Nathan_F

OK posted here before. Got some excellent advise. Thanks to the guy who suggest I try to fix my VAF. Was able to do so. Any way 1989 Toyota 4x4 extra cab, base model 5spd, 3.0. Code 52 removed knock sensor, checked with a volt meter, like was suggested here. Good. Checked pigtail lead, fairly new, as is the sensor, no resistance. Checked, shielded wire, good continuity, no continuity to ground. Sensor is not Toyota. Wanna drive this thing, new engine just got running.

Nathan_F

Just checked another way. Used ohm meter between ground side of shielded wire and ground. Good continuity. But this is unplugged from the ECM ???? Normally grounded thru another circuit? Appears that shield is connected to 2 other shields and then to a brown wire to the ECM plug. Really stumped now.

scope103

How did you check a knock sensor with a voltmeter? There IS a way to do it (sorta), but requires an RMS voltmeter and a vibration source.

I've never disassembled the actual KNK wire and shield, but my circuit diagrams show the KNK signal goes to the ECM through connector B, but the grounds connect directly to a Brown wire which is independently grounded at the camshaft bearing cap. So "unplugging" connector B doesn't affect the shield ground to the engine. (Also, connector B doesn't contain the ECU ground.)

Nathan_F

I set my meter on the lowest a/c voltage touched the negative to the threads and the positive to the connector terminal and it registered 25mV when I tapped it with an end wrench it jumped to 50mV then settled back to 25mV. I performed this same test with the sender in place and the pigtail connected, only ground probe to block and positive to the pigtail. 2 wires. One is a dead end the other checked about 30mV and when I tapped on the intake with a screw driver it jumped to 100mV. I don't know what the parameters should be. Any answers. Don't want to buy another sender if I don't need to.

scope103

Good job. That isn't an "official" test (the sensor is "tuned" to a certain frequency), so there is no "parameter," but I would say your sensor is working. (The dead wire on the pigtail is the ground, and it is not connected at the sensor end. Only at the ECU end.)

Nathan_F

Didn't follow you on the ground conversation in first response. There are three shielded wires that have the ground shield connected to each other and to a brown wire that goes only to the ECM. Unless it comes back out at another pin it doesn't go to the can cap. I also read somewhere that if the TPS has an issue that it might cause a code 52. Is this possible. Next step for me might be to bypass the knock sensor in the truck with another externally mounted. This would definitely rule out the sensor that I have and the pigtail. Any other suggestions? Had this thing apart 3 different times (for various reasons). Getting good at it but don't need another habit.

scope103

My schematic (which is not the same as looking at your wires) shows that the knock sensor, and two O2 sensors (if you have a Calif. vehicle) have shielded wires. The shields all end up on a brown wire that connects to lots of places, including the cam cap, and also to one of the plugs on the ECU. But there is no particular reason that ground wiring in the harness will exactly match the schematic. Ground is ground.

Sort of. The tricky part of the knock sensor is that the shield is only grounded at one end (the ECU "end"). It is not connected to ground at the sensor (that's why the sensor connector only has one pin, but the pigtail-to-harness connector has two). This is probably to prevent what audio guys call "ground loops."

I know that people occasionally mention "bypassing" the knock sensor with one mounted elsewhere (like a lifting hook), but I just can't imagine that would ever work. (Why would Toyota mount it in such an inaccessible location if the lifting hook would work?) But I can't prove it won't work, so you're welcome to try it, but please report back.

Shady Cadence

Get a timing light, and a tachometer if your truck doesn't have one. Connect the tools and get the truck to operating temperature. Have someone hold the throttle at 2000 rpm and note the position of the timing mark. Now tap on the exhaust manifold with a wrench while watching the mark and see if it retards. If it does not, try replacing the knock sensor(it could be weak).
If it still throws the code, or if it DOES properly retard timing yet still throws the code, replace the ecu.

HappyCamper

If you get the code, doesn't that mean it's working? That it actually detected a knock and protected your new motor?

When I replaced my motor, I had drained all the gas because it had been sitting for a while, filled it up with new low-octane gas, set the TPS and timing, and was getting code 52. I added some octane booster (because tank was full, otherwise would have just added some premium gas) and the problem went away. I've been running mid-octane ever since. I've been meaning to try 1/2 tank of low octane and 1/2 tank of mid octane to see what happens but haven't gotten around to it yet.

PS. I've got to remember Shady's method -- excellent!

Shady Cadence

Code 52 is defined as "Open circuit in knock sensor signal (knk)", meaning the ecu is not seeing what it should. The knock sensor is Piezoelectric, meaning it generates voltage from vibration. When the engine starts and runs, the knock sensor generates a very small signal which will spike larger if detonation occurs, prompting the ecu to retard timing. If the ecu does not see that small signal on startup and while running, it will throw code 52.

scope103

That is exactly wrong.

Shady Cadence's explanation is spot on.

Lots of people seem to think that code 52 has something to do with the engine knocking, so they retard the base timing or use high octane gas. A complete waste of time. Code 52 means "the ECU doesn't know if the engine is knocking (or not)", so it dramatically retards the timing in a desperate attempt to save the engine (in case it IS knocking).

HappyCamper

Thanks for the correction! Since I had gotten that code, I'll replace my knock sensor (cheap insurance). Perhaps that's also the solution to the OP's problem. Additionally, I wouldn't trust a cheap aftermarket part in such a critical function.

50kta

this code 52 is bull crap im ssooooo tired of dealing with itbut im going to try shadys test tomorrow. fingers crossed

Shady Cadence

Well, where did it bring you?

catwrangler

iTrader: (1)

93-4rI have a new Toyota KS mounted externally, shielded cable wired directly to ECU harness 5 inches from ECU , new Toyota Pigtail, new (reconditioned/tested) ECU...still have a code 52...WTF, so good luck.
The only thing I can think of is a bad ground somewhere. Hope you find your problem

Shady Cadence

What do you mean, mounted externally?

catwrangler

iTrader: (1)

Secured under the hood, but external to the normal mounting point under intake.

Shady Cadence

Is it firmly attached to the engine?

catwrangler

iTrader: (1)

secured using two metal hose clamps to engine lift point(I didn't want to drill a hole as others have suggested).
I'm doing my timing belt and valves this week so while I am in there I am putting everything back to normal using the new parts.

Shady Cadence

Put everything back where it's supposed to be and you shouldn't get a code. Attaching to the engine lift point isolates the sensor from the engine noise, and it won't generate the small running signal the ECU looks for.