Ive just finished wiring in a knock sensor today (using shielded wire thats grounded at ECU end (i used an eye connector and grounded on an earth). The ECU (after clearing the memory and re-running a diagnostic) does not show up the knock code anymore... so i have no codes showing up. I set it back to regular mode, and started the engine, and went to check timing. Now, i dont think this is directly related to the knock because i was having a bit of trouble getting it to sit on 15 in the past, (it would move around with idle speed changes). So, i give it my best attempt at setting to 15, then it flys up to about 17-20, then back to 15, i lock it in, then after a bit, the idle slows down (without me doing anythiong) and it is sitting on 10. .... so i try again, same deal.... I know the whole "knock takes off 5 deg" thing when its detecting knocking, but cmon... at Idle.... I also read heaps on them being pretty unreliable, but this was off an R33 RB25DET so its kinda new-ish... and i tested with a multimeter and tapped the end and it was producing voltage. so, i guess i have a few questions... why is the timing changing with engine speed differences? and how does the ECU know how to change what seems like a mechanical thing? and finally.... WTF is going on hehe anyone else have these issues or could give me some advice on setting the timing.... it shouldnt be this hard... ---------- Post added 18-01-14 at 12:15 ---------- strange, i went and did it again, after letting it sit off for 10 mins, and it let me lock in 15 this time, and i reved it a few times after, and rechecked ... still on 15.... Some original photos have not been recovered yet.

The ecu changes timing with load scale and RPM. If you look at the picture below (standard VLT timing map) you can see that if you get a bit of TPS creep, say up to 20 and the rpm goes up a touch to 1000 RPM, bang you're at 24 deg timing. Now the ECU works in a '4 cell' window so it interpolates around a bit so that's why you see 17, 19 etc etc. Basically to set your timing you need to have the RPM <800, the workshop manual states this, 750 RPM I think is stock idle. At this RPM the ecu should be looking for 15 deg timing and that's what you should see at the crank. Of course if you are trying to ghetto a high comp setup with a bit more boost you obviously want to set it at 10 deg at the crank etc...or 15-whatever advance you are trying to take out. Obviously this trims the entire timing map which isn't ideal, but ghetto high comps be ghetto haha. Not suggesting you are doing this. Some original photos have not been recovered yet.

The Ecu won't use the TADVH (IGN timing map) at idle unless the TPS sensor is not set correctly or faulty. it will only use the Idle/decel table which is set to 15deg till about 1000 rpm then it goes to 18deg up to about 1600rpm. Make sure 2 of the 3 TPS contacts is making short circuit at idle and goes open circuit at a touch of the pedal. from memory its middle pin as the common. the other pin will go short to middle when the throttle is about 75% open. also Check TPS connector for corrosion. (Black Box Bosch style connector not the grey one dangling off with grey connector)

wow, thanks for the info guys.... that's much more than I expected and not the usual "set it at 15 bro!" type stuff. I was also thinking perhaps the knock was giving false positives and interfering but im thinking there was something to do with the idle speed and possibly that it had "warmed up" (been running for 15 mins or so while I did other stuff) it was running a little higher rpm or in a different mode? and after letting it sit and re-doing it, it was in a more stable state for setting the timing. I'm still a little foggy on how the ECU can change a physical setting of the timing, I thought the cam spun a gear / shaft / CAS wheel etc. but don't worry about explaining, ill just believe you And yes, its a ghetto setup, however its using all legit VLT parts with correct ECU, Injectors and my engine although high in km's has all working sensors and plugs (I did all these diagnostics and tests using multimeter etc. long before starting this project). Some original photos have not been recovered yet.

come to think of it, One other cause for ignition timing bouncing around can also be the CTS (2 pin coolant sensor or connector as they often suffer electrolysis & corrode. The engine could also be jumping to and from warm up timing tables. easily checked by disconnecting CTS and allowing engine to warm up for no less than 4 minutes and checking timing again. This may help or just confuse ignition timing: 1. The CAS or Dizzy sensor sends 2 low voltage digital signals to the ECU (60deg & 1deg of cam rotation). 2. The ECU makes a decision based on CTS (Coolant Temp), TPS (Throttle position), TADVH (ignition 3D map) & RPM (revs per min) as to when it will trigger the Transistor (ignitor) attached to the Dizzy's body. 3. The ignitor's (transistor) base pin receives the trigger from the ECU, then grounds the ignition Coil to cause the coil to induce a current (good ole Lenz's law - magnetic flux yadiyada). 4. The coil will naturally find a path to Ground, the spinning rotor button in the dizzy is perfect as it passes other spark lead terminals it will find the shortest path to ground through one of them as the spark plug is attached to the block at the other end.