Bill Ball

I went out and ran the Spanner with the rear knock sensor disconnected. It gave me a bunch of fault codes, including "rear knock sensor faulty". So, I cleared the EZK to run a fresh test. Now it insists there are no codes. I reconnected the sensor and drove around, pulled over and pulled the plug and drove around. The Spanner still insists there are no fault codes in the EZK even though the knock sensor plug is disconnected.... except now it indicates I have a faulty WOT switch in the actuator tests. Not sure how long it should take for the EZK to record the knock sensor disconnect, but I can't add anything here now.

Sorry, I think I misled everyone about the Hall and rear knock sensor relationship. I have a creative mind that does a good job of trying to make order of the universe and it did a good job earlier of implanting the notion that the Hall and knock sensors were linked based on what I observed last year. Man, I should just remove all of my posts for the last few days after a good series of similar misconceptions.

Bill Ball

JohnS:

If you are reading here, I have a question about the LH/EZK diagnosis that may help others as well as me. In the WSM (vol. 1: 30-1 to 30-40) I see discussion of separate error codes for the LH and EZK relative to things like the WOT switch. I get no error codes there, but the separate actuator test shows a WOT test timeout. How would that happen?

Also, I see now a note on 30-39 & 40 that says after clearing error codes you must drive the car at operating temperature a for at least 6 minutes then let idle for 60 seconds before checking for codes again. That explains why pulling the knock sensor plug after clearing the EZK did not register and why I am not seeing a WOT error code the EZK memory. That's rather important as I have cleared an ECU after thinking I corrected faults, then immediately re-run the tests and seeing no fault codes assumed I had successfully corrected the fault when I might not have.

worf928

Bill don't sweat it. Everybody gets a wire crossed from time to time.

Bill Ball

I don't mind making mistakes as long as they get corrected w/o harm. I'd been carrying this odd notion about the Hall and knock sensor around for the last year. Investigating this is getting me more familiar with the wiring diagrams and the Spanner, so it's all good in the end. And maybe we'll get some clarity on DrMark's issue.

John Speake

Hello Bill
I switched off my computer early yesterday .....


There are many points to be answered here...

First, Mark - If you obtain another EZK (only after your present unit is confirmed faulty) and this new unit is fitted with a GTS EPROM, then you will have full diagnostics.

Also, don't forget that the mapping of igntion advance with the GT EZK you used is different to the GTS. This might also affect pinking on your test drives

However, as I discussed with you on the 'phone, your GTS should not suffer from audible pinking because the knock control system should retard the igntion to prevent this. If any of the sensors are faulty, then the igntion is retarded by 6 dgrees at high load/rom to stop pinking.

Dave and Mark - At what rpm do you hear pinking ? I haven't heard of this GTS problem before. What octane fuel are you both using ?

There are no differences in the wiring harnesses from MY to MY that I am aware of. The EZK didn't change either.

The diagnostics are pretty good, but they should not be taken as an absolute. It is quite likely that once knock happens, spurious codes may be registered.

Bill, you are correct, I should add a note in the Spanner User Manual about clearing fault codes on the EZK and then going for a test drive, then reading codes again.

Tails

Guys,

When I was undertaking diagnostic fault finding on my 1990 S4 Auto in 2005, as the car started to miss badly and would not hold idle and when I was driving the car would just cut out.

By using basic testing techniques to try and find the problem without the aid of a Hammer, Spanner or O/scope, on removal of the connectors to the two knock sensors and the Hall Sensor, the female part of all the connectors all fell to pieces.

In addition I inspected the knock sensors whilst checking for vacuum leaks, MAF function, earth connectors etc., and when the MAF was removed as well as the two aft fuel regulators, fuel pipes and throttle cable and wheel good access to the aft knock sensor was available. The rear knock sensor plastic cover was heat affected and broken into many pieces as well as the shielding was unwinding.

To gain access to the front KS I removed the oil fill pipe and the cooling water cross over at the front of the engine and found that one of the crankcase vent pipes was badly split. When the knock sensor was exposed the cover cap was also broken up due to heat. In addition its connection to the EZK harness was also heat affected and broke up when removed. The only connection that did not fall to pieces was the speed sensor fitted to the bracket just in front of the bell housing.

