Crazy knock counts... what gives?
 

So I'm finally getting my car on the road... slowly. I was supposed to get some miles on it yesterday, but had a slight encounter with a skunk that took priority. (Those of you on OT know...)

Anyway, when I took it out on Saturday, I hooked up my diagnostic computer to it and it showed me knocking like crazy! It show 3,000+ knock counts for like 1,000 "ignition events" in about 10 seconds. How can this be?

Per my math, at 800RPM, that's 6,400 ignition events in one minute (8x800). So that's roughly 100 ignition events per second. So in 10 seconds at idle I should see 1,000 ignition events, right? If every cylinder knocked, the max I should see is 1,000 knocks over that same time period, right?

But I'm seeing well over 3,000. I thought maybe the rebuild would somehow magically fix this, but it hasn't. Some of you may recall I had a problem hooking up the sharktuner to do the EZK because of crazy knock counts. I tried a different EZK and had the same results. The only thing I can think of that would cause this is a bad knock sensor (they're only a couple years old, but one could be bad I guess) or a bad hall sensor or bad wiring/interference. Any other suggestions? Any way to diagnose?

Davek9 has built a home occilliscope that maybe we can try. Any advice on that?

Hi Andrew,

The logic is fine, but each cylinder fires once per TWO revs (it's a four stroke remember :) ).

So I make it (800 RPM x 8)/2 = 3200 ignition events per minute = ~50 per second. So yeah, your knock event count sounds WAY off.

Even if it were knocking on EVERY power stroke, the knock sensor would need to be seeing multiple pulses every stroke.

I know that doesn't help, but I have NO idea what is going on. Back to lurking for me until Mr Speake chimes in.

Steve.

Oh yeah...I know that... i was just trying to see if everyone was awake out there! :icon107:





Remember to divide by 2 next time... :banghead:

To monitor knock events with my Innovate Motorsports gear I tapped into Pin 4 on the 12 way diagnostic plug on my '88. Just above pin 2 (ground pin). Got the location of that signal from John Speake.
Not sure if that info is any use to you but if you have some way to "watch" that signal for 5 volt pulses you can see if your car is sending that many knock counts from that point in the signal chain...not that I can tell you where in the signal chain that is relative to the LH or EZK or knock sensors...lol.

disconnect one knock sensor at a time and see if it still does it, this will tell you if one sensor is defective.

Andrew,

Do you have any timing pulled back on the EZK chip?

As for the hall and knock sensors, I thought they pulled out timing as a default when they are bad, but I guess a wiring check wouldn't hurt.

If you have aftermarket stereo/gps, why don't you pull the fuses and try it with them out of the loop......maybe some odd interference?

I guess disconnecting the knock sensors one at a time to isolate each one would work.......When you replaced them, did you use a torque wrench on the bolt?

This diagnostic computer, have you checked knock on another car, how have the knock cycles shown up on another S4/GT/GTS?

If you do a systems check with this diagnostic program, do the hall and knock sensors check out fine?

Hopefully it is just some sort of electrical interference, not true knock. If you have access, can you sharktune a chip with the timing pulled way back until the knock goes away (if it does)?

I'm just rambling here, I'm down to getting a new alignment and cleaning my car up for that little road trip. What did you do, mention the "D" word in front of your car again?:icon501:

Jim

If you disconnect just one knock sensor the EZK will just retard the igntion 6 deg at mid to high rpm and load. You can't fault find that way :-)

If you're running high boost, and have a knock sensor faulty, then maybe the 6 deg isn't enough.

But if you'll logging lots of knocks even at idle, you have some wierd problem..... are you sure about the tester you are using ?

Generally speaking Knock sensors react to vibration , the shock waves which ripple through the engine block. They can misread vibrations from ANYTHING bolted to the engine. Knock sensors are carefully located and tuned to a specific application to try to avoid being triggered by other engine noise. Some are of the opinion that a bored and stroked engine for example may be enough change in normal engine noise to make the sensors less accurate.

Yes. I firget offhand how much, but I have at least 9* pulled out at the top end.

Damn it! I forgot to do this. I will try this. and see if there is any difference.

I have, but not for a long time. Most of the Detroit boys are running 86 or older cars.

Yes, they seem to be fine. Although I did get a temp II fault, but I ohm'd it and it seemed fine. Even checked at the LH and EZK. All was good. I cleared the error and will see if comes back.

Funny! I think it's some sort of interference. I'll try the stereo thing and see if there is any change. Also I think my Innovate WB took a crap, but is still connected, so that could be it too.

Yep lots of knock at idle especially when cold. It's better after it's warm, but even a slight rev and the knock count goes crazy!

As for the tester... well I'm not 100% sure, but the last time I used the STer if you will recall I had strange EXK symptoms that we couldn't resolve too. This seems oddly familiar, but that time I didn't have issues with my tester, now I do... go figure.

John, is there a way to scope or test the knock sensors? I was wondering if you hooked up a DMM or an occilliscope to the knock sensor connections, if one could tell if it's putting out the correct signal? If one was going crazy, that might be the simplest way to tell.

I don't think DR has had any issues, but I'd say I'm not even close to being stock. :) SO I guess there is always that possibility. I never checked it, but I wonder what if my Intercooler was resting on top of the knock sensors. Any vibration could send the knock sensor into a tizzy.


Keep the suggestions coming.

You can for sure watch the output on an Oscope. It will start ringing (get excited) when their is true knock. Otherwise it should have basically zero output. I can borrow one from work if you need me to.

John

Andrew, My 90S4 is up and running if you need to test your program, give me a bump. Mac

Thanks John. Dave K has just finished building one that plugs into your PC. I may build one myself or just borrow his.


Thanks Mac. I'll let you know.

There is no substitute for a good digital Oscope when troubleshooting a fuel injected car. Good luck.

Well then, **** man... bring one over! I'm free this weekend. You can show me how it's done. :D

Another thing I was thinking of is that this kind of supercharger that's bolted on to the top of the engine instead of the intake manifold is so close to the knock sensors that if it began to produce noise due to wear which could be normal for the supercharger, may cause the noise to confuse the knock sensors.

Dan
'91 928GT S/C :eek: