Here's the situation - Most people are aware of the problem with the stock knock sensors in the 1MZ-FE blowing so easily. I have the knock sensors replaced now with one that is much more durable. The problem is, it is MUCH more sensitive or at least the knock signal coming out is MUCH higher (measured with SFAC2) than it was before. Now you can imagine that the ECU is programmed to take certain actions with pre-defined levels of knock. With higher knock levels, it is pulling massive timing and dumping fuel trying to correct for what it believes to be a lean condition or pre-detonation. It makes it very difficult (read - impossible) to get a good A/F tune.
I'm sure it's not knock that I'm seeing. The motor's compression has been reduced from 10.5:1 to 9.0:1 plus I ran 115 octane race gas in it and I'm still seeing the exact same amount of knock (about 190 on the SAFC2) as before and it is pulling my timing advance to about 10*.
If it were a DC signal, I could simply attenuate the voltage level with a resistor. Does anyone know how I can attenuate the 0-5 volt AC output of the knock sensor?
I know 1 way to fix this is to go to a stand-alone EMS. I've considered that but with an automatic transmission, the cost and complications are making it seem not worth it. Any help, please? -JoeB
__________________ Please stop by and rate my ride:
I have always thought that the supercharger causes false knock readings since its so close to the knock sensors. I used to see knock on my obd scanner with race gas in the tank.
Great! Thanks! For some reason I didn't think simply adding resistance would attenuate the analog signal. I guess it's not really an analog signal exactly is it. I think it is just a variable 0-5VDC level. Glad I'm wrong! This shouldn't be too difficult at all! Thanks again, I really appreciate the help! -JoeB
__________________ Please stop by and rate my ride:
The signal from knock sensor is actually just like a signal you would get from a microphone (yes, a small AC voltage).
The resistor network in that picture is a voltage divider that attenuates the signal from the knock sensor... That's a reasonable way to do it.
You can also simply pull our the knock sensor and wrap it in a small amount of teflon tape - it'll dampen the vibrations of the motor mechanically. Its actually subaru perscribed fix for over-active knock sensors.
Good luck with either path you choose.
-Charlie
__________________
2003 Impreza WRX Wagon 5spd - 2.2L stroker + other goodies
1989 Camry Alltrac LE 3S-GTE 5spd - SV25/ST205 hybrid
1990 Camry 3S-GTE 5spd - parted out / junked
1990 Camry DX 3S-FE 5spd - The original white90dx; gone but not forgotten
I'll just build the attenuator circuit but thanks for the info. Getting to the knock sensors on the 1MZ-FE is no picnic. I did a little searching around on that link from gLok and found that Mike Chaney also wrote a program to determine the resistor values. The program allows you to plug in the resistance value of the knock sensor and plug in the amount you want to attenuate the knock signal (0-100%) and it gives you the value of the resistors you need to build the divider circuit! http://www.charm.net/~mchaney/f-tech.htm (for future reference)
I'm not encouraging anyone to do this. The knock sensing is very important and can keep your engine alive. Our ECU is actually pretty agressive in trying to advance the timing as much as possible, then backing off (retards timing and dumps fuel) when it sees knock. If you decrease it's ability to read knock, it could advance the timing too much and the resulting knock could damage your engine. My circumstance is somewhat different in that I'm using a non-standard knock sensor that is even more sensitive than the stock sensor.
I think with the supercharger, I was reading too much knock anyway. At least at certain RPMs. I don't think that was really knock either so I may have considered this anyway. Although with this solution, you cannot desensitize certain RPMs or RPM ranges, it is a constant reduction for all RPMs. The XEDE unit for the STi has the ability to reduce the sensitivity at certain RPMs as I understand it. Too bad we don't have something like that availavble. -JoeB
__________________ Please stop by and rate my ride:
As a side note, I just broke my rear knock sensor.
RIP... You were always good to me for not throwing the CEL's. LoL!
__________________
"The lamest twice banned, non-female member of-all time." -Ekam, Thanks, I <3 you too! AIM/Yahoo Toysrme257th
for anything, anytime; including camry turbos Now with Turbo!
190 seems quite high. I believe the S-AFC II just reads raw data (not neccessarily knock). I read this in the owner's manual. I get small numbers sometimes (comparatively speaking) like 25-50 sometimes. I think when the ECU runs in closed-loop it may not adjust timing. I'm assuming from how high your numbers are that it's not in closed-loop mode though. If this is the case, perhaps your ignition is being a bit zealous with its advance at WOT and just backing off once it realizes it is over-advanced (does it dart to 190 then maybe back down quite a bit?)
190 seems quite high. I believe the S-AFC II just reads raw data (not neccessarily knock). I read this in the owner's manual. I get small numbers sometimes (comparatively speaking) like 25-50 sometimes. I think when the ECU runs in closed-loop it may not adjust timing. I'm assuming from how high your numbers are that it's not in closed-loop mode though. If this is the case, perhaps your ignition is being a bit zealous with its advance at WOT and just backing off once it realizes it is over-advanced (does it dart to 190 then maybe back down quite a bit?)
Just a couple things to think about.
190 is definitely high, that's why I'm trying to bring it down. It's definitely not the ECU, nothing changedd there. It's just new knock sensors. I get knock readings on the SAFC2 sitting at idle, I'm very sure it is simply an over sensitive knock sensor. I get numbers as as high as 70 with extremely easly closed loop driving. Then when I step on it, it will quickly shoot up to around 180-190. Of course that only lasts for a second, then the ECU steps in, dumps fuel and pulls massive timing to correct what it thinks is knock or a lean condition. -JoeB
__________________ Please stop by and rate my ride:
I think I may try white90dx's suggestion and see if that works well enough for my application. If not, you can always get a GM knock sensor. Supposedly a good fix for knock sensors that break, and I'm assuming it'll likely be a pretty good (not over-analytical) sensor.
I think I may try white90dx's suggestion and see if that works well enough for my application. If not, you can always get a GM knock sensor. Supposedly a good fix for knock sensors that break, and I'm assuming it'll likely be a pretty good (not over-analytical) sensor.
The GM knock sensors are the problem. The writeup was written specifically for 4rth gen f-bodies. That solution worked GREAT! It'll still read a peak knock of about 70 on a hard pull but that sounds about right. At least that's the way it was before. Thanks again for the help. -JoeB
__________________ Please stop by and rate my ride:
The AutoGuide.com network consists of the largest network of enthusiast-owned enthusiast-operated automotive communities.
AutoGuide.com provides the latest car reviews, auto show coverage, new car prices, and automotive news. The AutoGuide network operates more than 100 automotive forums where our users consult peers for shopping information and advice, and share opinions as a community.
ToyotaNation.com is an independent Toyota/Lexus enthusiast website. ToyotaNation.com is not sponsored by or in any way affiliated with Toyota Motor Sales, USA, Inc. The Toyota, Lexus and Scion names and logos are trademarks owned by Toyota Motor Sales, USA, Inc.
Attenuating Knock signal?
Linear Mode
