Arnoud

Fred: good information to have about the workings of a catalytic converter in general, and thanks for highlighting the part you did. As I wrote: I do not think that both my catalytic converters are "clogged", but was showing here what happens when I continue to drive around for some minutes (within idle - 2000 rpm or so) once my engine goes into this fault mode. Looks like total over-fueling / non-burnt fuel to me...

Sean. IMS relay was and is in place, and none of the red nor green leds where on when this happens = both leds remained off (then again: I have never seen those red or green leds come on, so how do I know it even works?). Both coils + Beru wires + rotors + distributors were all changed for new during April 2014, as part of my preventive maintenance then. I have checked the rotors + distributors + measured the Beru wires many times, they are all good. Both ignition amps I have swapped with new ones in the past, which made no difference either. The coils I have not swapped till date, I can put my old ones back - just to make sure it is not them either....

Mrmerlin/Stan: thanks for these follow-up tests suggestions. The alternator I have just changed out for a new Greg Brown one (pictures some days ago in "above" thread), but I will measure it's voltage and report here on that and all other suggested tests.

I should be able to loan a local Sharktuner tool (or if not/need be: buy me on). Mmm...the full throttle switch is currently still not showing to work on Theo's tool, so I still have to adjust the various cables on the throttle quadrant as per your Rennlist instructions from 2013 (link: https://rennlist.com/forums/928-foru...l#post10498854 ). The throttle idle switch is working absolutely fine, exactly as it should.

FredR

Arnoud,

FYI- You cannot use sharktuner [ST2 version] unless you have PEMS fitted in the LH and EZ units or so I understand

Whatever the problem both units are glowing red so both have the same issue.

Mrmerlin

please put the old coils back in,
and the old distributor caps back on one at a time then test

Seems that the new coils are not as good as the old versions.
I have seen where new Bosch caps had a shorted lead in the cap this shorted a plug and made the cat glow
you could have 2 bad caps it showed up after the caps got hot

Carl Fausett

Please attach a vacuum gauge to the motor and run the car at idle and warm (the conditions when the idle stumbles). I'd like to know if the vacuum is stable, or whether the needle vibrates, or fluctuates. I'd like also to know the value shown.

GregBBRD

I would "chase" down and repair anything/everything abnormal, at this point in time. Leaving some component with a fault, while trying to repair a problem makes no sense...especially if that component has any possible affect on the problem.

Why isn't full throttle "showing to work"? Remember that the switch merely connects the trigger wire to ground....what happens if the trigger wire is grounded all the time?

The fuel injection system is simple...there are very few components that can "richen" the mixture and make the cats glow red, yet allow the engine to run fine some of the time. Especially true if the problem occurs to both sides of the engine, at the same time.

Temperature sender.
Throttle switch.
MAF
LH computer
Injection wiring loom

GregBBRD

I'm assuming that all of these items have been checked...more than once, after two years.

Any main ignition failure, on either ignition system (the two sides are essentially separate) will cause two cylinders, on each side of the engine, to misfire.

Certainly, ignition amps, coils, main coil wire from coil to cap, distributor cap, rotors should have been tested/replaced/substituted long ago?

Mrmerlin

Greg with every post the OP makes it seems new info comes to mind and a few more tidbits are released.
That said its been proven that the new coils are not as good as originals and there was a bad batch of Bosch dizzy caps as a few have found out,
We had one on a car from NY one cap was shorting to another plug once it got warmed up,
and this would have been very difficult to figure unless both header pipes were monitored by a dedicated O2 sensor,
thus being able to see the one sensor reading quite differently once the engine got warm,
in this case one cat was glowing.
In cases like this its better yo go back to last working condition as you may uncover the issue

GregBBRD

Both cats glowing red equally certainly got my attention...

The overheat protection system probably isn't kicking in, since both sensors are getting equally hot, at the same time, so the voltage difference isn't "seen".

