PDA

View Full Version : Splicing wire for knock sensor questions



tuner4life
10-28-2011, 08:12 PM
Alright, I finally got Camry #2 driving. It just has 1 smallish issue. It throws a CEL code 52 (knock sensor). and it runs like crap below like 2000 rpm. I did notice while it was all apart that the wire going to the knock sensor had been taped up. I remove all the tape and find that it had been spliced. The main wire was spliced using a crimp "sleve". and seems solid. What I am confused about is there is some metal "stranding" I guess you would call it, that doesnt seem to be in wire form. Instead it seems to run around the inner wire sort of covering it. Anyways, where it was spliced that "surrounding" wire was just pulled back and not connected at the splice.

Could that cause the issue with the knock sensor? It looks like a single wire plug at the sensor, so what is this extra wire stuff? Where does it go? What does it do? and should I try to splice all of it together?

Sorry if this sounded like jibberjabber. I hate electronics.

Hookecho
10-28-2011, 08:44 PM
That outer wire is the shielding. It is supposed to be grounded. All it does is keep other signals from the wires in the loom from interfering with the knock signal. It does not carry a signal at all. The inner wire carries the signal. It's possible that it could be the problem but i doubt it. Either the sensor is bad or the inner wire connection is bad.

tuner4life
10-28-2011, 09:02 PM
ok thanks.

Hookecho
10-28-2011, 09:10 PM
Here's a diagram on how to rewire the knock sensor with new shielded wire. I got this idea from Luni.

I suck a MS Paint but you get the idea.

http://www.celicatech.com/gallery/files/4/1/4/3/knock_diagram.png

tuner4life
10-28-2011, 09:18 PM
ok. I may try this tonight. You still end up splicing the signal wire to the factory plug though, correct?

Hookecho
10-28-2011, 09:37 PM
Yup.

l0ch0w
10-28-2011, 11:26 PM
PM me, I can make you a new patch harness that doesnt contain any splices (which can be bad for knock sensor wiring). I have the correct crimper tools and terminals.

Reason you dont want any splices is because the greater the mass of unshielded cable present the higher likelyhood that your knock sensor might pick up harmful interference. Splices add alot of unnecessary metal (butt connectors, solder, etc...)

alltracman78
10-28-2011, 11:34 PM
You fucking meatheads.
There's no need for that, the shielding is grounded to the engine ground wires at the ECU.
If you ground both ends you may end up sending a signal through it.


Here's a diagram on how to rewire the knock sensor with new shielded wire. I got this idea from Luni.

I suck a MS Paint but you get the idea.

http://www.celicatech.com/gallery/files/4/1/4/3/knock_diagram.png

alltracman78
10-28-2011, 11:37 PM
And no shielding shouldn't cause a MIL [not impossible though].
Not that you shouldn't fix that splice correctly.

The MIL is caused by a LACK of a knock signal.
No shielding usually would usually cause either nothing or more of a knock signal.

If the crimp is on the rubber coating instead of the wire that may be the problem.
Or you may have a bad knock sensor.

Hookecho
10-29-2011, 12:06 AM
Jeremy. If you look at my diagram you will notice that only one end of the shield is grounded. The other end does not go to the ecu. This diagram is for running a new and independent wire. Meathead.

On the factory wire i think the shielding is grounded by way of a splice wire to the ground wire within the loom.

Hipster Lawrence
10-29-2011, 01:36 AM
Replacement knock sensor harness are available from the dealer for some cars. Every Toyota I've had with a knock sensor code (rare) was fixed by replacing the harness. The sensors don't seem to go bad that often.

l0ch0w
10-29-2011, 02:05 AM
Those replacement harnesses are only for v6 trucks... they dont go all the way back to the ecu, they are merely an extension piece so that the tecnician can test the sensor without having to remove the whole freaking intake manifold

l0ch0w
10-29-2011, 02:21 AM
You fucking meatheads.
There's no need for that, the shielding is grounded to the engine ground wires at the ECU.
If you ground both ends you may end up sending a signal through it.

You are retarded....

The shielding splices off the ecu side of the cable but thats really the only difference here... the brown wire that comes off the sheilded line then splices into the main ecu ground (which goes both to the back side of the manifold and the back of the ecu) ITS ALL GROUND DUMMY..