I carried out repairs to all of the connections and knock sensors covers and checked the continuity of all the wiring to the EZK and all was in order.

The fault was in the LH and was repaired by John and upon refitting the whole system was tested by one of John's Spanners and all was satisfactory without any faults.

I just completed a 3,000 kilometre round trip with no real problems and averaged 11.5 litres per 100 kilometres which included freeway driving at 115 kl/hr, B roads at 105kl/hr and stop start city driving at 50 and 60 kl/hr. Speed monitored via GPS, so missed quite a few radar speed traps.

The best fuel consumption was 11.2 litres/100kms on a 600 kilometres section between fills. Travelled 900 kms in 9 hours with included two pit stops and one fuel stop.

Back to the subject in hand, all of the above KSs and HS were not interconnected externally, as far as I could see and they were only connected to the EZK. As mentioned previously I checked for continuity to confirm no shorts or grounding existed in the wiring of these units as well.

I would not venture an opinion regarding the frying to the EZK on the GTS's, I will leave this to more knowlegable electronic people, however if you need a good hammer and chisel man "I'm da man".

Tails 1990 928S4 Auto

Drmark

John,
Pinking is in 3500 - 5000 range on half throttle or more (I think it goes at WOT).
I am using Optimax 98 ron.
I think I have enough evidence that my EZK is at least part of the fault but I didn't realise that the diagnostics were on the EPROM - that is good news.
I am going to swap my EPROM into Paul Anderson's non diagnostic EZK from a 87 S4 and run with that for while before doing anything - hall sender and knock sensors aside, I shouldn't get pinking if the ECU is doing it's job (which the GT one seems to be doing). In the meantime I will run some injection and valve cleaner through the system.
If the pinking persists I will check / replace the Hall sender. If it's still there a MAF swap is my next course of action. I don't belive the Hammer when it comes to the new knock sensors - particularly as a second one seems to have failed within half an hour of the first test.
I will report back. Can I hold your MAF for a couple of weeks just in case?
And, last but not least, does anyone have an easy to follow guide / thread showing steps for opening up ECU and replacing EPROM?
Thanks
Mark

John Speake

Hello Mark,
There is the possibility that oil in the inlet is causing detonation. The GTS engines have some reporrted oil issues.

On some occasions my S2 has pinked around the max torque point - briefly, during a hard motorway run.

I agree to run the car with your GTS EPROM in Paul's EZK is a good idea.

Before spending money on a HAlls ensor change, check out the connector. behind the distributor on the driver's side. Usually the problem is the connector, rather than the sensor itself.

On one car I saw, the Hall sensor had been changed, but the fault persisted. We checked the 5v supply was getting to the Hall sensor, and ground. In the end we found that the Hall signal wire inside the 35way plug on the EZK was hanging loose.....

When you get pinking, is it transient for a second or two, or does it persist for as long as you keep your throttle at the same setting ?

How long has the car had this fault ? Any work done on the car imediiatley at the point at which it
started ?

BTW don't buy an EZK from an OPC. I am sure a good one can be found...

John Speake

I am very intrigued by Dave C's comments about GTS's and pinking. I have not heard of this problem before.

There is one aspect of the EZK knock system that should be explained. The knock prevention igntion retard appears to only work at high rpm and load. This means of course below these threshold levels the ignition will not be retarded, even when the Hall signal, or a knock sensor signal are absent.

The GTS has very good cylinder filling at low rpm due to the longer stroke, maybe this is a factor in its tendency to pink.

I am in the process of comparing the fuel maps for S4, GT and GTS..

drnick

john, id heard there was a 'big hole' in the mapping all through the midrange where timing was pulled back. can you confirm this?

John Speake

Hi Nick,
I can't say yet, as I haven't run a 928 on a dyno and optimised its igntion mapping. I am sure there are some safety margins in there, so the timing may be more conservative than it needs to be.

But still need to be mindful of charge desnity versus air temperature, fuel quality etc.

