zchead

In an older engine without o2 sensors you would be correct. But in the new age we have sensors that measure the amount of fuel in your exhaust. SO for an ecu to make your engine knock and then back it off is pointless. It has no need to do this with all the other sensors on it. The knock sensor is a fail safe. Nothing more.

thechief86

Ok, so help me make mine turn the lights off, lol

RJR

No, scope is correct. You're confusing mixture control (which is measured by the O2 sensor) and ignition timing control (which is measured by the knock sensor). It is perfectly possible to have a perfect mixture going into the cat converter while the timing is so retarded that the engine is running well below optimal efficiency.

Knock (pinging) occurs when the fuel in the combustion chamber stops burning progressively with a uniform flame front from the spark plug, and instead starts igniting in an uncontrolled fashion at multiple locations in the cylinder, vastly increasing the rate of pressure increase in the cylinder. Primary causes of this are:
1) hotspots in the cylinder caused by carbon or (in the old days) lead deposits which get hot and act like glow plugs. I have a good story about blowing a hole in an aircraft piston due to lead deposits.
2) low octane fuel which ignites spontaneously due to the pressure of the compression stroke.
3) ignition timing that is advanced too far, which is what we're talking about here. The reason this can cause knocking is that with too much spark advance, the combustion starts so far before the piston down-stroke that the pressure builds up high enough to cause spontaneous ignition of the rest of the fuel, as in (2) above. However, in this case even higher octane fuel won't prevent it if the timing is sufficiently far advanced.

The engine is most efficient when the ignition is timed at the optimum point just before knocking starts (or actually when mild - inaudible and non-destructive - knocking is just occurring). That allows the engine to extract the most energy from the expanding gases before exhausting them out the tailpipe.

Because of this, changing fuel grades will not cause or eliminate audible knocking in an engine with a properly working knock sensor, but it will affect power and fuel mileage. On my '94 3.0, the additional fuel mileage is just enough to offset the extra cost of mid-grade fuel over the basic grade, but the power increase is noticeable. So, running 91 vs 87 is essentially a way to get some free (much needed) horsepower. Easiest performance mod you can do on the 3vze.

And, BTW, to address another point made by someone earlier, if you ever put a timing light on a running engine and accelerate it, you will see that the timing advance far exceeds 10 degrees above the idle setting. Timing advance of 40-50 degrees is not at all uncommon. This has been true ever since the days of mechanical distributors.

scope103

Maybe I should put that in my signature?

And you don't even have to take my word or RJR's (though you should). Here's what Toyota has to say about the knock sensor:
http://www.cygnusx1.net/Media/Supra/...taTech/h38.pdf
http://www.cygnusx1.net/Media/Supra/...taTech/h23.pdf

dr1553

We've got some great new info here but I'm still a little curious about a few things I asked about before....forgive me for repeating myself....I know it's not cool but.....


Since the signal from the KS is so weak, or so sensitive to interference....

Is there a good chance that most of these code 52 could be from a break down of the shielding in the main wire itself? Not the signal wire? Looks like most people fix the problem by replacing the KS and the short wire, but not all. I guess the most likely place for the shielding to break down is in the short wire. But it could be anywhere between the KS and the ECM?

My ground at the plug under the TB tested good, no resistance, and the signal wire tested good back to the ECM, just a tiny deflection off zero on my meter......but that doesn't mean the shielding isn't compromised somewhere along the way? It could still be grounded even if it had a split somewhere that didn't go all the way around, but a big enough split to let interference in?

Is that all correct?

Scope, you said the shielding is only grounded at the ECM, correct? Where does the shielding end and go to ground on the ECM end? I just had my kick panel off and it's single wire into the ECM so it ended somewhere before I could see it.

When people run a new coax all the way from the ECM to the KS, where are they grounding it?

