Speedtoys

Yes, fuel has "life".

https://www.bp.com/content/dam/bp-co...icle-tanks.pdf
https://www.sunocoracefuels.com/tech...w-octane-fuels

merchauser

thanks everybody for the gasoline education!

back to the knock solution........back in the weeds with the knocks. here is the chronology:

as mentioned, I make some runs this AM with 3 week old gas. knocks averaged 18. I got about 5 gallons from a new to me Shell station.
recorded knocks were 0 to 1 with an over abundance of runs. I did everything I could to replicate this over and over, and this was the first time
I had ever seen the ZERO!

so my car is capable of zero knocks...

just before I got home this AM I stopped at a Shell near my house to put in a few more gallons. thought I would put in some of more of the good stuff.
7 hours later, I go out and knocks are back at 18.....WTF.

I thought maybe I had some loose cables at either a knock sensor or the hall, but they are fine, and if there was an issue, the hammer would have displayed
a fault code: it did not. maybe the hammer is the problem?? I seriously doubt it. anything is possible, but I really don't think so.

I have been having an issue with the WOT portion of my TPS. most times contact will close properly at pedal down about 80%, but other times, it will not. for
sure it needs to be replaced. (of course, I should have done it with intake out, but it tested fine) could this be the issue??

FredR

The cruise maps are referenced until such time as the WOT kicks in. In cruise mode fuel cruise map is read but actual fuelling is modified by the input from the O2 sensor to produce an AFR of 14.7 [stoich]. Similarly in cruise mode the ignition map is read and values applied. When the WOT switch kicks in the LH brain ignores the O2 sensor, reads and applies the fixed cruise map values and adds additional fuel such that on full load the AFR goes to about 12.4. On the EZ unit [ignition] the cruise map is ignored and it applies the full throttle map which is a fixed degree of advance irrespective of load for each given rev band in the map. Thus if the brains do not receive that WOT signal, the engine will run a bit leaner with a bit less advance. Based on what I have seen with ST2, the fuelling does not seem to make much difference with respect to the onset of knock but the ignition timing most certainly does.

If you are getting erratic readings, then I suspect that it may just be that when the WOT switch is not working you do not see any knocks and when it is working you do. The hammer tells you something but not everything. ST2 data logging would show where and when the knock events occur and whether the WOT switch was working [or not].

Speedtoys

The number of variables...is lengthy.

Nobody doubts the Hammer data, but the testing is purely and entirely...interesting.

Detailed logging will show the larger story.

merchauser

Fred, you mean the inverse, correct??? or if your statement is accurate, what does this mean?

unfortunately, I do not have access to ST2, so the hammer is all I can rely upon.

I thought at WOT timing is retarded because the cylinder temps and pressures are higher than at mid-throttle, so you have to back off the timing, the fuel ignites faster and the flame front travels faster at higher load???

928 GT R

Great topic for continuing exploration and gaining the knowledge that will allow our engines to survive longer.

Many current high performance engine management systems incorporate individual knock sensors for each cylinder. Is there any way to retrofit individual cylinder knock sensors to our cars? My initial thoughts are that it would be too complicated and invasive, but I am continually amazed by the inventiveness and creativity exhibited by members of this forum.

How can we Improve our knock sensing/management capabilities?

Thom

If the WOT switch is not working then the EZK will not reduce the ignition timing to below the knock threshold by switching to the WOT map, as the EZK will believe that the engine is still running under cruise, where the knock threshold is higher than at WOT.

The 32V engines already use one cam position sensor per bank of cylinders. I ignore if the EZK is programmed to run sequential ignition already, but using 2 cam position sensors would be enough for a proper ECU to calculate in which cylinder knock occurs in function of engine position using even only just 2 knock sensors on a V8 engine.

ETA : The EZK would not natively handle sequential ignition since the engine is not running one coil per cylinder anyway.

FredR

I mean exactly what is written. When the motor is running on the cruise map the fuelling is leaner and the ignition timing is not quite as advanced compared to when the WOT is engaged and then more fuel is thrown in and a bit more advance deployed thus more likely to show signs of knock. Remember we are looking for a logic that explains why one minute you are clear and the next showing some signs of knock. Even if knock is prevalent with the WOT made the next question is whether what you are seeing is reasonable [or not].