I was playing with the ST EZk on Adrian's s/c car yesterday. It was the first time I have used it on a car, so rather feeling my way. Everything worked as it should :-)

There is still lots of investigation to do...but it is coming on nicely.

worf928

I spent a little while searching for posts on GTS' pinging behavior. Rennlist's search function is too broken to allow me to efficiently accumulate data to support my recollection of a number of recent reports of pinging/pinking and/or EZK problems with GTSs.

I'm going to post a pole thread presently. Let's see what it says.

I'm fascinated to hear that the knock sensing system only functions at high-load.

My GTS seems to audibly ping during part-throttle operation and during throttle tip-in.

I only use the highest octane gas available: 93 octane (MON + RON) /2 (?))

In chasing this pinging problem I have:

- run many bottles of Techron through.
- confirmed correct fuel pressure (static tests - not under loaded conditions)
- replaced the fuel filter (possibly the one installed at the factory as the inspection paint on the bracket nuts was intact.)
- cleaned and balanced the injectors
- replaced both fuel pumps (in-tank pump was removed in two pieces, main pump electrical connection was buggered)

After each of these steps the problem diminished and thus they seemed to help a bit. But, the problem still exists. On our GTS the pinging only appears once the engine is very heatsoaked. The engine needs to be run on the road for 10 or more minutes after it reaches operating temperature for problem to appear.

I have the following 'theories' left to chase (others are welcome):

1) Accumulation of carbon deposits leading to a higher-effective compression ratio that when coupled with heat leads to pre-detonation. I plan to use a 'top end cleaner' such as that used by Randy V. I have a feeling that the engine runs hotter than my GTs. However, I have not measured any abnormal temperatures with my infrared gauge.

2) Some issue with the knock-sensing system and/or EZK. (Supported by failure of the Spanner to complete tests on the GTS. But, possibly debunked by John's assertion that the knock-sensing system functions at high-load only.)

3) An(other) engine Harness problem and/or common faults with GTS harnesses.

As we all know - and have posted about previous - the connectors for the Hall sensor, knock sensors and speed sensor turn to dust as the get older. Also, oil (or other types of) contamination will destroy the knock sensors. I suspect that, since the knock sensors are piezoelectric, grossly over-torquing the knock sensors will also compromise their operation but have no experimental data to support that suspicion.

In my GTS's case the knock sensors are not bathed in oil. There is a little residue but as far as I can tell with an inspection mirror they are in good condition.

John Speake

OK Dave
The poll is a good idea.....

Elevated engine/air temps with certainly make the tendency to ping greater, as will carbon build up, as you state.

I think the better low end cylinder filling of the longer stroke GTS engine also is a factor.

Most of the work I did measuring igntion mapping was with the S4 maps. Tommorrow I will try checking out the GTS map, and switch off the Hall signal on my test jig to see where the 6 degree retard points are. I'm too tired this evening...

I haven't heard of pinging problems as you describe on S4s or GTs. Many owners are blissfully unaware that a knock sensor or Hall sensor is faulty, and that the top end performance is missing.

I certainly see cars with spurious EZK knock and Hall fault codes. Bosch recognised this shortcoming of the diagnostics. so that every 50 or so igntion switch offs, the stored fault codes are deleted.

Try disconnecting the Hall sensor on your car Dave, you will see where it starts to kill the performance..

worf928

Hi John,
We should also note that the GTS is spec'd with a 10.4:1 compression ratio versus the 10.0:1 or lower ratio of GTs, S4s, etc. Thus, the GTS would naturally be more sensitive to variations of the variable factors that contribute to pinging.
Reports seem to be rarer than for GTSs - at least in my mind. Given the very low number of GTSs running around as compared to '87-91 even if the same number of reports occurred the ratio would favor GTSs pinging more.
I've never had an audible knock problem on my GTs.
Don't need to. BTDT with a dead Hall sensor on my '89. I know exactly what it feels like.

worf928

Actually, John, while you are poking around in the EZK...

Other than under conditions of knock sensor faults are there (do you know of) any conditions under which the EZK would not record knocks?