Arnoud

Fred: yes, I know: need to have PEMS fitted in the LH and EZK units for the SharkTuner to work. I will either borrow such locally, or then order from new - if/when I start testing with it.

Thanks Carl, for chiming in and having me check on that. In underneath update thread I report exactly what I saw, once I had bought such vacuum gauge.

Greg: thank for your follow up. I agree fully with your advice, and have fixed the full throttle switch, by adjusting the cables on the throttle quadrant as per your post instructions on how to set those correct. It now shows to work OK on both Theo's tool, and on my multi-meter when connected to the LH and EZK connectors.

Arnoud

What have I tested in the past days:

Removed all patrol from the tank (as I reported before: I unfortunately had a mix of 20 liter old + 20 liters new 98-octane patrol), by taking it from the PS fuel rail and removing the fuel relay and jumper it. The flow of the fuel was about 3 liters per minute, as I measured here. This means to me: in spec (it should be at least 2700 cc per minute) and therefore nothing wrong with my fuel pump and filter.
After having emptied all, I put 18 liters of new 98-octane patrol into the tank. But the patrol never was, and is not the root cause of my problem - but good to have eliminated that now.


I checked the thermo elements E1 + E2 of the ICM for their combined resistance: 3,9 ohm. A bit low, as the WSM states: " Approx. 5 to 10 ohms between El and E2 for perfect condition therm0 elements and leads. ". I also measured (but not shown here) that the ICM "relay" box has pin 87 "connected" to pins A1 + A2 when in rest (i.e. no ignition nor engine running) and when the engine is running: true so. Hence: all working as it should.


Checked - once more, done many times already - temp II when the engine is stone cold and not running, and at ambient temperature (meaning: around 18-20 Celsius in the garage): 2776 ohm = all good.


Checked the voltage at the jump post, when just having started the engine and it running @idle: fantastic 14,28 V (that new alternator is really super good, Greg!).


Checked the engine vacuum at the front fuel damper, at this time the engine had been running @idle for 20 minutes or so: about 22 inch Hg. The needle is stable and does NOT swing around wildly at all, but: when the engine idle goes down every 5 seconds or so the needle does drop a few numbers (to around 20 or just under that) for a part of a second and then comes back here.


And then - with the engine switched off, and once more as done many times already in the past near 2 years - measured the coils and their leads.
Left to right upper row is the DS coil, middle row is the PS coil and the bottom row are the DS (2025 ohm) and PS (2019 ohm) coil leads. All in spec, as far as I can tell...


And also - with the engine switched off, and once more as done many times already in the past near 2 years - measured all eight (8) spark plug leads. All in spec.
Lead 8: 3K80 ohm
Lead 7: 3K87 ohm
Lead 6: 3K95 ohm
Lead 5: 4K00 ohm

Lead 4: 3K91 ohm
Lead 3: 3K91 ohm
Lead 2: 3K82 ohm
Lead 1: 3K86 ohm



Last test I did yesterday evening: with the engine running @idle, one by one I disconnected the electrical plug to each injector. The result was the same for all eight (8) of them: when disconnected the engine started to run a bit slower rpm wise and sounded like running on 7 cylinders, as it actually was. This indicates to me: no problem with any of the injector wiring, nor with the injectors.

More testings coming up in the next few days...

Lizard928

As a note to those following this thread. It is best to unthread the ends off the ignition leads and ohm out each end and then the wire individually. This can spot problems with the leads easier. However his issue has nothing to do with the leads.

siscogts

I remember my friend had similar problem after fpr change,when flooring the pedal. Inner diaphragm was nearly seizing (guessing how much time that fpr was stored before being sold) and was not behaving correctly... He cured it with some wd40 sprayed from vacuum port, then moved several times diaphragm with mityvac... that fixed the issue. Not sure if symptoms are 100 percent the same, but worth a try, maybe.