If you are installing a patch harness, or looking at a wiring diagram, the way he drew the picture is correct and functional. Its not like shielding is directional or something...

Luni
10-29-2011, 05:33 AM
Ive personally confirmed with a DMM by ohming the shielding out, it terminates to the E1 ground on the intake manifold. Bruce rewired his KS with the same exact instructions from me and it worked perfect.

Jeremy, if the shielding fails it WILL cause a CEL. Ive seen it too many times for it not to be true. Do it the way Bruce said to, and you will be fine. If your wiring is truly the issue, and you do your wiring repair job properly it WILL fix your issue. I promise.

l0ch0w
10-29-2011, 05:53 AM
I can confirm this too because ive torn a whole harness apart and am rebuilding it...


Ive personally confirmed with a DMM by ohming the shielding out, it terminates to the E1 ground on the intake manifold. Bruce rewired his KS with the same exact instructions from me and it worked perfect.

Jeremy, if the shielding fails it WILL cause a CEL. Ive seen it too many times for it not to be true. Do it the way Bruce said to, and you will be fine. If your wiring is truly the issue, and you do your wiring repair job properly it WILL fix your issue. I promise.

alltracman78
10-29-2011, 11:05 PM
Ive personally confirmed with a DMM by ohming the shielding out, it terminates to the E1 ground on the intake manifold. Bruce rewired his KS with the same exact instructions from me and it worked perfect.


There's no need for that, the shielding is grounded to the engine ground wires at the ECU.

Of course it's going to work. A ground is a ground. You can ground the shielding anywhere on the car and it will work. You can run a wire from it all the way to the trunk, ground it and it will "shield" the wire. If you ohm out the shielding to the ground point in the L rear of the trunk it will give you the same reading as if you checked it at E1 [assuming your grounds are all good]. Same goes if you put the other end of the multimeter on the neg battery terminal. My problem is not that it won't work.
This is why I have a problem with doing it that way:
1-The shielding is already spliced into the E1 ground wire [one of the ECU grounds] near the ECU connectors. If you don't touch that and ground the other end of the shield you're running the risk of a ground loop. If you want to get rid of the connection at the other end you need to uncover that end of the harness, uncover the splice, cut it, cover that end of the shield, cover the splice and recover the harness. Unnecessary excess work.
You could just have 2 separate shields on the wire, each one grounded at 1 end. But that brings me to problem #2.
2-Grounding at the intake manifold vs the wiring harness inside the car exposes the ground to corrosion. Take a look at your E1 ground. Even if your car is a Southwestern car it will have some corrosion [unless the harness is new]. Not necessarily enough to cause a problem, but the shielding is more vulnerable to corrosion than regular wire. Splicing it into the existing shielding and covering it will help to prevent this. No point in chancing it when you don't have to.


You are retarded....

The shielding splices off the ecu side of the cable but thats really the only difference here... the brown wire that comes off the sheilded line then splices into the main ecu ground (which goes both to the back side of the manifold and the back of the ecu) ITS ALL GROUND DUMMY..

If you are installing a patch harness, or looking at a wiring diagram, the way he drew the picture is correct and functional. Its not like shielding is directional or something...
If you could [or had?] actually READ what I posted, I never said it wouldn't work, or that it wasn't a ground. I said there was no need for it to be done that way.
Questioning someone elses intelligence when reading comprehension of simple sentences escapes you doesn't speak very well of your own intellect......


I can confirm this too because ive torn a whole harness apart and am rebuilding it...

There's no need for that, the shielding is grounded to the engine ground wires at the ECU.
I can also confirm it, I've torn apart more 3SGTE harness than you've owned.

alltracman78
10-29-2011, 11:28 PM
Whoops, forgot this.

Jeremy, if the shielding fails it WILL cause a CEL. Ive seen it too many times for it not to be true. Do it the way Bruce said to, and you will be fine. If your wiring is truly the issue, and you do your wiring repair job properly it WILL fix your issue. I promise.