The next step may be to take a look inside the combustion chamber with a boroscope if you can get hold of one. At the moment we are largely in the dark.Just remind me whether the injectors have been u/s cleaned [or not?].

Do not confuse the fact that higher load reduces the amount of advance that can be deployed with the advance used for the max power case where performance takes precedence over economy.

FredR

No need for that with ST2- the stock system can detect what cylinder knocked and when, ST2 can log the data so that one can actually work out what is going on. The engine management system was remarkably advanced for its period and despite what some seem to think it is still very fast today considering it can pull back timing on the next firing stroke. Unfortunately it cannot handle sequential firing although Ken [Porken] had a partial go at that by rewriting the code and making room by ditching superfluous coding.

928S MN

Didn't see where, if at all, it was stated the original fuel was E10 or not?
If it was E10, there is always a small chance that there was a phase separation in the Gas Stations tank. And what you filled up with was lacking the ethanol octane addition.

merchauser

I should be able to look inside this weekend. yes the injectors were just cleaned

am I wrong to think that because I recorded 0 knocks yesterday, that I should not have any carbon build up on the pistons?? I assume that is what you would like to me look for?
and those multiple zero recordings have nothing to do with the fuel??? remember, after I got the zero runs, I added more fuel from another station?

would a good test be to run this tank down and fill up with some 100 octane? that would seem to eliminate the fuel theory?

merchauser

hers is my unscientific testing this AM: I made many (lots!) of runs within each of the following parameters. no idea if it means jack,
but without a sharktuner, this is my best attempt at trying to isolate where the knocks might be occurring

-first I drove shifting in a near back and forth fashion with max RPMs at 3k making certain not to engage WOT; 0 knocks
-next, same idea with shifts @ 4k; 8 knocks
-then shifting to 4k, letting it drop to 2500, then back to 4k over and over, but with pedal floored to engage WOTS; 8 knocks
-finally the thrashed WOT, near redline; 18 knocks


nearly all items that might contribute to knocks have been repaired, or replaced. TPS and ISV are original, as are plug wires. I hope to
smoke test again, and believe that I need to replace the throttle shaft bearings, but I don't think any of those items are responsible for knocks?
how much carbon build up on piston tops does it take to increase knocks?

hoping this might offer a few more clues.

jcorenman

^^^ This. Having more variables than data really wrecks the logical process.

Paul, if the throttle switch shows any flakiness, then replace it. That removes one variable. The effect of a non-working WOT switch is (a) less fuel at WOT, which can certainly cause knocking; and (b) a slight (1-deg) difference in timing which likely has little effect. It does not effect knock-detection or retard.

The EZK, with a working Hall (cam) sensor, knows which cylinder is firing and knows which cylinder is knocking. It then retards that one cylinder in steps until it stops knocking. This is the same thing that any modern system does. The hammer only tells you that EZK is detecting knocks, but not which cylinder(s) or what RPMs. These would be useful things to know.

Fuel is a variable, but octane is not easily verifiable. In any event an S4 should have few if any knock-retards on premium gas, although 91 AKI is getting edgy. We do see differences in knock tendency between stations or one area to another, or temperature, altitude, season, etc. And I am not convinced that a pump labeled "93" is any better than one labeled "91".

But a lean mixture-- either in general or one weak injector-- would definitely be an issue. But again, no data. A wbo2 AFR gauge would be very useful here.

Have you run the WSM fuel-flow test? (Disconnect the return line at the tank and divert it to a measurement jug, jumper the fuel-pump for 30 sec and measure quantity-- should be 1.25L if I recall correctly). A weak fuel pump or a clogged filter will cause fuel pressure to drop at WOT, and a lean mixture. Knowing the L/min flow through the whole fuel system is an important data point.

Did the injectors come with a test-sheet showing flow rate per injector? Where did you install the weakest ones?

How do the plugs look? Are they all the same or some whiter and less-carboned than others? Any signs of leanness? If that correlated with which cylinders are getting retarded then that would be a great data point. (Note that plugs are beastly hard to read with all this "clean" fuel, so this is probably futile).

Certainly peek at the pistons, what you are looking for is any that do not look like the others. Post pics.

jcorenman

Part-throttle and you are getting knocks at 4K rpm? Are these stock brains or did someone slip in a set of Autothority chips?

FredR

Paul,

After you did this run what number the display show?