knock sensor help

[fullmetalx]

Problem: code 34(knock sensor and/or circuit) keeps on reappearing. Clear the codes and after a test drive. shut off engine recheck codes and code 34 pops up again.

Vehicle: 1991 240sx se, no mods, ecu# 26, 89 octane gas

Procedures already performed: replaced old knock sensor with a new one because old one has it's housing was melted (old one still good though, reads ~520 ohms), checked for continuity on the harness between plug on knock sensor and at the ecu and it is good. checked continuity of knock sensor to ground and it is good. idle is stable ~750 rpms, timing is at 20`BTDC. Car runs fine but not the best. Been this way since i bought it. Puff of bluish white smoke coming from the exhaust once in a while usually from a first time revving or start to accelerate in 1st gear.

I'm stumped as to why the code keeps on reappearing. could there be another reason or another part that could trip off this code. Would my lower octane gas (89) do it. Highest here in san diego is 91 octane. should i try that? I've searched thru Zilvia and the FSM. I've ran out of ideas. Any help would be greatly appreciated. Thanks.

[johngriff]

Try the higher octane gas.

White/blue smoke may be a sign of a real knock occurring. When you tested for continuity did you do it on the contact side of the pins. I have seen alot of bad contact pins on the ecu plugs as of late, where you get a good reading from the back, but a bad reading from the contact side. (steve injector?)

Timing was setup @ 1000rpm w/tps disconnected correct?

Do a compression (leak down if possible), just to double check the blue/white possibility of a chamber related failure.

[daryl337]

also, try to retard the timing a little bit.

[Z U L8R]

20 degrees? aren't ka's supposed to be 15 degrees btc? shit load of timing + lowest octane possible = knock

[gunluvS14]

no.. factory base timing is 20 BTDC

[Z U L8R]

i checked the fsm it said 20degrees. i stand corrected, thx =]

still. you may want to retard it a degree or two. you may have a lot of carbon build up on the tops of the pistons, also the little bit of bluish smoke is most likely blow by and excess oil in the combustion chamber also heightens the chances of detonation..................but....i just had an epiphony (spelling?)

i bet the little 90 degree elbow on the bottom of the diaphram that's part of the EGR system. i bet that hose is cracked/burnt/broken. therefore your egr valve is not receiving vacuum. therefore your egr valve is not opening allowing exhaust gas to recirculate back into your engine thus lowering your nox, therefore the cylinder head temp is raised causing your car to spark knock therefore setting your infamous knock code. that 90 degree hose always breaks. if you can't find it/have no idea wtf i'm talking about or checked it and it's ok, rev your motor up with your other hand on the top of the egr valve. curl your fingers and feel if the egr valve is opening as you give it a little gas. oooOOOoo bukake ^_^

[fullmetalx]

Problem solved. code 55 now.

I did a little more reading on zilvia and read some post by mr. ps13silvia. he said to check for a short on the harness between the ecu and to the other end of the plug by the knock sensor. he said that it's a common problem he has dealt with a few number of times on ka24de engines.

Lesson learned. i know yesterday i checked for continuity but what i didn't check was voltage. I shoul've looked into that also. there could be contuinity but the right amount of voltage is not coming through, hence there could be a short somewhere. now i didn't want to open up the entire harness throughout the top of the fuel rail. so i made a jumper wire from the ecu pin #27 straight to the knock sensor itself. cleared the codes, checked timing with tps off at 20'. test drove car. also noticed bluish white smoke at the beginning of acceleration is gone. came back shut engine off. recheck codes and code 55 shows up. whew. fixed it after almost 5 yrs of running like that. my old knock sensor could have been still good eventhough the housing was still melted. it was just a short in the line. wasted $60 on a new one that could've been put to better use.

Thanks everyone for the replies.

[johngriff]

Awesome.

Wasted $60 on a new one that could've been put to better use.

That is why I always say, Test, Test, Test the wiring! Good News.

[fullmetalx]

here's a pic of my old knock sensor. it's still good i believe reads ~520 ohms. i doubt anyone would put this in their car back but i'll give it free to whoever that might need a working one.

[Z U L8R]

lol cool. i was rechecking the thread to see if you needed more info on testing your egr system. glad everything worked out. i got in a big debate about your car with the matco tool guy. he couldn't understand how egr system malfunctioning could cause spark knock. to which i replied, that's the difference between a B tech and an A tech rofl

glad u had the guts to tackle that.
Dave

[UNISA JECS]

Problem: code 34(knock sensor and/or circuit) keeps on reappearing. Clear the codes and after a test drive. shut off engine recheck codes and code 34 pops up again.

Vehicle: 1991 240sx se, no mods, ecu# 26, 89 octane gas

Procedures already performed: replaced old knock sensor with a new one because old one has it's housing was melted (old one still good though, reads ~520 ohms), checked for continuity on the harness between plug on knock sensor and at the ecu and it is good. checked continuity of knock sensor to ground and it is good. idle is stable ~750 rpms, timing is at 20`BTDC. Car runs fine but not the best. Been this way since i bought it. Puff of bluish white smoke coming from the exhaust once in a while usually from a first time revving or start to accelerate in 1st gear.

I'm stumped as to why the code keeps on reappearing. could there be another reason or another part that could trip off this code. Would my lower octane gas (89) do it. Highest here in san diego is 91 octane. should i try that? I've searched thru Zilvia and the FSM. I've ran out of ideas. Any help would be greatly appreciated. Thanks.

FYI a good knock sensor reads 560 Mega Ohms thats 560,000,000 ohms