Rage

Does the GM/Merc engine knock sensor.....
1.) emit a signal only when a "knock event" has been detected or....
2.) emit a signal when any engine vibration of sufficient amplitude is detected?

Please only respond if you are positive about the information.

GOODT

here you go,cant get more positive then straigh out of the manual

Rage

Some additional info.

The Merc knock sensor is a Delphi unit, product sheet attached.

Under product sheet "Description" "...it produces an output voltage in proportion to the engine vibrations caused by knock. When knock is present on the sensor signal, the PCM controls knock by retarding spark timing."
The "When knock is present on the sensor signal" seems to indicate that the sensor can have other vibration signals on the "sensor signal" in addition to knock signals.

Under product sheet "Performance Advantages" "...has a flat frequency response over the range of 5 to 18kHz. This allows the same sensor to be used on different engines by simply adjusting the filter frequency of the signal processing electronics (PCM) to match the knock frequency of the engine. In addition, the sensor responds to knock frequencies higher than the primary knock frequencies (harmonics). This allows the higher knock frequencies (harmonics) to be used by a control system (PCM), either individually or combined with the primary knock frequency."

This seems to indicate that the PCM has to be programed to recognize what is a knock signal for the particular engine in question rather than calibrate the knock sensor for a particular engine's knock frequency.

Maybe Merc has a custom knock sensor????

In theory, if we knew the "knock frequency " of a 496HO an electronically savvy person could design a signal filter to hook up to the knock sensor output to filter out all but the knock frequency signals and then data log the filtered voltage signal as "knock events"

Anyone know what the knock frequency of a 496HO is?

audacity

note sure what you are looking for???

a knock sensor is nothing more than a piezoelectric element that out puts a voltage. your EEC/PCM takes that voltage and looks at its frequency and amplitude. most engine calibrations that look at this output do so during partial and steady state throttle. not WOT. just as other adaptive strategies are not used during WOT.

again, not sure what you are looking for but it's a lot more complicated than it may seem on the surface.

Dave_N

What about using a stand alone system like this?

http://www.msdignition.com/tool_8.htm

Dave

Rage

audacity,

Please see the thread 'Looking for expert on Merc PCM555 ECU' in this group which will explain why I am trying to find a means to monitour/record 'knock events' for a PCM555 eqiuiped 496HO.

It is simple with a GM ECU which provides an output sigmal of 'knock events' that cam be recorded real time along with other engine parameters. The PCM555 does not. I am looking for a workaround to get essentially the same "knock event' info from the PCM555 unit.

It would seem to be very problematic for an engine if the PCM555 can not monitor 'knock events' at WOT as you state. Am I wrong?

Rage

Dave,

Thanks for the continuing suggestions.

I also found this item with the help of Google. The problem is in the fine print: "...Since all engines make noise, some more than others, MSD added a sensitivity adjustment to the Alert. This adjustment allows you to "tune" the Alert to match your engine's characteristics...." What this says to me is that YOU have to calibrate the MSD knock detector to recognize what is and what is not the frequency of a "knock event' for your particular engine. This might work in a vehicle where you can hear a knock event when it occurs and dial in the unit to warn when that frequency occurs so you can retard the timing as they suggest. But if you can hear the 'knock event' what do you need the detector for?

Snake oil.

I think my only hope is to some how monitour the injector pulse width for abrupt increases that the PCM555 will generate in response to a 'knock event'. The LM-1 o2 sensor reader and data logger that I have only has six auxilliary data inputs and I will need eight since the PCM555 adjusts injector pulse width in respose to a 'knock event' on a cylinder by cylinder basis.

audacity

"It is simple with a GM ECU which provides an output sigmal of 'knock events' that cam be recorded real time along with other engine parameters. The PCM555 does not. I am looking for a workaround to get essentially the same "knock event' info from the PCM555 unit."

why????

they don't work that well in a modern-day vehicle with a very advanced EEC.

Rage

audacity,

I will be increasing the flow of the intake manifold, heads, and exhaust, And installing a hotter cam and associated hardware in a stock 496HO. Target ~500-525hp. With this will be associated an increased risk of engine knock with the stock PCM555 ECU settings. I will have A/F ratio recording capability along with other engine parameters with an LM-1. A real time knock event data stream would obviously be invaluable in the tuning process which is why this line of inquiry. Before I pull the trigger on this project I want the means in place to accomplish it. This is pretty much the last hurdle unless more pop up.

audacity

real time knock data is not an 'invaluable' tool what-so-ever. get yourself some dyno time.

spend more of your time on more important things.