SS21

Rennlist has been invaluable during the 2+ years of my 993 ownership, and it has lent its hand to resolving various issues I have had, however I continue to be stumped with an ongoing engine pinging issue.

Car details: 1997 C2S Manual, 106k miles, UK car, no engine modifications other than Gemballa silencers, BMC air filter.

So, this is going to be a long one, and in an effort to be concise I will try and list all the parts that have been renewed, confirmed working, and work undertaken.

Engine ping noticed during first few weeks of ownership. Looking under car plug wires are in terrible condition and arcing. In due course, entire ignition system renewed (plugs, leads, caps, rotors), remaining parts tested or confirmed working with known working parts (coils, ignition amplifier under seat). Engine runs better, no longer arcs, but pinging still present.

Codes read with Hammer, results – Knock registration not possible, Control Unit Faulty. This, coupled with an idle ignition advance at 36 degrees, led me to suspect a second rate ECU chip. Lo and behold the ECU had an aftermarket chip. Replaced with a standard chip, idle advanced confirmed to be in normal range 6-8 degrees.

From this level of work the car runs better, however engine pings regularly when under load (more present when hot) at approx. 2k, 3k, 4k rpm, revving clean to redline – this has been confirmed recently on the rolling road, where the car only made 256hp hot. After a cooldown and at 60 degree C head temp, car made 264hp.

• ​​​​​​​​​​​​​​Car run on 98 or 99 octane, plus additional 2 point octane boost additive.
• Distributor belt and alignment checked – hall sender verified with oscilloscope
• Distributor advance counterweights checked
• DMF replaced
• Crank sensor replaced and gapped – and verified with oscilloscope
• Fuel rail pressure checked
• All earths cleaned or renewed
• Continuity checked between ECU and engine harness
• Vacuum pipes checked (rearmost solenoid and actuator replaced as faulty) – engine holds vacuum overnight, makes good vacuum at idle (+/- 18 inhg)
• Carbon deposits on piston crowns do not seem abnormal – although I can’t comment on sensitivity of these engines to carbon buildup and hotspots.
• Leakdown test results in spec
• All hydraulic lifters replaced
• No fault codes since standard chip fitted, other than knock registration – to be expected

Injectors and valve timing are the only things I can think of that I have not checked.

Compression test results:

• Bank 1: 230, 230, 230
• Bank 2: 217, 214, 215

Observations and thoughts so far:

Compression results are spread nicely across banks, however +/-15psi difference between the banks (also are these readings a little high overall?) Could this (and poor dyno results) be pointing towards slipped valve timing from pinless cam pulleys, resulting in engine ping?

Plugs, cap and rotor have been monitored for wear – all wearing nicely.

I will note, somewhere on the web my research dug up a point that disconnecting the hall sender would cause the injectors to default to batch firing instead of sequential (also stated in workshop manual), but that it would also retard timing 6 degrees as a preventative measure. This makes sense as the ecu no longer reads cylinder 1, however I cannot find this info in the workshop manual. In practice, I have not been able to see a drop in ignition timing with a strobe light, even at 2k rpm, nor a perceptible drop in rpm when unplugged. Can anyone confirm this? I have also tested this theory on a 964 with no results.

The most recent work has been the tappets, and this does seem to have diminished the pinging, although it is still there.

However, when testing the hall sender I unintentionally left it unplugged and drove home. I noticed the car did not ping. I kept it disconnected and the car did not ping under conditions that previously would reliably produce engine knock. I connected the hall sender and was able to produce engine knock. What does this tell us? The car is happy in batch firing injectors? The car is not happy with ignition advance? Or that it is not advancing ignition correctly?

Weirdly, since having it reconnected, the car is the best behaved it has been, rarely pinging, although still occasionally. Car feels a little flat in power curve.

It seems that high piston speed is fine (revs clean past 4k) however slower speed, with high load and little advance are where it suffers.

I appreciate this is a longwinded issue, with plenty of info to try and digest. I am keen to hear your thoughts and put together the missing pieces to this puzzle.

Thanks in advance.

IainM

Are you sure it's pinging as in premature detonation and not something else e.g. Exhaust/ heat shield resonance? Seems like the engine is running great.
maybe the knock sensor(s) have failed?
can you see a burst of knock pulses on the ODM line when you hear the pinging?
if the timing had slipped I'd expect much worse results

pp000830

Can you describe the "ping" sound or better yet record it for listening?
Ping could mean post ignition sound in the exhaust, misfire or other sounds all similar but of differing origins.
Are there any other symptoms when it "pings" such as momentary power drop etc?

SS21

Certainly is engine ping, very much brought about by specific conditions - engine load, throttle, rpm,
etc. Whether it is pre ignition (hope not!) or detonation I can’t determine. The real confirmation came on the rollers where a gizmo was plugged into OBD port that would flash when the engine registered knock, hence confirmation of rpm. Hammer, etc will only register knock, not tell you when it is happening - also does not tell you which sensor registers knock. I’m not convinced knock sensor(s) are faulty. Does anyone have a suggestion to test without throwing more time and money at working parts?

