Quote: Originally Posted by CIVICEATER
Its not showing i have any knock retard on the aeroforce. I do have w/i with an m3 nozzle and spiking 21 holding 17. My timing is about 31 at redline mods are in my sig.
It's kind of a pet peeve, but this thread is already a year old and I bet some people have upgraded or changed things on their car over a year. So unless your car will
never change it's a good idea to put your modifications into your post instead of referencing your signature line.
That way in another year if someone reads this it will still make sense.
With that said your "performance" modifications:
Quote:
2003 Solar Yellow SRT-4
E1 Turbo
Kinetic WGA
Devils Own W/I with Progressive Controller
MPx 3" Exhaust with Cat
Walbro 255lph Fuel Pump
Stage 2 Injectors
Back to your statement. If you're watching the knock retard degrees at the same time you're also watching the raw knock voltage and seeing the PCM not pulling any timing but still getting voltage from the sensor it might be from how the knock detection works. The sensor (microphone) will pick up a lot of noise which is why most PCM's process it and listen for certain characteristics at certain times. For example some cars will essential "turn off" and not use the knock sensor at very high rpm's because there's too much noise to accurately determine if the knock is real or not. The torque peak of an engine corresponds closely with the peak cylinder pressure (which is where detonation is mostly likely to occur), so if those cars had a torque peak in the middle rpm range that's where they would most likely develop knock. The PCM may be programmed for this and if it's not seeing knock in the upper mid-range it won't do anything at higher rpm's where there is less load on the engine and less chance for knock even if the knock sensor is picking up and producing high voltage reading.
I'm not saying that's the case in the SRT-4 but it should give you an idea of how some sensor-based knock detection systems work and why raw sensor voltage doesn't tell you much. It's very well possible to have knock voltage and no knock. With that said if you can watch two variables at the same time I'd mostly pay attention to knock retard degrees and overall timing advance. The raw voltage may be helpful under a couple circumstances.
Now let's say you had an aftermarket big turbo upgrade on the car but were still relying on the stock or "Stage" computers to run everything along with a piggyback. In general most PCM's determine knock by listening at certain times or rpm/load conditions when knock is most likely to be present, they listen for a certain frequency depending on physical engine characteristics (primarily cylinder bore size), and by the voltage produced by the knock sensor. If all those match up it will determine there's knock.
When adding a big turbo you can alter the torque curve and torque peak of the engine quite a bit but the PCM knock algorithm might still be programmed to be more sensitive for the stock turbo torque curve. If the PCM was using engine load (which is calculated using the MAP sensor voltage) to know when knock is likely to occur that could also be affected by a piggyback altering the MAP signal. About the only variable that upgrades wouldn't affect is the knock frequency since that's determined by the engine characteristics.
This is all speculation on the SRT-4 and generalized statements on how most knock detection systems work in modern PCM's but it does bring up a point that knock
could potentially be not getting picked up or the PCM isn't as sensitive too it as it should be. Or it's picking up knock in the mid-range and applying knock retard and continues to apply that amount above a certain rpm range even if knock gets worse.
To test this hypothesis of what could happen (since we don't know the knock algorithm used) if you are getting heavy solid knock voltage and the PCM is not retarding timing you could increase the fuel octane (i.e. use race gas) or turn the boost down slightly and see if that voltage drops. If it does there may be knock present the PCM isn't or can't catch. Before worrying about that too much I'd make sure the knock sensor operation appears to be functioning normally, it's properly torqued, etc. Also if the knock voltage is overall high but all over the place and spiking around a lot this won't be very useful. This would only be good to test if it is always holding a high voltage under the same conditions.
Even better would be to put the car on a dyno that can load the car to simulate road conditions (in other words an eddy current Mustang-style dyno vice an inertia Dynojet-style dynamometer) and use a mechanic's stethoscope to listen to the block (if you can safely do so with a long enough stethoscope such as from inside the car). The human ear is pretty good and it's really easy to notice detonation that electronics sometime can't get right.
Linear Mode
