rb20 knock sensor problem

[jayplanedin]

so ive been having a hesitation above 5200 rpm for a while and i think ive got it figured but want to double check with the smart guys on here.

i have a 91 240sx with rb20det... basically stock without fmic and running 12.5 lbs of boost.

i checked the ecu and its giving me code 34(knock sensor). i have gone through every single thing on the "fix your hesitation" list and no results. i am running perfect a/f ratio and i know my coils are good... running the proper spark plugs, have enough fuel(obviously since no leaning out) ect ect...

so heres my hypothesis. ive been reading up a lot on nissan motors (sr's ka's vg's and b13 motors) and it seems that the obd1 versions of the motors apparently commonly will throw the code for the knock sensor, and although the resistance of the knock sensor is fine, the pinout on the ecu is fine, and after replacing everything, it still gives the code.

i have only found this on 2 or 3 threads which do not involve rb's, but the ka's and b13's. apparently the harness(being a shielded wire to eliminate some interference) can wear out quite bad and actually make the wire have a higher continuity and therefore cause a code to be thrown. additionally, the ecu apparently will only throw a code when the knock sensor or wiring is faulty, not when the motor is knocking. and as a result of this shielding killing the connection, replacing with ordinary 16 gauge wire can actually completely eliminate this problem (as some have stated).

i am just wondering if anyone on here has heard of this with rb's or could confirm any of the information written above. ive searched all of gtrc and google and not come up with anything.

[Skym]

Fault codes are triggered by incorrect voltage from sensors. You can stop ECU reporting the error code if you have Nistune, Nistune tuning software, laptop, consult cable, consult plug wired in, etc.

It's common for wiring that's close to engine to get heat damaged, brittle (common problem with injector loom, coilpack loom that can cause missing). Usually loom needs to be rewired or replaced. Or knock sensor gets damaged by heat. Have you reset the ECU to clear the error code??? If you know someone with Nistune tuning software, cable, can tap knock sensor with engiine off, but 3 on ignition barrel and note if the knock count level rises on laptop.

[jayplanedin]

I have tried resetting the ecu and no change. I figured I'd start with the loom first since starting with the free/cheap stuff is usually better than the 150 dollar sensor lol. So it is a common occurrence that the knock sensor wiring becomes faulty with time/heat?

[hotrodstu77]

I think I have the same problem but on start up. I have nistune and it seems when the car is cold it richens itself out like crazy and seems like the timing is badly off. But usually once it gets warm it will be completely better. And I'm pretty sure my ecu is setting a knock code as well.

Skym you said something about you can tune this out with nistune? How would I do that??? This has been a bad problem for my car and it's really pissing me of cause I keep fouling plugs. Can ya help me out? Do you need me to send you my bin? I'm gonna try unplugging my sensors tonight and see if it changes anything.

[Skym]

That's sounds like warmup ignition, fuel tables. They need to be retuned with mods on engine. Even just upgrading exhaust can make engine do that weird start part due to running too rich. First start fuel table is usually ok, but when you let go of ignition key, ECU switches to after start fuel table (I think there's an ignition table to suit) and that's where engine bogs, struggles to start. If you crank engine longer to get pistons pumping, sucking air in, engine starts better which confirms it's running too rich.

Should be the same tables on any ECU, including aftermarket ECU's. From what I understand you lean them out, adjust ignition timing tables to suit until engine starts correctly (make sure you have wideband, so don't lean tables out too much, as engine has to run rich during warmup). But has to be done when engine is cold, as get one chance to do it, so has to be done over a few days if adjustments are needed. Most tuners have worked this out, so use settings they have done in the past on same engine that work and fine tune if it needs it over 1x night. That's if they tune it properly. But that comes down to what customer wants, sometimes they don't get the chance to. But one thing I haven't tried is resetting ECU and see if it allows you to go throught the first startup again.

Best way is to datalog A/F rato via wideband during warmup one day, do adjustments to fuel, ignition to suit, then see if they work the next day and adjust if necessary. Then check again the next day, etc.

