Emissions Troubles
Looking for some assistance with a long standing fuelling issue I have with my 98 MGF which is causing the CO to be sky high.
Apologies for the length of this post but I want to share all the checks I've performed in the hope that someone can help narrow down what is wrong.
Parts replaced:
- Cat replaced for MGOC standard i.e not an ebay special
- New O2 sensor
- New Coolant temp sensor fitted (brown one).
Despite the above the CO remained unchanged.
However, noticed that whilst at the testing station the cooling fans didn't kick in at any point. The temp gauge remained at 'normal' but my spider senses were alerted. When I next got an opportunity I tested by unplugging the coolant sensor (brown) and the fans kicked in immediately.
Plugging the PSCAN in I could see that the coolant temperature was being reported but the O2 sensor voltage remains at a constant 0.49v regardless so I went hunting.
I have tested the individual conductors from the O2 sensor back to the ECU (pins 27 & 28 on red connector) and these are good.
I have tested the heating element in the O2 sensor and this looks good using Dieter's pages as a reference.
Placing a multimeter across the O2 sensor I can see the voltage steadily rise as the device warms to peak at 0.9v. From this I consider the sensor to be good?
Engine earthing point has been cleaned to ensure a good contact with ground.
I have attached a couple of images as I'm confused by the behaviour of the ECU. My understanding is that the ECU should operate in open state until the coolant temp reaches 'normal' running temp (approx 90 degrees). However, if I'm interpreting the data correctly, mine appears to operate in closed mode when idling regardless of temperature. When I then pick up the revs the ECU drops to open mode - returning to closed mode when the revs return to normal.
Am I correct in thinking that the O2 value of 0.49v is not necessarily coming from the sensor but is actually a default value from the ECU itself?
My thinking is that the ECU is at fault here as it appears to be ignoring the coolant temperature despite the fact it is reporting what I believe to be true values.
I do have another ECU for a VVC but unsure whether I can simply swap them over or whether that will cause other issues?
Any advice would be greatly appreciated.
Apologies for the length of this post but I want to share all the checks I've performed in the hope that someone can help narrow down what is wrong.
Parts replaced:
- Cat replaced for MGOC standard i.e not an ebay special
- New O2 sensor
- New Coolant temp sensor fitted (brown one).
Despite the above the CO remained unchanged.
However, noticed that whilst at the testing station the cooling fans didn't kick in at any point. The temp gauge remained at 'normal' but my spider senses were alerted. When I next got an opportunity I tested by unplugging the coolant sensor (brown) and the fans kicked in immediately.
Plugging the PSCAN in I could see that the coolant temperature was being reported but the O2 sensor voltage remains at a constant 0.49v regardless so I went hunting.
I have tested the individual conductors from the O2 sensor back to the ECU (pins 27 & 28 on red connector) and these are good.
I have tested the heating element in the O2 sensor and this looks good using Dieter's pages as a reference.
Placing a multimeter across the O2 sensor I can see the voltage steadily rise as the device warms to peak at 0.9v. From this I consider the sensor to be good?
Engine earthing point has been cleaned to ensure a good contact with ground.
I have attached a couple of images as I'm confused by the behaviour of the ECU. My understanding is that the ECU should operate in open state until the coolant temp reaches 'normal' running temp (approx 90 degrees). However, if I'm interpreting the data correctly, mine appears to operate in closed mode when idling regardless of temperature. When I then pick up the revs the ECU drops to open mode - returning to closed mode when the revs return to normal.
Am I correct in thinking that the O2 value of 0.49v is not necessarily coming from the sensor but is actually a default value from the ECU itself?
My thinking is that the ECU is at fault here as it appears to be ignoring the coolant temperature despite the fact it is reporting what I believe to be true values.
I do have another ECU for a VVC but unsure whether I can simply swap them over or whether that will cause other issues?
Any advice would be greatly appreciated.
Last Edit:2 months 1 week ago
by paulusx2
Last edit: 2 months 1 week ago by Cobber.
Please Log in or Create an account to join the conversation.
The O2 sensor voltage should fluctuate. If it stays steady in PSCAN, but fluctuates while measuring with a multimeter, I would check the connection wires from the O2 sensor to the ECU connector, perhaps the wire is broken, or too much resistance in the connector. You can also clean the connectors, including the ECU connector, with contact spray.
I think you cannot use a VVC ecu on a standard engine, even swapping two identical ECU's may not work due to not matching security codes with the instrument cluster of ingnition key. But I have no experience in ECU swapping on my MG, I did encounter this on my Audi.
I think you cannot use a VVC ecu on a standard engine, even swapping two identical ECU's may not work due to not matching security codes with the instrument cluster of ingnition key. But I have no experience in ECU swapping on my MG, I did encounter this on my Audi.
by tnjk
Please Log in or Create an account to join the conversation.