Could you just cut it back and ground it anywhere under there near the ECM? Splice in the new center signal wire to the single wire going into the ECM plug? Maybe cut the wire a few inches back and solder it to the new coax center wire? Then at the other end how do they get it connected to the plug at the end of the harness? I'm guessing at the KS end you don't need to worry about grounding the shielding, but it needs to connect to the ground side of the plug? That way the shielding on the new short wire to the KS itself is grounded thru the plug and then back to the new ground near the ECM.

If anyone has run a completely new wire, and it worked, and has pics please join in! I've written down in my notes that RG-58 is the wire to use. Correct?



edit to add....

Well.....after a little more research it appears rewiring the KS is common on other Toyota's also, especially Supras. Yes, they are using RG-58, stranded core if possible, solid core will also work but it's a little harder to run due to stiffness of the wire.

After seeing what they are doing I wonder if it would be better for us to just run the new wire all the way to the KS. That way you would eliminate the connector under the TB, the less connectors the better I'm sure. The only bad thing would be if you needed to pull the engine you would have to either cut the wire or pull the intake to disconnect it from the KS. Probably wouldn't be a big deal to rewire it again if you ever needed to pull the engine. Just cut it and pull the engine. Hmmmm.....

Just have to be sure that the new wire doesn't get cut where it passes under the intake and is protected with a little piece of rubber or something like the original. Might have to make the little cutout in the intake larger....not sure as I've never had one off to check it out.

scope103

I've always suspected shield damage on some of these intractable code 52 cases, but there is no way to know for sure. If the shielded wire had the insulation worn through the shield itself could be picking up current from -- something. If the wire was damaged enough that water or dirt got in, that could produce a low-resistance path from the core to the shield. Which one? I have no idea. If I was every cursed with this I would replace the wire from the pigtail connector all the way back to the ECU -- heck with splicing.

My EWD shows the shield ground eventually makes its way back to the camshaft bearing cap, but it's spliced somewhere in the harness (not at the ECU connector). Connector B (I show the KNK connection at B5) has no grounds, but C has a bunch of them (C11, C13, C26). You could ground the new shield just about anywhere, but to be SURE its using the same ground as the ECU I would splice into one of those BRown wires.

The shielding in the "pigtail" is connected to the shielding of the long wire through the 2-pin connector, but there is no ground connection at the sensor itself. I suspect this is to prevent what audio guys call "ground loops," but I don't really know.

Ideally (but not easily) you wouldn't splice to the old wire at the connector, you'd remove the connector pin and wire your new wire to a replacement pin. I'm pretty sure these are the correct pins: https://www.theplugdealer.com/shop/p...rass-tin-25190 (but I would visually compare them before I did any cutting on the old one). You'll need a tool like this: http://www.aliexpress.com/item/SN-28...383094433.html or http://www.aliexpress.com/item/SN-28...244923537.html (shop around; all you need is the jaws that fold over the tabs; the build-an-iPhone ones with replaceable jaws can be very expensive).

Good luck!

thechief86

Well, i just ordered a new knock sensor and short harness. I'm going to try to get the sensor to shut off the lights and the engine to run strong again by putting the sensor on the engine hook first, and if I can get a signal to the ecu, I'll pull the intake off after the first of the year.
I wish I had known to do this while I was doing the head gaskets! I didn't have a problem with it until I got it running again after that...

zchead

I have placed my knock sensor on the engine hook. My truck is running strong and does not have a knock. My check engine light is also off. I will post pics later today. Thanks You.

MudHippy

The 90 I bought a couple years ago had one on the engine hanger. It was giving me code 52 constantly too. Until I replaced the wire. Which, as previously stated, is ALWAYS the problem.


