S4 hot power loss, glowing cats.
#211
Rennlist Member
Thread Starter
Thank you for your comments.
Yes, I must test known good items... Again. Because all testing of known good components was done AFTER the h pipe was put in, and therefore is void.
At this time, with * no o2 sensor connected *, it runs and drives. It does seem to warm up and get worse, but more subtle in manner until you cannot get it to Rev properly when accelerating.
That said, the idle is ( by my senses) very good. Which was NOT always true before I found out about the Erroneous o2 position.
Yes, I must test known good items... Again. Because all testing of known good components was done AFTER the h pipe was put in, and therefore is void.
At this time, with * no o2 sensor connected *, it runs and drives. It does seem to warm up and get worse, but more subtle in manner until you cannot get it to Rev properly when accelerating.
That said, the idle is ( by my senses) very good. Which was NOT always true before I found out about the Erroneous o2 position.
When you say the car idles fine, do you mean it idles fine at first startup, or continues to idle fine once the car has warmed up, or continues to idle fine even after the car has been driven a couple of miles?
The O2 sensor being unplugged does make a difference on a properly wired and properly running S4. Disconnecting the O2 sensor does not make the LH go open loop - it will make the car run progressively richer until it hunts badly, loses power, then stalls (at least with a manual - I've not owned an auto so not sure what effect the torque converter and auto timing map has on the behaviour).
If you're still seeing no change with the O2 sensor disconnected, then there is still a problem with the O2 sensor signal reaching the LH ecu or with the ECU itself. As you state, the previous tests were invalidated by the O2 sensor being in air, so you need to repeat those tests first to determine whether the problem is the ECU.
Also worth noting that even if there's no change with a swapped ECU, that doesn't mean the original ECU is fine, it just means there is a problem elsewhere which needs to be resolved before the ECU can be tested adequately. You could be facing multiple issues here.
The pot on the MAF is irrelevant in an S4 as there are no wires on the car side of the MAF connector for it. The wires for the LH ECU potentiometer input on an s4 are on the (unused on a US car) 3-pin harness connector loose behind the CE panel.
Given you had the car running well (albeit briefly) recently, my gut says its intermittent and likely a wiring issue - I'd be very carefully testing and wiggling the O2 sensor harness all the way to the LH ECU, looking for anomalous resistance/shorts/voltage between the signal pin and the sensor heater ground/supply pins.
Are you using a separate narrow band sensor for the ECU input, or a simulated one from the innovate?
The O2 sensor being unplugged does make a difference on a properly wired and properly running S4. Disconnecting the O2 sensor does not make the LH go open loop - it will make the car run progressively richer until it hunts badly, loses power, then stalls (at least with a manual - I've not owned an auto so not sure what effect the torque converter and auto timing map has on the behaviour).
If you're still seeing no change with the O2 sensor disconnected, then there is still a problem with the O2 sensor signal reaching the LH ecu or with the ECU itself. As you state, the previous tests were invalidated by the O2 sensor being in air, so you need to repeat those tests first to determine whether the problem is the ECU.
Also worth noting that even if there's no change with a swapped ECU, that doesn't mean the original ECU is fine, it just means there is a problem elsewhere which needs to be resolved before the ECU can be tested adequately. You could be facing multiple issues here.
The pot on the MAF is irrelevant in an S4 as there are no wires on the car side of the MAF connector for it. The wires for the LH ECU potentiometer input on an s4 are on the (unused on a US car) 3-pin harness connector loose behind the CE panel.
Given you had the car running well (albeit briefly) recently, my gut says its intermittent and likely a wiring issue - I'd be very carefully testing and wiggling the O2 sensor harness all the way to the LH ECU, looking for anomalous resistance/shorts/voltage between the signal pin and the sensor heater ground/supply pins.
Are you using a separate narrow band sensor for the ECU input, or a simulated one from the innovate?
#213
Rennlist Member
Thread Starter
#214
Rennlist Member
Thread Starter
Rob Edwards took time from his family and the (now) very nice cobalt gt to help me out today.
He has a shark tuner.
Nice piece if kit. Hadn't worked with one before. Nor still, as he ran the show.
Essentially, the long short of it is that it's 99% clear - with the swap to the h pipe with no exhaust hitting the o2 very early on after some glowing cats, the fact that the LH is JUST slightly bad has been masked for months.
With his working units and the shark tuner hooked up (but no o2 - I tried to weld a bung on with the pipe in the car, and I'm not that talented) - we logged for quite a bit and it runs as well as a stock 80k mile 928s4 auto runs - boring but reliable.
