MrN0body

I purchased the car from a friend who had let it sit for a few years. It’s a 1999 ES300 and it’s giving me a knock sensor code. I know knock sensors are sort of infamous on this engine so I want to make sure whether it’s the sensor or actual knock.

MIL only turns on when car is under load. Usually freeway ramp or a road going up. P0330/P0325 are the codes and turn on at the same time. Knock sensors are tied together to bypass one that had already gone bad years ago when he drove it. I’ve checked the freeze data and the only thing I notice out of the ordinary is the high ~20% ltft on both banks but stft are good. I’ve used the code scanner to clear the codes and left battery unplugged to try and clear the ltft but it always remains around 20% so I’m not sure if that’s related.

The fuel pump was pouring fuel out due to the housing having hairline cracks. No more leaks. I’m not sure if this caused the fuel trim readings but I just replaced the pump today. I had assumed since it leaked gas, under load the car wouldn’t feed enough fuel due to the leak possibly causing some knocking and while the car does feel better power wise as soon as I get on the on ramp I fee it sluggish and light usually turns on followed by tranny limp mode. I attempted several times clearing the code and driving but it always comes back on the freeway, even just coasting eventually it goes on. If I drive around the city a few days it’ll sometimes go off, I’m assuming because it’s not really under a load.

I drive the car pretty calmly and I’m slowly working to restore it. Attached are a few screenshots of the freeze data at different MILs. Anything in particular I should log? I have a tech stream cable and a laptop with the software I could bust out but I’m not sure if that logs knock or anything better.




No MIL just some idle readings such as Maf

Arsenii

Hello,

I just got a little theory of what is going on, but I am not sure if it is correct. P0325 - Knock Sensor 1 Circuit Malfunction, Toyota Service Manual says it is triggered if "No knock sensor 1 signal to ECM with engine speed between 2,000 rpm and 5,600 rpm", which means that the cause of that is either a bad sensor, or a bad wiring that goes to it. The whole manual can be found here. This code is quite severe, so once it is triggered, the transmission always goes into limp mode, and the ignition timing is set to the maximum value. What I would probably do is check the wiring, and also check why and how the other sensor was bypassed, since it is a bad idea to drive like that for a long time, it needs to be addressed as quickly as possible.

If I understand correctly, Short-Term Fuel Trim (STFT) is what being changed constantly to achieve perfect air-to-fuel ratio (stoichiometry). If the ECU sees some patterns ("trends") in the STFT, it learns to compensate those with Long-Term Fuel Trim (LTFT). Please correct me if I am being wrong. In short, judging by your LTFT, it seems like you may have some unaccounted air entering somewhere, so that you need more fuel to achieve stoichiometry. It may be a bad MAF sensor, as well as a vacuum leak. The most basic way to check for vacuum leaks is to start up the car with the air box off and clog the throttle with something, the car should stall almost immediately. You can also spray all the vacuum hoses and the intake with something like a starting fluid (be careful not to catch fire), if there is a vacuum leak nearby, it will draw the fluid, and you will either see a change in engine behavior, or a spike on the A/F sensor graph. Granted, it is not the best method, but certainly the easiest. You can also visit a shop that uses smoke machines to see if there are any leaks.

Those are basically my guesses, and what I would check first. I am also learning as I go, so I would appreciate if someone can correct me if I am being wrong.

Hope this helps and best of luck!

MrN0body

Thanks for taking the time to respond I really appreciate it. Knock sensor 2 input at the ecm was spliced with knock sensor 1 signal. Essentially knock sensor 1 is feeding both inputs on the ecm. I’m planning on fixing the wiring today once the sun settles. AZ is too damn hot.

I have checked for vacuum leaks using carb cleaner and have replaced one caused by a formed hose for the booster, this was prior to replacing the pump. I believe the issue was not in unmetered air entering, but the car needing to add more fuel due to the leak at the pump. At the moment LTFT are currently 12%~ so they are slowly adjusting, I just thought maybe there was a way to clear the trim values to speed up the learning.

EDIT: I was unable to repair the splice, cable was cut right at connector. I’ll need to figure out if I can depin just that cable and replace. I was able to log to see data when MIL gets triggered, freeze data doesn’t provide much. Here are the logs if anyone wants to take a peek. I didn’t log all parameters, just a few because I was testing the logging feature. I went by code # to distinguish light coming on.