Arnoud

Thanks Colin, for pointing this out.
And indeed: these Beru leads (which I replaced as part of preventive maintenance during April 2014) are still measuring as good as when they were brand new. I also measured them all individually towards ground, just to make sure none had any signs of being shorted out to ground: all measured infinite M ohm, which is correct. Here is one such measurement:



Francesco: thanks for providing this information, everything helps so to unravel my mystery problem. The thing is: I have replaced the FPR and both FPD's for brand new ones as part of my made Intake Refresh (as documented fully in this now very lengthy thread). So what is the change that the old one was wrong, and the brand new one too? However: I will test this, as measuring provides factual data.

Arnoud

Other tests as done, so to further eliminate items:

Coding plug checked: part number is: 928.607.439.00 which is correct for MY1994 AT ROW. It measures correctly to be 150 ohms, and looking from the back (and having measured all): rest of the wiring is correct too.


I read in one of the many Rennlist threads, that Greg Brown mentioned that an often overlooked source of vaccum leaks is the following one, see his post #30 here: https://rennlist.com/forums/928-foru...l#post15198847

" BTW....there's another "invisible" air leak that happens in these engines. Right on top of the water pump is a black solenoid that opens to vent the charcoal canister. They fail and will stay open all the time. If the "secondary" valve over below the overflow tank fails (and the "Y" in the passenger fenderwell fails, which is super common), you will have a huge vacuum leak that you will not see with a smoke tester. The quick way to start looking at this possibility if to disconnect the hose that runs between the primary and secondary valves, at the valve above the water pump (non engine side) Run the engine. If you 'feel" vacuum leaking through the valve itself, that is wrong. The valve should not allow vacuum to pass, when the engine is running at idle.....except for a few seconds when it is first started. "

So I measured that here @idle: no vacuum leaking (nor did I felt anything) - hence the problem remains:


Took all the cables off from my TechEdge controller, that go to the CE panel (RPM, oil pressure and temperature). Problem remained (nor is any of those wires having any sign of being shorted to ground):


Measured the DS (1134 ohm) and PS (1115 ohm) rotors, in spec:


Measure the wiring from both 2nd-stage ignition controllers (which I have swapped for brand new ones, problem remains), pins 5 to the EZK connector pin 15 and 32: wires measures fine (the resistance from my DVM wires themselves is around 0,5 ohm):


Not shown here but also measured: both 2nd-stage ignition controllers pins 2 to ground: measured fine too.

More tests ongoing...

Arnoud

After having passed the compulsory yearly inspection (as we have here in Finland, including full smog test) by the end of April, last month May turned out a very busy work month for me. So I was only able to order and receive PEMs from John Speake (thanks John!). I borrowed a Sharktuner2, so to get more data and insight of my root problem(s ?).

Finally had some time this week to put the PEMs into a spare set of EZK:


And into a spare LH (and yes: this still has the old hybrid circuit in there. My normal LH has that replaced with John Speake's excellent permanent solution).


Connected my WBO2 (which is in the right hand side of my Finnish developed X-pipe) via TechEdge 2J9 controller to the Sharktuner2, while - on advice from John Speake - kept the factory NBO2 (which is in the middle of my Finnish developed X-pipe) connected directly to the LH. The Sharktuner2 software started fine, and I loaded the default EPROM for a GTS into the PEM's:


I did not changed any of the default parameters:


I then started to drive around, while just data logging. Attached is one file: "Run 7-driving+accelerate-to-100km-row-2232-20190605.txt" , which has a total of 6820 rows (including the header row).
- Just before row 2232 I accelerated (not WOT, but close enough fast throttle) onto a highway towards 100km, and it can be seen that all eight (8) cylinders have knock-retard applied then. Is that a normal log for a GTS?
- It can also be seen in this log file, that my idle does drop every now and then below 600 rpm. Still do not understand why that is happening too...
- Note that I have by-passed the ICM "relay" module, and thus far have not had my engine going into total "crap running mode".

Any comments and feedback super appreciated (and I will drive and data log more in the days to come)!