I have two logs of at least 2 dozen minutes each from the way home. I can post them for the more interested.
If not for humorous or special effect, the original ecu started to act up at around camp pendleton, and was really bad until about Highway 73. At that point it perked up like I hit the nos button and was fine after.
So. As always, it's a relay or the ecus.
He has a shark tuner.
Nice piece if kit. Hadn't worked with one before. Nor still, as he ran the show.
Essentially, the long short of it is that it's 99% clear - with the swap to the h pipe with no exhaust hitting the o2 very early on after some glowing cats, the fact that the LH is JUST slightly bad has been masked for months.
With his working units and the shark tuner hooked up (but no o2 - I tried to weld a bung on with the pipe in the car, and I'm not that talented) - we logged for quite a bit and it runs as well as a stock 80k mile 928s4 auto runs - boring but reliable.
I have two logs of at least 2 dozen minutes each from the way home. I can post them for the more interested.
If not for humorous or special effect, the original ecu started to act up at around camp pendleton, and was really bad until about Highway 73. At that point it perked up like I hit the nos button and was fine after.
So. As always, it's a relay or the ecus.
#215
Pro
Join Date: Feb 2010
Location: Atlanta GA metro, OTP North
Posts: 618
Likes: 0
Received 1 Like
on
1 Post
I believe the consensus is that the S4 LH computer WILL fail. It's just a matter of time as to when the hybrid circuit will go intermittent as yours seems to have. Mine did that, then failed completely at an inopportune time.
Have it rebuilt now and save the trouble later.
Dave
Have it rebuilt now and save the trouble later.
Dave
#216
Rennlist Member
I am not sure that I understand a "slightly bad" LH. Intermittently bad I get, if that's what you mean.
Sharktuner logs? Sure, post them up. Or send them via email, would be interesting to see what they say. (jim at siriuscyber dot net).
Happy New Year!
Happy New Year!
#217
Archive Gatekeeper
Rennlist Member
Rennlist Member
Jim- was about to e-mail you - I put my spare pair of PEMs in those spare brains, and loaded them with a stock '90 GT map, but checked the box for 24 lb injectors on the fuel parameters tab. We were in non-cat mode as there is only 1 bung in BC's pseudo-H-pipe so the NBO2 was completely removed.
The car ran reasonably well with that combination, we went out and terrorized Irvine (the car sounds _great_ with the pipe on it. ) Brendan has an LM-2 that we couldn't get a reliable reading out of, so we don't know the AFRs but judging how well the car ran they cannot be too far off. (We did log a _lot_ of knocks in cylinder 6, nothing more than 6 degrees but there's something amiss in that cylinder)
I sent BC home with my laptop, ST2 and the PEM'd brains, he's going to swap the cat back into it and install the WBO2 sensor into that. The WBO2 law for an LM-2's analog output is 0 V for an AFR of 7.35 (gasoline) and 5.0V for an AFR of 22.39. The LM-2 has a 14 pin pigtail for its analog outs as follows:
Analog Out 1 + (Lime Green)
Analog Out 1 – (Yellow)
Analog Out 2 + (Brown/White)
Analog Out 2 – (Dark green)
Analog In 1 + (Purple)
Analog In 1 – (Black)
Analog In 2 + (Grey)
Analog In 2 – (Brown)
Analog In 3 + (White)
Analog In 3 – (Red)
Analog In 4 + (Peach)
Analog In 4 – (Orange)
RPM + (Black/White)
So he will need to connect the LM2 pigtail to the pigtail of a female 1/8" stereo jack from radio shack or something, so it'll plug into the ST2.
The question for you: On the ST2, do you know which of the tip/ring/sleeve contact surfaces on the 1/8" male stereo jack is the +V in, and which is the ground? Maybe it's in the manual and I've missed it.
The car ran reasonably well with that combination, we went out and terrorized Irvine (the car sounds _great_ with the pipe on it. ) Brendan has an LM-2 that we couldn't get a reliable reading out of, so we don't know the AFRs but judging how well the car ran they cannot be too far off. (We did log a _lot_ of knocks in cylinder 6, nothing more than 6 degrees but there's something amiss in that cylinder)
I sent BC home with my laptop, ST2 and the PEM'd brains, he's going to swap the cat back into it and install the WBO2 sensor into that. The WBO2 law for an LM-2's analog output is 0 V for an AFR of 7.35 (gasoline) and 5.0V for an AFR of 22.39. The LM-2 has a 14 pin pigtail for its analog outs as follows:
Analog Out 1 + (Lime Green)
Analog Out 1 – (Yellow)
Analog Out 2 + (Brown/White)
Analog Out 2 – (Dark green)
Analog In 1 + (Purple)
Analog In 1 – (Black)
Analog In 2 + (Grey)
Analog In 2 – (Brown)
Analog In 3 + (White)
Analog In 3 – (Red)
Analog In 4 + (Peach)
Analog In 4 – (Orange)
RPM + (Black/White)
So he will need to connect the LM2 pigtail to the pigtail of a female 1/8" stereo jack from radio shack or something, so it'll plug into the ST2.
