ClioSport.net

Register a free account today to become a member!
Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

  • When you purchase through links on our site, we may earn an affiliate commission. Read more here.

F4R knock sensor





As the title says, I need some info on the knock sensor fitted to the F4R engine in terms of:

what rpm range or throttle opening does it operate under: 0 to max revs and WOT or does it stop before WOT

does the sensor feed an increasing voltage into the ECU as it senses more vibration and the ECU decides to back off ignition, or does the sensor have a preset "value" for knock that causes it to send a pulse to the ECU?

Any details on the mode of operation and ECU/sensor interaction would be much appreciated.

Many thanks,

Philip
 
  172 cup,s2 rs turbo


a knock sensor is a peizro crystal type sensor.it is a vibration type peizro sensor so unlike other peizro crystal sensor(map sensor)it doesnt req a voltage to work.the sensor contains a ceramic element which is clamped between the sensor base and the sensor securing nut(internal cannot be seen).the vibration from the engine(specifically engine pinking) causes the crystal to vibrate and generates a voltage between the top and the bottem of the ceramic crystal.the voltage is amplified and sent to the ecu where the voltage signal is converted into a meassurement signal(this signal is displayed on certain screens that you access with the renault clip etc).the ecu adjusts the timing depending on the signal sent from the sensor.special care must be taken when fitting these sensors as if the sensor is overtightened or to loose it will give false information to the ecu.
 


thanks Jay do you have more details on how the ecu adjusts the timing depending on the signal it gets from the knock sensor, and how to find out if my knock sensor is working ok?
 
  172 cup,s2 rs turbo


the ecu adjusts ignition timing and injection duration.the only way i know of checking a knock sensors output is by looking at the readings from the sensor through the renault clip.why dont you think it is working?
 
  172 cup,s2 rs turbo


that would show its working but not show if the readings are correct or not.the clip ossiliscope has a programmed curve tracer which gives you examples of a tracer curves exspected from each sensor when operating correctly.will have a look.got my scanner working now so ill see if i can figure out how to post pics and post one of the exspected curve.
 


heres the question.

as you know from previous threads, my engine was initially damaged by det caused by a bad unichip mapping that caused too much advance = running too lean = det = changed pistons etc etc.

Now that I am fully ok after mucho ££, I understandably want to be 100% certain we are not/never/ever getting det. The engine has been fine tuned, the mapping is spot-on, the lambda readings are aok etc, BUT I want more diagnostics.

So I started sniffing around Renault technical manuals for the F4R and surprise-surprise, the engine is meant to have a knock sensor! So why didnt the knock sensor and ECU do their job in the first place and r****d timing sufficiently to avoid the det that caused the original damage?

My theories are:

1- the knock sensor was working, but the ecu is programmed to r****d ignition timing by only so much (say 4 degrees) whilst my timing may have been off by 10-12 degrees (in fact we know it was). So the knock sensor/ECU was overwhelmed..

2- my knock sensor didnt work (so I would need to find out)

3- The unichip piggy back somehow overrides the knock sensor interaction with the ECU, but I really cant see that happen.

So heres what I need:

a) when does the sensor decide its detected knock, and what does the ecu do with the signal it gets from the sensor. Does it have just one setting for timing retardation, checks to make sure the knock is gone and they gradually ratchets timing forward until it detects knock again etc. I need to know maximum timing backoff values, and how it ratchets up

b) can I tap into the ECU knock detection circuit and get myself a diagnostic light in the car to warn me that were getting into knock etc.

Any help would be appreciated.
 


there is a misconception here. knock sensors are there just as a safety feature, but they never prevented engines from getting knackered by severe detonation.
moreover, it takes a lot to make a NA engine to ping enough to get damaged, and this tells a lot about the quality of some aftermarket chips.
also, pinging can occur in just one cylinder, due to uneven mixture distribution, faulty injector, faulty spark plugs, faulty valves etc.
actually, when knock sensors detect knocking, it is often too late.
if knock sensors were so good, ECUs would simply r****d until detonation started to occur then back up a little. unfortunatly is not so, therefore the maps must be conbservative enough to allow for some crappy fuel or other thing to happen without destroying the engine. many aftermarket chips are notorius for destroying engines, expecially turbocharged ones.
they were probably written from a comfortable desk making a lot of assumtpions and little tests.
as i said, knock sensors are far from precise, and can get very confused by engine noise and vibration that occur at high rpm, where knocking can be really destructive. in fact, most of the tuning firms use a mike and headphones plus a skilled operator to detect pinging when setting a car on a rolling road.

another thing to keep in mind is that pinging can occur at different levels of severity, and that there is something else called pre-ignition, which can really destroy any engine in seconds and which has completely different causes.

so when engine damage of that type occurs, an investigation should be done.
broken engines is just one of the consequences of homemade tuning...

Quote: Originally posted by philip on 02 March 2004
heres the question.as you know from previous threads, my engine was initially damaged by det caused by a bad unichip mapping that caused too much advance = running too lean = det = changed pistons etc etc.Now that I am fully ok after mucho ££, I understandably want to be 100% certain we are not/never/ever getting det. The engine has been fine tuned, the mapping is spot-on, the lambda readings are aok etc, BUT I want more diagnostics.So I started sniffing around Renault technical manuals for the F4R and surprise-surprise, the engine is meant to have a knock sensor! So why didnt the knock sensor and ECU do their job in the first place and r****d timing sufficiently to avoid the det that caused the original damage?My theories are:1- the knock sensor was working, but the ecu is programmed to r****d ignition timing by only so much (say 4 degrees) whilst my timing may have been off by 10-12 degrees (in fact we know it was). So the knock sensor/ECU was overwhelmed..2- my knock sensor didnt work (so I would need to find out)3- The unichip piggy back somehow overrides the knock sensor interaction with the ECU, but I really cant see that happen.So heres what I need:a) when does the sensor decide its detected knock, and what does the ecu do with the signal it gets from the sensor.  Does it have just one setting for timing retardation, checks to make sure the knock is gone and they gradually ratchets timing forward until it detects knock again etc. I need to know maximum timing backoff values, and how it ratchets upb) can I tap into the ECU knock detection circuit and get myself a diagnostic light in the car to warn me that were getting into knock etc.Any help would be appreciated.
 


thanks for the input (homemade tuning ... not quite). I still need details of the sensor/ecu interaction Ben/Jay if you have the info.
 
  172 cup,s2 rs turbo


your first assumption was probably right.the ecu carried out the max amount of adjustments it can but it wasnt enough to fully eliminate the pinking.the chip probably created a large amount of knock that the standard ecu/knock detection wasnt designed to cope with.the sensor is more there to help fine tune the engine running rather than stop an excessive amount of knock caused by a foreign factor.the knock sensor constantly sends a voltage signal to the ecu but it is the size of the signal that determines when the ecu recognises knock.the knock sensor is precise enough to detect the difference between normal engine vibrations and actual knock(pinking).i cant see a way that you could wire up a warning lamp to be honest.
 


Many thanks Jay. