614Streets

Is the ESC system affected by aftermarket parts such as Headers, louder exhaust, and more engine compartment noise?
I tried to do a little research on the WEB as well as in a few books and on this forum but all I have really been able to dig up is How the system works.

I have the belife that my exhaust system and general engine compartment noise is causing the sensor to activate and retard the timing. I figure I'll have an assistant strike the block with an extension with my timing light hooked up but all that tells me is that it is working.

The reason I belive it is retarding the ignition false is because with the timing light hooked up twisting the throttle with my hand the timing mark moves down .

I read that loss of the ESC signal to the ECM will cause the ECM to constanly retard the timing and thus set a code 43.

However besides the statement I read that the ECM throws a check at the system to "Test" the "Normal" Operation , how could a single wire signal being unplugged from the sensor set off a code signal to the ECM that the signal was lost??

It is to my understanding that the sensor is basically a vibration ground(dont know of internal sensor parts read something about crystal) and when vibrated by the frequency a "knock" makes closes (opens?) the circut and thus the signal of voltage is read by the ESC unit and converted to a signal to the ECM Spark Advance area.

How likely is it that my sensor could easily read knock when infact no knock is occuring? The price of power loss through retard is to crital for me to overlook as I am trying to get the engine running in a optimal power output/ and correct manner.

P.S. It is also my assumption that even though I belive the ESC retards timing only when knock is present and returns to normal in a timely manner , If something other than knock triggers a signal , then it would be obvious that the signal would be nearly constant therefore meaning alot and often Ignition Retard. What gives?

614Streets

Also can the system be tested in this manner?
Unplug the blue wire to the knock sensor and ground it.
Which way does it work, by being grounded or ungrounded?

funstick

do a search on pizeo electric. its basically the same thing as a acoustic guitar saddle pickup. in short when its vibrated it puts out a very low voltage. the ecm see this voltage and know it now has spark knock depedning on the output FQ of the knock sensor. its very posiable if you have a niosy valvetrian that whats really going on is lifter clatter or even stiff valve springs are slamming the valves shut is tripping it. try wrapping a bit teflon tape on the sensor threads to mute it a bit. if that helps then try a tiny bit more. also make sure to give the motor a good solid tap to see if its working correctly. there are some companys that makes sensors that are a bit more muted. im not sure who but i always use teflon tape to mute it down bit.

614Streets

First off let me say THANK YOU for a very intellegent response. I now better understand how the sensor works. Crystal under pressure emits low voltage electricity.

I made an attempt to remove the sensor today in its unforgiving position and came up short. I plan on buying a stubby 22mm to get it out and to place teflon on the threads.

Can you or anyone else anwser this one question and I think I'll be set. If the sensor is unplugged will the ECM notice no emits of current and therefore piggyback to a preprogramed function/table to constantly retard timing for safety issues? Or did the programmers not go that far. I ask because a manual of mine says in short word , but isnt paticulary clear on the manner.

614Streets

In addition I want to run a light off of my aldl or ecm to a small security light on my dash so when the engine is sensing knock I can manually record it. My ESC system is suspect , I notice to much false reading under scan and manually with the strobe light. Does anybody have any direction/ Prepreperation before I have to spend a day reading at the library?
GM S10 1227747 p/n AND c3 family......

funstick

without a good scanner i doubt you could chart spark knocks. have you tried

www.winaldl.webhop.net

its a very good free scanning program designed to work with laptops or pc's runnign windows. the interafce is simple to build. i dont think there is any easy way to chart the spark knock any other way. also i dont think without the knock sensor in the engine the ecm will notice. the 1227747 uses and off board esc knock module that sends a 2.5 volt signal anytime there is knock. the ecm can only detect the failure of the module. and the module can detect the continuity failure of the actuall sensor and trip a code by not checking out . but physically removing it from the engine and leaving it plugged in should not net a error code.

1bad91Z

iTrader: (1)

Do yourself a favor and just leave the stock knock sensor alone. Then buy a new knock sensor and install it on the driver side of the block (there is a plug on the oposite side of the stock knock sensor and it plugs a water jacket). Then extend the wire to the new knock sensor, plug it in, refill coolant and you're done! Thats what I did and no more codes for ESC!! Just my $.02

Mike (1bad91Z)

614Streets

As far as scanners go I had it hooked up to a LCD hand held unit from Monroe. The spark knock check box lit up just off Idle and throughout the rpm band, almost constantly.

I wonder how much a new sensor will cost me and If the original unit is bad. I would think It would throw a code . Also why place it on the drivers side?