But the problem with drilling the massive hole in the engine hanger for it is severely compromising the strength of the hanger. Try pulling your engine, or engine/trans/tcase combo, now. I dare ya! Don't forget to record that for us when you do. I'd love to see it.

thechief86

I actually just did this, too, but I'm waiting for my new radiator before I try it out. But I foresaw the weak hanger problem, and actually picked up a spare engine hook to keep in the glove box, just in case I need to pull the engine.
FWIW, I welded a nut to the existing hook, and threaded the sensor into that.

dr1553

Did you go with an oem sensor and also replace the short wire between the main harness and ks? How bad was your truck running before? No power, terrible mileage....the norm?

thechief86

Mine felt like a bad miss under load at low rpm, but power and mileage weren't too bad.
I got an OEM wire and a $12 sensor off ebay.

zchead

I did the same thing the chief did. I think I paid $35 for everything. And yes all the norms of a bad knock sensor or signal wire. Would drive ten feet kick a code and go into limp mode. I only got around 200 miles to the tank. Could literally watch the hand go down.

zchead

I don't mind the week engine hanger. I always use my own lift supports when pulling an engine anyway

thechief86

No luck on my relocation efforts. Still having the same problem, and for some reason it seems to run way worse with a knockoff toyota style sensor w/oem wire than it did with the gm sensor w/spliced in wire.
It still runs fine above 2k rpm, but below that it feels like I'm running on 3 cylinders.
Snow is coming... we've established this isn't the ideal location for the sensor, but I will continue putting it here at least until I can get some recognition that it even has a sensor hooked up. Again, this problem didn't exist until I did the head gaskets.
Safe to assume the problem is under the hood? I'll do some research on how to test both the ebay yota sensor and the gm one, and if they check out, I'll turn the evil eye on the wiring.

MudHippy

Who said a GM sensor would work? I can't imagine it will.

Anyway...it'll work just fine on the engine hanger. If your connections(electrical/mechanical) are good. And it's putting out a signal the ECU can recognize.

thechief86

Some other thread I read a guy used one for an s10 because it was cheaper, so I tried it. No dice on my end.

dr1553

You used a non oem knock sensor? What kind? I've read countless threads that an aftermarket ks didn't work, then installing an oem fixed it. I'd sure like to know what worked for you.

What kind of mpg is 200 per tank?

cbh148

Sorry if this sounds like a dumb question, but have you been resetting the ECU to clear the stored knock sensor code (52) then seeing if it comes back after you've made a change to the knock sensor setup (whether it be relocating the original sensor and original knock sensor subharness, or replacing the 2 with brand new units)?

Just checking because I have a pretty good amount of experience with knock sensors on Nissan 240sx's, and it sounds like they're basically the same as these Toyota knock sensors -- not only in their construction and theory, but even down to the often-confusing style of shielded wiring used for the knock sensor part of the engine harness and the knock sensor subharness, where it's using 2-wire types of harness plugs and even the plug that goes directly onto the knock sensor is 2-wire style, but one of those wire ports on the plugs will often just be for a wire that comes out of the plug and envelopes the signal wire to protect it from interference.

On those early 90's Nissans, the majority of the knock sensors have bit the dust. It's always the knock sensor, never the wiring harness or knock sensor subharness. And it's not uncommon for people to just delete the knock sensor by wiring in a resistor that is of the particular resistance value needed to satisfy the ECU from throwing the knock sensor circuit error code. On my SR20DET, for example, a 1 Mohm (aka 1 Mega-ohm, aka 1 million ohms, aka 1,000,000 ohms) resistor does the trick. I just de-pin the factory knock sensor wire out of the ECU plug, then install another ECU wire from a junkyard car or from a random circuit that isn't even in use (like for the over boost solenoid thing that nobody uses) right in place of the one I de-pinned, then a couple inches up from the ECU plug I soldered one side of the resistor to the wire, then soldered the other side of the resistor into one of the engine harness's ground wires by teeing it in. No more knock code, no more limp mode, and I drift the car regularly and it's never had less than stellar compression numbers.

thechief86

I did disconnect the battery, and the light stays off until I hit 1600 rpm or so.
I'll play with it a little more tomorrow