Toyota Forum banner

2VZ-FE code 52 -- knock sensor

13K views 55 replies 8 participants last post by  white90dx  
#1 ·
Got the dreaded code 52 today on the wagon, specifically "Open or short circuit in knock sensor signal (KNK)". I've been trying to find the part number for the knock sensor pigtail, anyone know what it is? I failed to find it in the diagrams.
 
Discussion starter · #3 ·
Does the engine go into limp mode at all? Sometimes the CEL stays off when I clear the code, I've noticed when the CEL is not on the engine idles properly, when CEL is on it idles too high.

Gonna have to test the wiring from the ECU to the connector just before it goes into the pigtail. If that checks out, then either the pigtail has melted (probably) or the knock sensor has gone bad (not likely). If it's anything like the 3VZ/1MZ and other Toyota engines the wire has heat damage.
 
Discussion starter · #5 · (Edited)
Think I might have found it, 82219-32060 less than $10. I called the dealer and told them to order it, if it's the wrong part no biggy. Contemplating if I should replace the knock sensor while I'm in there, don't want to they are not cheap.

That looks like it might be the right connector Charlie. I went through some parts I saved from the 2VZ/3VZ tear downs I did and found the knock sensory pigtail.

Image
 
Discussion starter · #8 ·
You need an oscilloscope to test a piezo sensor properly. If you put a test meter on the knock sensor and set it to A/C, rapping it gently on something solid will cause it to generate a few millivolts. Not an accurate test of course but it will tell you if the sensor is alive. Doesn't work for the newer knock sensors that are frequency specific.

The addition of the knock sensor on this engine gained what, a whole 3-4 HP? :lol: Why did they even bother.
 
Discussion starter · #10 ·
Well in the next few weeks I'll tackle this repair, such a pain in the ass on these engines. I'll be pissed if a new pigtail and sensor doesn't cure the problem but I've read about people that had this happen. The wiring going across the engine and to the ECU sometimes goes bad.
apparently it will retard to 0BTDC. that explains why my 2vz feels so underpowered....
You could try mounting a knock sensor externally like on the engine lift hook so the code goes away.
 
Discussion starter · #12 · (Edited)
Good link, will test things out in the next few days.
That does look like it should be the right part - though there aren't any good pictures of it online, oddly enough. Let us know how it goes. It seems odd to me that they have a separate short pigtail for the knock sensor, but I guess that allows you to take the harness off the motor without removing the intake manifold...

-Charlie
Forgot to respond to this. I wish the pigtail could be easily replaced, but the intake manifold has to be pulled. The intake gaskets are surprisingly expensive over $50 for the set.
 
Discussion starter · #15 ·
As for me, I'm glad my knock sensor is 'easy' to get at on the back of the motor, and away from the worst heat of the engine bay. :D

-Charlie
the 2VZ is a real pain to work on, worse if you have ABS then you want to shoot yourself. Imagine what a dealer would charge to replace a knock sensor. :lol:
 
  • Like
Reactions: John M
Discussion starter · #17 ·
Some people are completely unable to fix the issue, even going so far as to replace the sensor, the pigtail, the entire engine harness, and the ECU without luck.
That's what I'm afraid of. When you have it as an intermittent that spells trouble.
 
Discussion starter · #27 · (Edited)
That harness is discontinued unless you get lucky at specific dealers. Can't test the cable since it may fail when heated, replace it with OEM if possible. If you can't find it then de-pin the connectors and crimp and solder high temperature wires in place of the originals.

This is what mine looked like

.....that knock sensor says OE spec not actual OEM part. That's a roll of the dice. Original knock sensor part number is 89615-35040 I might have a used one kicking around I'll see if it looks like this one
 
  • Like
Reactions: John M
Discussion starter · #29 ·
That Ebay listing I posted is probably a fake, this is what the original looks like.

Image
 
  • Like
Reactions: John M
Discussion starter · #31 ·
Has to produce an accurate frequency and survive many heat cycles not a cheap part to make. Probably costs Toyota $25. That is not sarcasm the knock offs I'd say are made for $2.
 
  • Like
Reactions: John M
Discussion starter · #36 ·
When I had the code it would come and go which tells me if the ECU is getting a good knock sensor signal the code will go away by itself. I'd like to know why the sensor got so corroded is there very high humidity where you live? The pictures I posted are of a used sensor it was a bit greasy but otherwise looks almost new.
 
  • Like
Reactions: John M
Discussion starter · #40 ·
1988 ECU? It didn't have a knock sensor don't know what would be involved if anything when it comes to wiring differences.
 
Discussion starter · #44 ·
ECU looks for a waveform not resistance you can't fool it short of using a frequency generator.
 
  • Like
Reactions: John M
Discussion starter · #46 ·
A bit surprising, but it does look like Toyota went with a resonant type sensor. I wonder how knock disturbs the signal.... assuming it just increases the amplitude?

It wouldn't surprise me if Toyota put in an analog bandpass and RMS detector and fed that to a low-ish speed ADC input. Then it would just have min and max expected values for each RPM range.

-Charlie
Not sure how the circuit actually works. Consider this, if you disconnect the knock sensor you get a code right away not when knock happens. Which tells me the computer is always sampling the sensor.

Of course I could be completely wrong but that's my experience with the 2VZ, 3VZ and 1MZ.
 
  • Like
Reactions: John M
Discussion starter · #48 · (Edited)
That could be the circuit missing the 'pull down' of the sensor - basically, the signal goes to full-high when it should be somewhere mid-range. There isn't much on the sensor at idle in terms of AC signal. Testing involves revving the engine to at least 2k RPM and as high as 4k RPM.

-Charlie
The sensor has no conductivity it's an open circuit, if you rap it against something (do this gently) it does generate voltage.
 
  • Like
Reactions: John M
Discussion starter · #52 ·
I drove around for a few weeks with the ECU sitting on the floor mat.
 
  • Haha
Reactions: John M
Discussion starter · #55 ·
I had previously, to start to fix the error 52, removed the upper intake, removed and inspected the original knock sensor, swapped it with an OE (see previous posts), and, while I was there, tested the knock sensor pigtail the same flexing way. That pigtal, and the wiring harness, are 100%. There is nothing wrong with them.
Intermittent failures can't be tested with an ohm meter, because well the problem is intermittent. Highly probable your knock sensor harness is bad. I would get a 1988 ECU it will probably work and didn't have the knock sensor circuit.