1bad91Z

iTrader: (1)

2 reasons why to put the new knock sensor on the driver's side of the block.

1. If you have long tube headers like myself, there is no room to plug the harness connector onto the sensor.

2. It's a pain in the butt to remove the knock sensor (I spun the inner casing of the sensor when trying to take it out, had to use a small pipe wrench on the outer casing to remove the old sensor).

It's just easier to get to and if you ever decide to use LT headers, the sensor has to be on the driver's side.

A new borg-warner knock sensor for TPI motors are ~$30.00 at your local auto parts store.

Hope that helped!

Mike (1bad91Z)

92 zzz28

I had to do the exact same thing when I put on the hooker LTs.
I did not need to extend the wire, you should be able to pull it out of the harness back by the distributor and then run it down to the knock sensor repositioned to the driver's side.

614Streets

Guy I bet your right , I think the sensor is bad. I sucessfully made my own WINALDL ALDL to Laptop harness tonight and I can now track the knock counts and Im positive it is bad.

614Streets

Notice the screen reads one knock count , at basic Idle 5 minutes later it had climed to 24.......

funstick

glad to hear that your ont eh right track. i also forgot one of the other goblins of knock sensors. double roller timing chains. they apperantly produce just the right amount of noise to set it off. although in adiable to us when the engine is running the knock sensor can in fact pick up onthe noise. and as ai said before if the knock sensor is not bad try a bit of teflon tape to mute it. good luck.

92 zzz28

Also those timing gear drive sets will make the knock sensor go crazy. Lots of noise and vibration.

borgx

iTrader: (4)

I'm having similar problems... I have a double roller timing chain too...

If they are just noisy, changing the sensor will not help. I wouldn't think the teflon tape would help that much either... some, but not that much...

I read the posts above and never came away with a direct yes/no answer... Can the knock sensor wire just be disconnected?

614Streets

I just ended up disconnecting mine , I get no trouble codes and the performance has jumped up.

borgx

iTrader: (4)

Here's a stupid question...

Where the heck is the knock sensor on an 87 GTA with a 305???

1bad91Z

iTrader: (1)

614Streets - what ?

How did you get away with that? What year make model car do you have? It is TPI right? If so, is it MAF, or Speed Density? I have a Speed Density '91 Z-28 and it will code everytime when I unplug the knock sensor! Just curious!

To Borgx -


It's on the lower part of the passenger side of the block, right above the topside of the oil pan where the dipstick tube plugs in. Just follow the dipstick down to where it goes into the pan and you'll see the knock sensor.

Mike (1bad91Z)

614Streets

Its a 747 TBI 1992 2.8 S10 With a supercharger 5 psi and nitrous100hit.

CHEVYTOWN

I hope ya'll don't mind moi jumping in on this one...

The L6(250cid) I have has a gear drive and is pretty darn noisy. The only thing is I dont have a knock sensor. I believe it is built into the harness/ECM because with WinALDL, I am getting knocking counts...a ton of them! It ranges from the hundreds to the thousands...What is the safest amount of knocks that an engine can produce...Cuz I ain't gettin single digit #'s nor double digit #'s....triples and the like...I think I might try running without the "can" on the dizzy to see if that helps any...

PS...What kind of supercharger are you spinning...Was it difficult to tune the ECM, Cuz I have the same one running a TBI as well, and would like to go blown in the future.

grumpygreaseape

im not certain what ecm we are discussing here, this is how the ecm will verify that the knock sensor and related circuitry and (if present) the knock module. this is the strategy that is used in all the ecm's used in out generation f-bodys-when you start the car and drive off the ecm will, for lack of a better phrase, overadvance the timing once the coolant is showing 194 it waits till a high load condition happens and does the test. no self test will be run if it detects knock prior to the 194 degree coolant temp. then it will assume the system is functional. i can tell you the test procedures for what system you have the problem with if you will say-one post above mentioned a knock module and 2.5 volts to the ecm which is actually half correct for each system-if you have a module the ecm gets a 12 volt signal from the module indicating ok. the 2.5 volt reference is for the s.d. cars where the knock sensor is directly wired to the ecm. if you have a module 165 style- take a test light with the clip on the battery positive and stick the k.s. wire momentarily and knock should show on the scanner or as retarded timing at the balancer. hope i was able to help make it all clear as mud. p.s. aint no way someone just pulls the wire off the sensor and dont flag a code, uhuh no way unless the ecm is ng or something along those lines.

grumpygreaseape

yet another cause of "false knock" is voltage induced into the signal wire from plug wires routed too close.