- generous_dad
- Offline
- Time served MGer
- Posts: 152
- Thanks: 19
I may be barking up the wrong ECU but I post ith good intentions.
If I read correctly, you changed the CAT.
I understand that original CATs are not available or if they are, several hundred pounds (compared to a couple of hundred).
These CATs do not behave the same as original CATs and will cause an error at the ECU.
If I am barking at myself again, apologies. If there is any grain of sense, then I strongly advise you to call Mark at Kmaps.
He was the fuel and emissions engineer for these cars when they were in production (citation needed) and knows more about ECUs, replacement CATs and anything else about what comes out of the engine and how the ECU deals with it, than I have had hot dinners.
Good luck.
If I read correctly, you changed the CAT.
I understand that original CATs are not available or if they are, several hundred pounds (compared to a couple of hundred).
These CATs do not behave the same as original CATs and will cause an error at the ECU.
If I am barking at myself again, apologies. If there is any grain of sense, then I strongly advise you to call Mark at Kmaps.
He was the fuel and emissions engineer for these cars when they were in production (citation needed) and knows more about ECUs, replacement CATs and anything else about what comes out of the engine and how the ECU deals with it, than I have had hot dinners.
Good luck.
by generous_dad
Please Log in or Create an account to join the conversation.
I suggest that you select only the oxygen sensor voltage so that it updates much faster, which will help you to see if it's fluctuating.
I'm not 100% sure but yes, I think 0.49V might be a default value, so what you get if there is no sensor present.
I'm not 100% sure but yes, I think 0.49V might be a default value, so what you get if there is no sensor present.
by pscan.uk
Please Log in or Create an account to join the conversation.
Have you made any progress with this? One thing to point out: the system should go into closed loop a minute or so after start up, as soon as the 02 sensor has warmed up due to its internal heater and the heat from the exhaust gases. The sooner you pull away, the quicker it heats up. And the ~0.49V is a bias voltage created by the ECU that it uses, when the ignition is turned on, to detect a short to ground in the sensor circuit. That bias voltage doesn’t allow the ECU to detect an open circuit (because, when cold, the 02 sensor appears as an extremely high resistance ie open circuit); however, it does allow you to carefully probe at the 02 connector** with the ignition on, engine off, to see if you detect that voltage - if not, you have an open circuit between the ECU and the connector. It won’t be 0.49V, your digital voltmeter will pull it down, depending on its input impedance (a cheap DVM has been known to pull it a couple of volts lower). That 0.49V bias voltage drops away to zero as soon as the 02 sensor starts to warm up (as its resistance falls from many MegOhms) and begins to switch between ~0.2V and ~0.8V, which it should do around once a second. The ECU should switch from open loop to closed loop as soon as it detects the fall in bias voltage and the beginning of switching ie around ~30 secs to a minute after startup, regardless of engine temperature.
**. The 02 connector disconnected, and very carefully probing the sensor signal and sensor ground contacts, without spreading them, in the connector that leads to the ECU.
**. The 02 connector disconnected, and very carefully probing the sensor signal and sensor ground contacts, without spreading them, in the connector that leads to the ECU.
Last Edit:3 days 20 hours ago
by Martinr
Last edit: 3 days 20 hours ago by Martinr.
The following user(s) said Thank You: pscan.uk
Please Log in or Create an account to join the conversation.
- Airportable
- Offline
- Senior MGer
- Posts: 2047
- Thanks: 457
And here is another jewel from a new member.
There are few members who are as conversant with the technical side of the car, (13.8v as the norm). Those who think in voltages less than one are a great asset to the ‘bar.
Thank you.
M
There are few members who are as conversant with the technical side of the car, (13.8v as the norm). Those who think in voltages less than one are a great asset to the ‘bar.
Thank you.
M
by Airportable
The following user(s) said Thank You: Martinr
Please Log in or Create an account to join the conversation.
That’s kind, but I’m no expert; in fact, I was a bit hesitant to reply. I’ve had a Rover 100 (MEMS version 1.6) for some 20 years. I first came to the T Bar when I bought a PSCAN a few years back so I could add new key fobs, but, of course, I now appreciate all its features. I thought the T Bar only related to PSCAN! But I recently got drawn back to this forum again because I’ve just fitted, with a minor mod to the buzzer, one of Bruce Cotsell’s low-level coolant alarms.
My knowledge is partly from using laptop-based free Rover MEMS diagnostic software over the years. I wrote 2 Car Mechanics articles about this free software, the later one of which should be attached. So that, together with delving into the MEMS systems using my PicoScope, and being a long-term follower of ScannerDanner.com (his premium channel and the Youtube free stuff), is where I’ve picked up my info along the way.