But that's just startup tables, there's other tables ECU uses when moving that need to be retuned and they are the tables which pull the engine back via pulling ignition timing (protect engine) during warmup until certain coolant temp is reached. Very time consuming, as supposedly took factory months to get it right. But that's working from scratch.

I would suggest you find a .bin file from a engine with similar mods that's already been retuned and see how they adjusted those tables if you haven't done it before.

Also if you haven't already, cleaning AACV and IACV help (idle valves). Especially if turbo has failed in past or engine has done high km or car is on racetrack (higher rpm, higher boost levels than stock) without catchtank, where oil can enter idle valves and clog them. But if engine starts fine when warm, or after you start engine after turning engine off after first start, then it's the ECU tune.

Jayplanedin, common age related problem with wiring due to heat. Most replace looms with new looms or rewire looms when restoring cars.

[hotrodstu77]

i think ive figured a bit of it out skym. thanks. my engine is actually failry modified and has been tuned by a reutable dealer but my car wasnt running perfect when he tuned it so now that i have fixed somethings its making the tune go off slightly. so ill just keep messing aroun witht he enrichment tables and see what happens.

[jayplanedin]

Update. I recently Bought an innovate lc-1 and found some odd things. The rb will run totally fine at times, then, when it does the hesitation at high rpm, it seems that the afr during idle at those times is extremely lean(16-18:1). This causes a rough idle and the terrible hesitation at high rpm. Any ideas? It will run fine for an hour or two then out of no where, runs like **** and is lean.

[Skym]

Check fuel pressure at idle (around 35.5psi is normal with vacuum hose to FPR connected). If low fuelpressure problem doesn't show up, check fuel pressure, A/F ratio on dyno, but be careful and back of quickly if see A/F ratio go lean. Failing fuelpump, fuelpump wiring problem or faulty FPR could be the cause.

[jayplanedin]

Well just installed brand new walbro so I'm not suspecting it's the pump. How does a fpr go bad? Clogged with gunk?

[Skym]

I think the insides fail over time (spring, diaphram, etc). I think the test for FPR is pulling vacuum feed to FPR and see if fuelpressure jumps 7.1psi on fuel pressure gauge (handy to have a fuel pressure gauge or rising rate FPR with fuelpressure gauge installed on car). On one car I noticed the tuner mentioned he had FPR problems with bigger fuelpump (adjustable FPR) and found that pulling vacuum feed to FPR solved the fuelpressure problem.

Also some fuelpumps can make engine go lean at higher rpm's if there's not enough voltage to pump. I think Walbro's are one of those fuelpumps. Also Walbro's don't like being run dry, so keep fueltank above 1/4 on fuel level gauge.

[jayplanedin]

I'll give that a try. Where should I connect the fuel pressure gauge? After fpr on the return? Or before it on the feed side?

Also, what is that second fpr looking thinger at the back of the fuel rail? It connects to the iacv?

[Skym]

Compound fuel pressure gauge or similar between fuel filter, fuel rail.

From what I understand, the second FPR is a damper or supressor as some call it, which keeps a steady flow of fuel inside fuel rail to injectors, even when fuelpump changes the pressure while engine accelerates. I gather it's similar to what happens in dual entry fuel rail.

On dual entry fuel rail, the fuel entering both sides of fuel rail does a similar job, as the fuel coming together in middle of fuel rail, (where fuel exits fuel rail), cancels each other out and provides steady flow of fuel to injectors. It's like putting 2 speakers together (aiming at each other) and using the same sound on both speakers, which cancels out the sound. At least that's how I understand it.

[jayplanedin]

Tested fuel pressure today. With walbro ran at ~42 psi. Removed vacuum line from fpr and jumped ~6 psi. Any suggestions on what to check next

[Skym]

Dyno run with A/F ratio check. Sometimes lean A/F ratio can mean the exhaust manifold is leaking and oxygen is entering exhaust manifold and wideband is reading incorrectly. Or intake manifold gasket between head, intake manifold has blown (common problem on GTS, GTR) and getting a lean A/F ratio.

[jayplanedin]

Mani gasket was changed less than 1000 km ago. Still doesn't explain the codes, and the terrible idle when leaning out.