The question for you: On the ST2, do you know which of the tip/ring/sleeve contact surfaces on the 1/8" male stereo jack is the +V in, and which is the ground? Maybe it's in the manual and I've missed it.
#218
Rennlist Member
I sent BC home with my laptop, ST2 and the PEM'd brains, he's going to swap the cat back into it and install the WBO2 sensor into that. The WBO2 law for an LM-2's analog output is 0 V for an AFR of 7.35 (gasoline) and 5.0V for an AFR of 22.39. The LM-2 has a 14 pin pigtail for its analog outs as follows:
Analog Out 1 + (Lime Green)
Analog Out 1 – (Yellow)
Analog Out 2 + (Brown/White)
Analog Out 2 – (Dark green)
Analog In 1 + (Purple)
Analog In 1 – (Black)
Analog In 2 + (Grey)
Analog In 2 – (Brown)
Analog In 3 + (White)
Analog In 3 – (Red)
Analog In 4 + (Peach)
Analog In 4 – (Orange)
RPM + (Black/White)
So he will need to connect the LM2 pigtail to the pigtail of a female 1/8" stereo jack from radio shack or something, so it'll plug into the ST2.
The question for you: On the ST2, do you know which of the tip/ring/sleeve contact surfaces on the 1/8" male stereo jack is the +V in, and which is the ground? Maybe it's in the manual and I've missed it.
Analog Out 1 + (Lime Green)
Analog Out 1 – (Yellow)
Analog Out 2 + (Brown/White)
Analog Out 2 – (Dark green)
Analog In 1 + (Purple)
Analog In 1 – (Black)
Analog In 2 + (Grey)
Analog In 2 – (Brown)
Analog In 3 + (White)
Analog In 3 – (Red)
Analog In 4 + (Peach)
Analog In 4 – (Orange)
RPM + (Black/White)
So he will need to connect the LM2 pigtail to the pigtail of a female 1/8" stereo jack from radio shack or something, so it'll plug into the ST2.
The question for you: On the ST2, do you know which of the tip/ring/sleeve contact surfaces on the 1/8" male stereo jack is the +V in, and which is the ground? Maybe it's in the manual and I've missed it.
And don't forget to go into Sharktuner's Tools menu, WBO2 law, and enter the proper scaling (0v=7.35 and 5v=22.39).
Also change Sharktuner's injector opening-time parameter to 0.84, the Ford Motorsports 24# injectors (which is what I think you have) are quicker. That, and the 24# setting, should get you pretty close.
The LM-2's Analog-output-1 is the WBO2 output by default, Analog-output-2 can be used for a simulated NBO2 signal.
So Brendan, if you have the Innovate accessory analog cable, connect lime-green (pin 12) to the ring contact on the plug to the ST2, and yellow (pin 5) to the sleeve (ground). That will get WBO2 to Sharktuner.
NB-sim is brown/white (pin 13) with dark green (pin 6) connected to ground.
If you don't have the pigtail I can send a cable on loan if you want, to connect LM-2 to ST2 and LH for WBO2 and NBsim.
Cheers, Jim
Last edited by jcorenman; 12-29-2014 at 01:33 AM. Reason: spellling
#219
Rennlist Member
Thread Starter
Happy new year. We were talking about you Jim. I've never met you but we were like "Jim would know..."
YES. Dead instead of iffy. But it has been running the car...
#220
Rennlist Member
Thread Starter
Thank you. I'll get the cable - I'll need one some time I am sure, and they are on Amazon even.
It would all be academic other than to just see what's what, as it seems clear now it's the LH. It is strange that the the unlabeled one I have is also unable to run the car, but the conclusive nature of 85 miles between 80-100mph with absolutely no issues on Robs ecus seems like an open and shut case.
Btw, it's Brendan.
I have the files ready to upload. After some Homeland episodes I'll get them up.
It would all be academic other than to just see what's what, as it seems clear now it's the LH. It is strange that the the unlabeled one I have is also unable to run the car, but the conclusive nature of 85 miles between 80-100mph with absolutely no issues on Robs ecus seems like an open and shut case.