I will try and make a habit of looking in on the forum just in case I can throw in my two-pennies-worth along the way.
Thanks
Martin
My knowledge is partly from using laptop-based free Rover MEMS diagnostic software over the years. I wrote 2 Car Mechanics articles about this free software, the later one of which should be attached. So that, together with delving into the MEMS systems using my PicoScope, and being a long-term follower of ScannerDanner.com (his premium channel and the Youtube free stuff), is where I’ve picked up my info along the way.
I will try and make a habit of looking in on the forum just in case I can throw in my two-pennies-worth along the way.
Thanks
Martin
Attachment not found
Last Edit:2 days 18 hours ago
by Martinr
Last edit: 2 days 18 hours ago by Martinr.
Please Log in or Create an account to join the conversation.
- Airportable
- Offline
- Senior MGer
- Posts: 2047
- Thanks: 457
I shall study that on a larger screen in the morning. I’ve played with “MEMS diag”, it was a while ago though, if this info casts any new light on it I will probably dust off the cable & reload the software, actually it’s should still be on the system.
Thanks.
M
Thanks.
M
by Airportable
Please Log in or Create an account to join the conversation.
MEMS Diag, that takes me back. I wrote about Pawel’s MEMS Diag app in my first article (attached). Not having any Android devices, I went to great lengths to run it inside a virtual machine running Android on my Windows laptop, and I did use it that way on occasion - it’s still a very attractive app.. It was more of a challenge than a genuine attempt to run the app, and it made me seriously consider getting hold of an Android tablet just for the sheer convenience of holding a tablet, rather than a bulky laptop, whilst sitting in the driver’s seat with the seat pushed all the way back to rest the laptop on my knees.
Martin
PS I replaced the attachment of the later Car Mechanics article, a couple of posts above, with a better copy.
Martin
PS I replaced the attachment of the later Car Mechanics article, a couple of posts above, with a better copy.
Last Edit:2 days 16 hours ago
by Martinr
Last edit: 2 days 16 hours ago by Cobber.
The following user(s) said Thank You: Airportable
Please Log in or Create an account to join the conversation.
Am I reading the above correctly that you can’t just swap out the ECU if you think it might be defective and that something else needs to changed at the same time?
If something else has to be changed at the same time can someone indicate what that might be and why please?
If something else has to be changed at the same time can someone indicate what that might be and why please?
by Delbourt
Please Log in or Create an account to join the conversation.
“Am I reading the above correctly that you can’t just swap out the ECU if you think it might be defective and that something else needs to changed at the same time?
If something else has to be changed at the same time can someone indicate what that might be and why please?”
I’m going off my 1996 Rover 100 fitted with the MEMS 1.6, and I’m presuming your versions of MEMS will be the same.
With my car, fitted with the Lucas 5AS immobiliser, if you fit a different ECU, the car will run for 2 seconds and then cut out. So you have to use the PSCAN to program the ECU to the alarm. I believe I have that correct: the temptation is to think you are programming the alarm and fobs to the ECU, but I believe that is not the case - the ECU is matched to the alarm. (I’m sure Mr PSCAN will correct me if that’s wrong.). Either way, the end result is the same: you need PSCAN or some similar kit capable of the ECU/immobiliser/keys programming. None of the free diagnostic software packages I’ve mentioned, MEMS Diag, Rover MEMS Diagnostic, MemsFCR…., can do this.
There is a fellow (Technozen Electronics) who will program the fobs, alarm and ECU if you mail them to him - if you trust any courier service that much. I bought my PSCAN specifically because of this feature.
If something else has to be changed at the same time can someone indicate what that might be and why please?”
I’m going off my 1996 Rover 100 fitted with the MEMS 1.6, and I’m presuming your versions of MEMS will be the same.
With my car, fitted with the Lucas 5AS immobiliser, if you fit a different ECU, the car will run for 2 seconds and then cut out. So you have to use the PSCAN to program the ECU to the alarm. I believe I have that correct: the temptation is to think you are programming the alarm and fobs to the ECU, but I believe that is not the case - the ECU is matched to the alarm. (I’m sure Mr PSCAN will correct me if that’s wrong.). Either way, the end result is the same: you need PSCAN or some similar kit capable of the ECU/immobiliser/keys programming. None of the free diagnostic software packages I’ve mentioned, MEMS Diag, Rover MEMS Diagnostic, MemsFCR…., can do this.
There is a fellow (Technozen Electronics) who will program the fobs, alarm and ECU if you mail them to him - if you trust any courier service that much. I bought my PSCAN specifically because of this feature.
Last Edit:1 day 9 hours ago
by Martinr
Last edit: 1 day 9 hours ago by Martinr.
Please Log in or Create an account to join the conversation.
Moderators: pscan.uk
Time to create page: 0.548 seconds