between sensor ground and the usual ground of everything else ( chassis ground).

 

Makes perfect sense, thank you


Cheers
DD

 

My added thanks for this concise explanation of the ground functions. More on electrckery. At times, I think I get it and at others, I realize I don't.


Added bits:


1. My conversion includes a bank of relays. I created a ground bus along side. Each ground of the relay, sans one goes to this bus. It in turn is connected at the wing wall where the battery - cable is attached. The one exception Is the starter. It functions a bit differently.


2. A woven shield should be grounded at one end only. Although counter intuitive, grounding each end defeats the shield!


3. In my "perfect" electric world, chassis grounds would not exist. A separate grou d path would be created for all grounds. And the loads insulated.




Carl

 

Well, I ground the mystery wire and no change.

However, I did find out that I have no 12v supply to the O2 sensor heaters. I'm not convinced that this has anything to do with the fixed feedback readings, but I'll trace and repair faults as I find them.

Cheers
DD

 

That should bring the O2 sensors into being active more quickly after startup, but shouldn't affect them too much. Although idling might not produce enough heat to make them 100% effective. Worth fixing anyway.

 

All sensors shields should only be grounded at ONE end, if you ground at both ends it can cause current to flow in the ground shield in 2 ways, induced by the AC signal in the centre conductor and DC current due to a slight difference in resistance in both grounding points.

Narrow band O2 sensors produce a voltage between 0.15 and 0.9volts. 0.9volts is rich and 0.15volts is lean, you need a scope so see these voltages, they are too fast to see on a DVM.

That black wire will be the heater ground and is totally separate from the signal ground.

Unplug the sensor and measure the resistance of the sensor element it should be between 6 and 20 ohms. Check the 12v supply to the heater. But as you are getting 4volts which is close to the ECU's reference voltage of 5volts, I suspect the sensor element is high resistance.

The ECU monitors the current draw and this is controlled by resistance of the sensor element, which inturn produces the voltage.

 

Update

Problem not solved but I might be getting closer

An apparently poor connection at the fuel pump relay cured the 'no voltage to O2 sensor heaters' problem...I simply removed and re-seated the relay..... but, not surprisingly, it didn't solve the problem of a fixed feedback reading.

However, I was noodling around, just sorta going thru some basics, and learned that my TPS is also at a fixed reading, .98 volts, regardless of throttle position. As we know, it should be about .32v at idle.

Hmmm. Other than knowing that another fault needed to be repaired I wasn't sure if this had any bearing on my problem. Until a vague memory trigged a bit of research.

From Roger Bywater:

"The throttle switch of D Jetronic was replaced by a potentiometer so the position of the throttle was converted to a voltage ranging from 0.32 volts closed, activating the adjustable idle circuit, to about 4.5 volts fully open. "


"...activating the adjustable idle circuit...." being the part that piqued my interest.

Could it be that the too-high TPS voltage is preventing the feedback monitor circuit from operating?

I have another TPS and will report back after installing it

Cheers
DD

 

Sounds logical, Doug . . . coincidence that I had just been re-reading your old thread on converting Ford TPS sensor to fit the Jaguar drive . . . I'm very interested in outcome to your current issue, as I've not seen this one to date.

Good reminder Warren . . . in audio, we called the result "hum" . . . cuz it did! Rather than messing with our ears, I wonder what such earth loops could produce in a Jaguar ECU?

Cheers

Ken

 

I have a new Mustang TPS 'in hand'. We'll see what happens. it worked well on my other V12 years ago although success in this case also means getting the voltage down to .32v....which might be a bugaboo.




Cheers
DD

 

Doug, I noticed on that earlier thread how some very careful grinding of the cw end of segment with a mini cutoff wheel managed to lower your .47V (IIRC) down to .32V but I was left wondering if similar at other end might also optimize the 4.5V top end also. Very keen to know your thoughts and outcome. Goes without saying, I wish you success in fixing (or is that, unfixing) your fixed feedback.

Best wishes mate,

Ken

 

Doug, if you unplug the TPS what voltage is on the G wire? should be 5 volts if this is incorrect then the problem might be the ECU reference voltage.

 

5v reference voltage is OK!

Cheers
DD

 

Which brings up another question and discussion.....somewhat rambling here as I'm my first cuppa coffee. I'm just tipping my head and letting stuff fall out

Just how precise do these voltage settings *really* have to be?