The ECU looks for a knock signal from .1-.9V. It doesn't care what the voltage is [as far as setting a code] as long as it sees something. If the correct signal is .2V and it's seeing .5V because of a contaminated signal it's not going to set at code, it's simply going to retard timing more. If it sees zero V [or possibly higher than 1.0V, but I'm not sure if that would set a code or burn out that part of the ECU] THAT'S when it will set a code [no knock sensor signal].
If you have no shielding on the entire wire, it may completely kill the signal [I already said this :)]. It also may not. The point of the shielding is to keep the actual correct signal pure and unchanged. Losing the shielding is more likely to contaminate the signal than completely block it. I'm not saying that if the shielding is damaged to not fix it, because it should be fixed whether it's causing a code or not. And maybe it is causing that code, I definitely don't feel like that's impossible.
What is more likely is the wire is damaged [or the sensor is bad]. What a lot of people don't seem to understand [you may, you know more about this stuff than most] is that if you have a wire that is damaged [or corroded] but isn't completely severed just measuring it with an ohm meter won't necessarily tell you that it's bad. An ohm meter sends such a low signal through the wire that 1 or 2 strands can be enough to give a "good" reading [I've had this happen to me on more than 1 occasion]. Then when that same wire is actually used to carry a signal those few strands aren't enough to correctly carry the higher load, causing a problem. Even when someone "repairs" it it isn't always fixed right; the crimp at the terminal is bad/damaged, or the repair crimp isn't done right.

Bottom line with all this is that yes, what Bruce outlined should fix the problem, and WILL work.
BUT, you're better off running the shield back to the original location for longevity purposes. And definitely don't run it to both ends.

tuner4life
10-31-2011, 01:41 PM
Wow, this thread got out of hand. Yes, the shielding grounds at the ecu. I was prepared to try rewire it all per hookecho's diagram, but my dad (a wiring/electronics guru/nerd) insisted that I let him have a shot at splicing it before I rewire everything. He got it spliced and even managed to splice the shielding somehow. there was only about an inch of shrink tube needed to cover the patch it was so precise and neat. Anyways, it works fine now, the CEL is gone and the car runs great. Thanks for all the help everyone.

Luni
10-31-2011, 06:34 PM
Jeremy the only problem I have with what you said is the ground point at the ECU connector ends up at the same point (intake manifold) as the location of the ground for the KS shield. So your argument about corrosion doesn't matter cause if that ground corrodes it affects the sensor the same no matter where you put it.

alltracman78
11-01-2011, 08:42 PM
the shielding is more vulnerable to corrosion than regular wire.
I THINK the shielding wire is steel. I may be wrong. It's not copper.
It's also spaced apart and not twisted tight together.
I'm assuming you can take it from there. :)


He got it spliced and even managed to splice the shielding somehow. there was only about an inch of shrink tube needed to cover the patch it was so precise and neat.
See? It's really not that hard. Even if you had done it it would have been less work than tearing that whole end of the harness apart.

I'm just trying to explain to you guys that there's an easier way to do this.
You could DEFINITELY cover the shielding [and ground] to help prevent corrosion. You could probably even just ground it there and it would be fine.
You could maybe even ground both ends and not have a problem. There could be a better place to ground the shielding that the factory location.
But stock works fine, it's not a problem as it's set up, so WHY tear apart the harness, ground to where it may corrode faster, make more work for yourself and have it potentially not do as good a job?

Luni
11-02-2011, 05:17 AM
Grounding both ends would definately be a problem. That would take on more interference than not having it grounded.

Truth be told, when I rewired the knock sensor for the harness thats going into Dutchess, I overlayed the shielding around the solder point I did and heat shrinked around the whole thing. I didnt even solder the shielding, I just kinda webbed it all together entwined and evenly distributed then shrink wrapped it to hold it all in. Then I ohmed it out from start to finish and I had perfect continunity with E1. I have no doubt it will work (He hasnt finished the swap yet).

But most of the time people have to rewire the KS cause the internal wire goes bad, or the shielding goes bad. Obviously you can use a DMM to determine which one it is, if you determine its the shielding, its far easier to just rewire direct to the intake manifold ground.

There are of course multiple ways to do it. The only thing you DONT want to do is ground both sides. One or the other. Doesnt matter. You can ground to the body of the ECU if you want. Just dont ground em both, or you WILL have a slight current going through the shielding, enough to pull a CEL. I know from personal experience :)