Btw, it's Brendan.
I have the files ready to upload. After some Homeland episodes I'll get them up.
For Sharktuner's WBO2 input, the tip contact is unused, the ring is 0-5v for the WBO2 input, and the sleeve is ground.
And don't forget to go into Sharktuner's Tools menu, WBO2 law, and enter the proper scaling (0v=7.35 and 5v=22.39).
Also change Sharktuner's injector opening-time parameter to 0.84, the Ford Motorsports 24# injectors (which is what I think you have) are quicker. That, and the 24# setting, should get you pretty close.
The LM-2's Analog-output-1 is the WBO2 output by default, Analog-output-2 can be used for a simulated NBO2 signal.
So Brian, if you have the Innovate accessory analog cable, connect lime-green (pin 12) to the ring contact on the plug to the ST2, and yellow (pin 5) to the sleeve (ground). That will get WBO2 to Sharktuner.
NB-sim is brown/white (pin 13) with dark green (pin 6) connected to ground.
If you don't have the pigtail I can send a cable on loan if you want, to connect LM-2 to ST2 and LH for WBO2 and NBsim.
Cheers, Jim
And don't forget to go into Sharktuner's Tools menu, WBO2 law, and enter the proper scaling (0v=7.35 and 5v=22.39).
Also change Sharktuner's injector opening-time parameter to 0.84, the Ford Motorsports 24# injectors (which is what I think you have) are quicker. That, and the 24# setting, should get you pretty close.
The LM-2's Analog-output-1 is the WBO2 output by default, Analog-output-2 can be used for a simulated NBO2 signal.
So Brian, if you have the Innovate accessory analog cable, connect lime-green (pin 12) to the ring contact on the plug to the ST2, and yellow (pin 5) to the sleeve (ground). That will get WBO2 to Sharktuner.
NB-sim is brown/white (pin 13) with dark green (pin 6) connected to ground.
If you don't have the pigtail I can send a cable on loan if you want, to connect LM-2 to ST2 and LH for WBO2 and NBsim.
Cheers, Jim
#221
Rennlist Member
Apologies, and fixed.
#222
Rennlist Member
Thread Starter
https://drive.google.com/open?id=0Bz...aXM&authuser=0
https://drive.google.com/open?id=0Bz...b0U&authuser=0
Raw Text. I have sorted it here and I don't see anything strange with the normalized ECUs and Tune other than a major amount of knocks, even at cruise.
I am severely tempted to somehow scale this to actually squirt enough fuel with E85 to see if the knocks disappear. My hunch is they will not because the system sees knocks that don't exist, and it is pulling timing because of other engine noises.
#224
Rennlist Member
Files ...
Raw Text. I have sorted it here and I don't see anything strange with the normalized ECUs and Tune other than a major amount of knocks, even at cruise.
I am severely tempted to somehow scale this to actually squirt enough fuel with E85 to see if the knocks disappear. My hunch is they will not because the system sees knocks that don't exist, and it is pulling timing because of other engine noises.
Raw Text. I have sorted it here and I don't see anything strange with the normalized ECUs and Tune other than a major amount of knocks, even at cruise.
I am severely tempted to somehow scale this to actually squirt enough fuel with E85 to see if the knocks disappear. My hunch is they will not because the system sees knocks that don't exist, and it is pulling timing because of other engine noises.
I disagree that the EZK is seeing knocks that aren't there, they look real to me. Random engine noise (or a bad knock sensor) would show up for all cylinders. In this case it is pretty much #6 alone, up to 9-deg retarded (max) when none of the others are retarded at all.
Something is definitely wrong with #6 cylinder. Perhaps that cylinder is very lean as a result of a weak injector? Whatever it is, I don't think fiddling with fuel is the right answer.
Check the plugs for sign of leanness, and also try swapping #6 injector with a neighbor, to see if the knocks follow the injector or stay with #6. And maybe have a look with a borescope.
Here's a plot of where the knocks are-- generally over 3000 rpm and mid/high load, which makes sense. What doesn't make sense is #6 alone. If you trace down column "KR6" on the right side you see it start from 8.6 retarded then slowly stepping down to 8.3 then bouncing back to 9.0 as a new knock is detected, then stepping down again. Nothing is happening on any other cylinder, in other places #7 or #4 retard a few degrees but nowhere near as much as #6.
#225
Archive Gatekeeper
Rennlist Member
Rennlist Member
The other funk in the data is the coolant temp, it jumps in 0.1 second from 186 to 192 to 181 and back many times. Ground issue somewhere?