Live data, along with power curve, show ignition being backed off as ping is registered, advance fed back in, ping, retard, cycle repeat until after 4K where it revs free to redline.

Engine appears to run fine otherwise, but the rollers don’t lie and it’s down circa 30hp when hot. Any comments on compression results?

Thanks!

BobbyT

I know this sounds like a remote possibility, but some owners have had "pinging" faults pop up caused by incorrect fan/alternator belts, even with belts of proper size in every respect. The belts need to be Genuine Porsche, the ones come in Porsche packaging, with "Porsche" stamped on the belts. One of our local PCA members had a career in a V-belt related business, and he has the same opinion. Something about the way the Genuine belts flexes, and the others don't...and yes, it does sound weird. Easy thing to check.

SS21

Yes I’ve heard this one before! And whilst I erred on the side of caution and bought genuine belts with Porsche stamped on them, I haven’t fitted them as the current ones (genuine) have plenty of life. They live in the front as emergency spares. I can fit them, because what do I have to lose, and it won’t cost me anything... but I’m struggling to understand the concept of belt size/tension affecting engine timing/ignition? As you say, something to do with flex, slippage...

BobbyT

Hope to hear good news after you fit the new belts! The incorrect belts don't really affect timing, but they apparently generate a noise or pulse that the knock sensors pick up, interpret as pinging, and trigger fault codes.

MarinS4

I suspect lean conditions are causing the pinging. Did you get AF trace when on rolling road?

Oil from BMC/K&N air filter coats the MAF causing lower airflow to register. This results is lean conditions. I’d suggest replacing MAF and Air Filter.
Clean and flow check fuel injectors. Over time deposits decrease injector flow. Don’t be surprised to see flow down 5-10%

IainM

The ECU constantly adapts, searching for optimum timing and so it is normal to see activity on the knock line in the ODB.
I built myself a plug in counter to monitor mine and was surprised how many pulses I counted.
I would see 1 or 2 every 10 secs or so at constant speed but on a WOT burn I would see a burst of pulses, as many as a 100 as the system tweaked timing as revs climb.
never got any hesitation, misfire codes or anything. This is just how the feedback system works to get optimum efficiency & power
havungvsaid that, I've never heard mine ping

pp000830

It would be interesting to see if you use fuel other than pump gas from your present source as a control if your problem goes away.
Is the gasoline sold in your area alcohol-free?
Can you buy a 5 gallon can of racing gas, high octane/no alcohol or moisture just once as a control,
Have you replaced your fuel filter any time in the past?

SS21

MAF readings, other live data, and engines ability to redline without issue suggest fuelling is fine. Fuel pressure at the rails has been checked. I do agree that injector spray pattern and flow rate is one of the few things I’ve yet to verify. Perhaps I should make this a priority for peace of mind. I have been running injector cleaner for quite a while now, but I appreciate this has its limits.

Fuel filter has been replaced, in fact I did it again last week during tappet overhaul.

MAF itself has been cleaned a number of times.

Just researching now and apparently Esso 97 fuel “has no requirement for renewable fuel (such as ethanol)” so I will give this a try (minus additive) and see how it drives. Bear in mind this is 97 Ron and lower than the usual 99 + 2 ron additive I run.

Are there any thoughts on the absence of knock with the hall sender unplugged? Does it in fact retard timing, or just switch to batch firing injectors?

Thanks all.

MarinS4

Compared to what? Sure you can read MAF data but can you compare to what it’s suppose to be? What about AFR’s? Lean conditions promote detonation. That’s the first check IMHO! Verify you have proper fueling. I like to see sub 11.5 under WOT peak toque and sub 12 towards redline. What are you seeing?

Lorenfb

Without the Hall sensor connected and if a knock is sensed, all cylinders are retarded by the same amount, verses just the knocking cylinder/cylinders.

-nick

Without the hall sensor timing is pulled “at certain load ranges” per the factory manual. That’s also why you won’t see any difference at idle with a timing light.

I second the air/fuel trace. That should have been the focus on the dyno. What did the result look like? With aftermarket mufflers, you should expect to be a bit leaner than stock, but not so lean as to get into pinging territory.

Timing at idle should be around 3° if I recall. It starts at something ridiculous like 35° at 0°C, then timing is slowly reduced as it warms up. I suspect that’s why you saw enormous idle timing on the aftermarket chip.

SS21

On the rollers Lambda trims were small so MAF and fuel pressure were ruled out.

I have borrowed a wideband lambda and a quick drive yesterday showed 12.5 at redline, 14-13.5 upto 4k WOT. When pinging would occur there was no perceptible spike in readings, it was however pinging at around 13.5-14.1 AFR. Idle and steady throttle are Lambda 0.99-1.

Disconnecting the hall sender made no noticeable difference to readings, the car did feel very flat understandably, again no pinging.