When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Archaelogical dig finds knock sensor in worst shape of all time
There is a quick story.... during the stroker rebuild, the last parts needed were on back order and only one knock sensor showed up before we were getting the engine together (other lagging by a week to 10 days) so wedecided to put the new one up front and the rear one could be the best of the two that could be changed at a later time (because some have said it was changeable without pulling the intake) . well the part came in, the car dynoed well without the new one in there, so the part sat and sat for many years...... 10 years to be exact. i tried to take a look at it on occasion, but all i could see was a lot of dirt and sand from laguna seca, and parts of the knock sensor that had broken off... finally during my last quest of finding a slight miss , and changing out injectors, plugs, caps rotors, etc, i had the fuel rails off and it prompted me to try and get to the knock sensor. when i saw it, it was pretty nasty. there was nothing left but two little wires attached to the exposed "guts" of the knock sensor. the wires were always ugly with most of the insulation gone exposing the wire mesh shielding.......... so now, affter some contorting of hands and body , and using a 13mm swivel socket, i was able to free the remains ......... in my hands the connection broke with one of the two wires to the sensor. I cleaned up the area as much as i could and the new one is in there torqued as hard as i could muster with a bad angle , but near the torque spec.
Last edited by mark kibort; 08-20-2018 at 01:18 PM.
what is the chance that the frayed wires , holding on by a thread, could still show the ecu that it was connected? ive always had a question, of the HP effects of the timing pull back at the higher RPM ranges. its known that it hurts power, torque at the lower RPM range 3-4kRPM range, but what about the higher range? could that actually add hp ? ive never tested it on the 928 and know we have seen some experimentation , but i cant recall the results.
thanks for posting
this is what they look like new. New one installed now.. what a pain. take the maf off, then losen the Throttle body rubber adapter "u", and squish it out of the way. the Throttle body is just above a quarter of the 13mm bolt, so a normal socket wont work, open wrench wont work, you need a swivel 13mm and it barely fits. way to go porsche, 2 more mm and it would be easy.
what is the chance that the frayed wires , holding on by a thread, could still show the ecu that it was connected? ive always had a question, of the HP effects of the timing pull back at the higher RPM ranges. its known that it hurts power, torque at the lower RPM range 3-4kRPM range, but what about the higher range? could that actually add hp ?
Possible. If the wires are still intact then that knock-sensor might have been working fine all along, or it might have been dead from the beginning. The problem is that without some way to check, you just don't know. If EZK detects a sensor fault (knock or hall) then it pulls 6-deg of timing under load. I have never heard of that adding power...
So one functional test is to do a timed 40-80 mph pull (or whatever), then disconnect one of the knock sensors and do it again... should be slower. If not, then EZK is detecting a sensor fault-- but you still don't know which one.
Possible. If the wires are still intact then that knock-sensor might have been working fine all along, or it might have been dead from the beginning. The problem is that without some way to check, you just don't know. If EZK detects a sensor fault (knock or hall) then it pulls 6-deg of timing under load. I have never heard of that adding power...
So one functional test is to do a timed 40-80 mph pull (or whatever), then disconnect one of the knock sensors and do it again... should be slower. If not, then EZK is detecting a sensor fault-- but you still don't know which one.
thats a great idea. .......i dont know why i didnt do that.... i guess i still can. i have TONS of data on 60 to100mph runs in 3rd gear 3500 to 6000rpm. that wire that was attached fell apart off the knock sensor just on route to the table. ill see if i can get a voltage out of it by hitting with a hammer. back to the power issues, most chips pull timing out at high RPM, right dont know if it is 6 degrees, but some for more HP at speed. has anyone done a dyno test of knock sensor connected vs not connected?
It's a Piezo sensor so it's unlikely you'll be able to see the millivolt audio signal response with home-mechanic tools. And DON't hit it with a hammer. It's a microphone after all. You'll get the same response from it tapping the bolt with the tip of a screwdriver. And you still won't be able to see it on a DMM.
It's a Piezo sensor so it's unlikely you'll be able to see the millivolt audio signal response with home-mechanic tools. And DON't hit it with a hammer. It's a microphone after all. You'll get the same response from it tapping the bolt with the tip of a screwdriver. And you still won't be able to see it on a DMM.
thanks. how sensitive are they, and do they get set off by a loud exhaust or an exhaust leak? im wondering if mine was showing a connection. what a great bonus if I've been driving around at 6 degrees retard,...........maybe that's why ive been able to run 91 octane for 10 years . however wouldnt Bill have seen something on the shark tuner when we were doing speed runs with his lap top looking for knocks?
thats a great idea. .......i dont know why i didnt do that.... i guess i still can. i have TONS of data on 60 to100mph runs in 3rd gear 3500 to 6000rpm. that wire that was attached fell apart off the knock sensor just on route to the table. ill see if i can get a voltage out of it by hitting with a hammer. back to the power issues, most chips pull timing out at high RPM, right dont know if it is 6 degrees, but some for more HP at speed. has anyone done a dyno test of knock sensor connected vs not connected?
Mark..thats the _fundamental_ action when a KS is not reporting healthy, -6d across the whole map. Its the DNA of the ECU to protect the motor. No dyno test needed.
You need both, we had a long talk on how to un**** that car with a Sharktuner last month, I have one. None of that stuck?
The sensors "hear" all the mechanical noise passed through the roof of the block. The signal is filtered for frequency and for timing relative to crank position so the offending cylinder can be identified. The sensors may be more or less sensitive at different signal frequencies, but it's really looking for "when" so it knows the cause a little better. Exhaust noise isn't mechanical at the valley so it's unlikely a leak would cause the sensor to set off the system and pull timing.
