worf928

siscogts

Good point. During my tests i noticed that knock count increased just after i floored the pedal quickly... The only conditions to get knocks.
But... Is there a protocol? i tested the car on road (logically, not in the garage at idle),operating temperature, and i use 100 octane ron(87-88Mon) gas, a test on heavy traffic and a test done on open road. What your protocol is? What do you suggest?

merchauser

siscogts, can you tell us your various test and how many knocks recorded for those tests? also, condition and miles of your
car would help along with perhaps recent maintenance?

siscogts

first test 2014, used a KTS 301: it had about 123 k kms, at that time everything was stock, recent timing belt (camshaft adjusted to get exact zero retard or advance), caps and rotor were originals , spark plugs 10 k kms (triple electrode Bosch DTC), MAF was working even if I didn't know If was working very well , well or within specs .....and ....ask me for more specific question.
2 test: 19°C , one in the usual traffic, max speed 40 Kmh using first 3 gears, maybe I pushed up to 2000 rpm. 8 knocks /10000 ignitions, they went on the display after I tried to accelerate in 3rd gear at maybe 10-15 km H, engine struggled so I changed to 2nd .

second test on open road using low gears up to 5k rpm and over... 27 knocks/10000 ignition

Third test with KTS 301 again this year , test was made on open roads and repeated when I did a mountain road , full of hairpin bends(I hope my english is good enough) , pulled hard... 34 knocks , then another test and gave me 31 ...

Did another test last week,6 °C, now the car came with new plugs ,new ignition rotor and caps, 9 months old TB, new sensors almost all around, injectors cleaned , new fuel pressure regulator, new fuel dampers,new hall sensor, 3k kms lambda probe. This time I used my newly purchased Theo Jenniskend tool.I got only 3 knocks ...but because front knock sensor seems to be gone and ECU retarded ignition . Had a problem before with connector, this time connector works .I also have no WOT contact working , in fact I am planning an intake refresh and new knock sensors, a new TPS , new ISV,new hoses within next 2 weeks.(the only sensors I havent' changed are those that need to be changed ).MAF seems to be still within specs , using some reference Theo Jenniskens gave me, but I think I will do a check sending this one to Theo or john Speake, later. I do not often floor the pedal with my GTS, so I was not able to tell you or remember the difference before/after knock sensor failure. This car is so much powerful (especially compared to my daily car)that honestly I cannot recognize the actual power loss .

siscogts

forgot to mention that When I did the TB job I put GB baffle under the oil filler neck. This reduced a bit oil consumption , tha may vary from nihil /1000km upto a liter every 1000km.in the last 1500 km used about 300 grams of oil.

merchauser

in the hammer menu, what is system adaptation and what does it do? thx

siscogts

Could be idle adaptation?

worf928

worf928

So, the first question is: When Porsche decided on the specification for maximum knock count what was their protocol?
I don't know. Maybe it was "hop on the autobahn and run flat-out in 5th for the knock test." It certainly was not a static test.
There's no protocol I've found in Porsche's documentation. Maybe there is one and I haven't found it. I'm not going to bet on it though..

Test protocols will be constrained by many real-world factors. In particular, I don't have an authbahn close to me

My protocol consists of 2 to 3 'test suites'. First I make sure that the subject 928 is warmed to operating temperature. Usually, this is a static procedure: I allow the car to idle while I check tire pressures, replace carpets, etc.
The first test suite is conducted on the way from my garage to the highway. There are a couple of places where I can come to a stop and then produce a high-load low rpm situation - a second gear start. I will attempt to do two of these on the way to the highway unless I get more than five or so knocks on the first test.
The second test is conducted on the highway and can only be conducted in light-to-no traffic. I slow to 40 or-so MPH, select 3rd gear and floor it (or just let an auto kick-down) and hold it to the floor for several seconds. I will do this test two or three times, subject to conditions, unless I get more than one or two knocks on the first test.
If, at this point, I have registered quite a bit of knocking (~>20) I will do an 'over-run' knock test: I accelerate without tripping the WOT switch to high rpm and then allow engine braking to slow the car. I'll do this several times. Any knocking in over-run is indicative of leaking injectors and/or lots of carbon buildup (and possibly other things.) To date, I've only ever seen over-run knocking on GTSs.

I will typically not do knock testing if I already know that there's a bunch of stuff wrong with a 928. There's no point.