The type of TPS system we have on our old Jags doesn't seem different than dozens of others out there yet, seemingly, only ours has this very specific .32v idle specification. Others seem perfectly happy with less precision. If you look at troubleshooting/info pages for any other car and these things are spoken of in broader terms; "Voltage with throttle closed should be between .6v and .9v" or similar broad ranges

For just one example, since I'm using an older Mustang TPS, I learned that, with 80s-90s Fords, any voltage from .6v to 1.0v is interpreted by the ECU as "Closed Throttle" and any increase of 2.71v or more is interpreted as "Full Throttle".

Does the Jaguar ECU truly need to see the precision of .32v? And, if so, is this only for the purpose of "activating the adjustable idle circuit"? Or could it also be that, if the idle setting is set at .32v then the voltages throughout the entire range will hence be correct? Sounds plausible, but one never sees voltage specs as a percentage of throttle opening; ie, "At 40% throttle you must see 2.3 volts". The only requirement is a smooth increase/decrease in voltage as the throttle is moved.

incidentally, a variance of +/- 3-5% is generally deemed common and acceptable simply due to production differences in mass-produced throttle position sensors.

Another thing that often comes up is the TPS never giving a full 5.0v feedback to the ECU at WOT. Apparently this is of no concern on any car. It seems that 4.5v is universally recognized as full throttle and efforts to achieve 5.0v are a waste of time. Apparently true for out Jags as well, as even Bywater mentions "....about 4.5 volts fully open".

If I understand correctly, the function of TPS on our oldie Jags is to *momentarily trim* the injector pulse width as the throttle moves. That's it. On modern cars the TPS plays a much more significant role where more precision would be needed.

Cheers
DD

 

On X300's we know it needs to be fairly precise, I can't see Jaguar setting a spec that tight if there wasn't a need for it.

I though the trim was just for the XJ6, but much more important for the V12 since it's a MAP style measuring system?

There is a full throttle microswitch and a full throttle vacuum switch, so I wouldn't have thought TPS was used to indicate full throttle.

 

Some interesting stuff fell out there Doug........

It's puzzled me why Jaguar used full load switches and a TPS, when the TPS would have done the job, I can only assume they did not do much work on the ECU circuitry and programming until they went to the Denso ECU.

There would be a TPS voltage threshold where the ECU shuts the injectors off above a certain RPM close throttle, my guess this is the critical voltage and the TPS needs to adjusted so this is achieved. I know if you just touch the accelerator when slowing the ECU injects fuel and you get that exhaust crackle.

 

Yes, and on many other newer cars as well, especially those with electronically controlled transmissions which are tied in, to one degree or another, with the engine management architecture



Agreed, but some needs are more critical than others



That makes sense

Perhaps the TPS reaching +/- 4.5 volts triggers closed loop operation? But I think the full load microswitch does that as well?

Cheers and thanks
DD

 

From Bywater:

"Over-run cut off above1500 r.p.m. was activated by a throttle closed signal and a simultaneous vacuum in excess of 21" Hg, but was disabled until fully warm. Once activated the cut off applied until the vacuum fell to about 10" Hg or speed dropped below 1100 r.p.m. Cut off in Lambda sensing versions was active over a much narrower range from 18" back to 15" Hg but in both cases the vacuum effect varied with speed. Quite why this vacuum requirement was thought necessary is a mystery, and it was abandoned for the later 16CU which relied simply on speed and the throttle-closed signal from the potentiometer"


So, perhaps closed throttle isn't recognized until/unless the ECU see .32v from the TPS?

Due to the extra injector pulse whenever the throttle is opened?

Cheers
DD

 

The S58 states the TPS is used to

A) "....determine the position of the throttle for correct fueling...." (not very specific, IMHO)

B) "....the rate of opening for correct enrichment".

Elsewhere in the text the TPS is described as a 'correction input' rather than a 'primary input'....roughly analogous to my description of 'trim' ?

However, I'm not sure about the 'momentary' part.

Cheers
DD

 

In the HE supplement TPS is listed as providing acceleration enrichment. Page 19-5

The full load switch/s enable open loop mode Page 19-3

I would disconnect the TPS and test it with a battery (3x 1.5v batteries in series will give you 4.5v close enough) if you don't have a variable power supply.