If you were doing speed runs with Bill's laptop looking for knocks in the SharkTuner software, remind me again why you didn't continue and actually tune the engine controllers then. If it connected then ther's no need for harness changes to get it working now.
Mark..thats the _fundamental_ action when a KS is not reporting healthy, -6d across the whole map. Its the DNA of the ECU to protect the motor. No dyno test needed.
You need both, we had a long talk on how to un**** that car with a Sharktuner last month, I have one. None of that stuck?
sure , it always sticks... but, you forgot that i was only able to see somethings with bills efforts, as we were left with issues with the harness. needs a new harness and connections to work. very disappointing.
but, why dont we need a dyno. 6 degrees at WOT and 5 to 6.6K rpm, might help HP, no? have you experimented with the knock sensor by pulling one?
Originally Posted by dr bob
Mark --
The sensors "hear" all the mechanical noise passed through the roof of the block. The signal is filtered for frequency and for timing relative to crank position so the offending cylinder can be identified. The sensors may be more or less sensitive at different signal frequencies, but it's really looking for "when" so it knows the cause a little better. Exhaust noise isn't mechanical at the valley so it's unlikely a leak would cause the sensor to set off the system and pull timing.
If you were doing speed runs with Bill's laptop looking for knocks in the SharkTuner software, remind me again why you didn't continue and actually tune the engine controllers then. If it connected then ther's no need for harness changes to get it working now.
You might have forgotten, but we ran into the harness issue that only allowed us to do minimal testing. no tuning was possible. we only checked for knocks and wide band fuel mixture reporting for WOT runs
...You might have forgotten, but we ran into the harness issue that only allowed us to do minimal testing. no tuning was possible. we only checked for knocks and wide band fuel mixture reporting for WOT runs
I guess we've all forgotten. Sharktuner either connects to the ECU boxes, or it doesn't. However, the car has two diagnostic buses and they both need to work. Maybe one of them is broken.
We're talking two wires, it is simple to check:
K-bus is pin-10 on the 12-pin diagnostic connector, pin-1 on the EZK connector, and pin-12 on the LH connector;
L-bus is pin-12 on the diagnostic connector, pin-7 on the EZK connector and pin-16 on the LH connector.
Two wires, three connection points each, check for continuity and short to ground. Less time than it took me to look this up and type it.
Then reconnect everything, fire it up, and check for 6-7 volts on pins 10 and 12 at the diagnostic connector.
I guess we've all forgotten. Sharktuner either connects to the ECU boxes, or it doesn't. However, the car has two diagnostic buses and they both need to work. Maybe one of them is broken.
We're talking two wires, it is simple to check:
K-bus is pin-10 on the 12-pin diagnostic connector, pin-1 on the EZK connector, and pin-12 on the LH connector;
L-bus is pin-12 on the diagnostic connector, pin-7 on the EZK connector and pin-16 on the LH connector.
Two wires, three connection points each, check for continuity and short to ground. Less time than it took me to look this up and type it.
Then reconnect everything, fire it up, and check for 6-7 volts on pins 10 and 12 at the diagnostic connector.
i dont remember the details, but one of the bus connectors was totally usless and was missing wires, the other had issues too. i think we were discussing jumping some of the missing wires, but that became too troublesome. so we took a pass . its been 6-7 years now
Originally Posted by Speedtoys
"have you experimented with the knock sensor by pulling one?"
No, because that is absolutely pointless.
This is the documented reaction of the ECU.
I dont care what Bill did, I offered to help last month.
I know it is documented, but i was more concerned or interested in the effects . again, there are many programs (general tuning for all performance cars) that pull timing at the higher RPM range, so i was wondering if the timing retard might actually help at the highest RPM power levels.... just curious. as far as help, we have to dig into the cables and see if we can find a connector to wire the loose wires into the proper form factor for the sharktuner, and hope wires are not missing as i remember that might be an issue. i know one of the harnesses was replaced by a 928intl replacement. but the other remained original (hacked or "modified" up by porsche or Porsche NA for the holbert car)
i dont remember the details, but one of the bus connectors was totally usless and was missing wires, the other had issues too. i think we were discussing jumping some of the missing wires, but that became too troublesome. so we took a pass . its been 6-7 years now
I know it is documented, but i was more concerned or interested in the effects . again, there are many programs (general tuning for all performance cars) that pull timing at the higher RPM range, so i was wondering if the timing retard might actually help at the highest RPM power levels.... just curious. as far as help, we have to dig into the cables and see if we can find a connector to wire the loose wires into the proper form factor for the sharktuner, and hope wires are not missing as i remember that might be an issue. i know one of the harnesses was replaced by a 928intl replacement. but the other remained original (hacked or "modified" up by porsche or Porsche NA for the holbert car)
'but i was more concerned or interested in the effects'
it pulls 6d of timing from the already slightly conservative timing map. It will not make more power up top, it feels like towing a 500lb trailer.
Anything you need to know not listed here?
You cant make a car idle & run right at speed at the same time, so lets stop believing there's something you could do better, using the proper tools, that you don't wish to use. ID like to see you get to the point of knowing what you dont know, and then making that car run right.
There is no theory is how it runs missing one or both knock sensors, or a hall sensor.
08-20-2018, 12:57 PM
Archaelogical dig finds knock sensor in worst shape of all time
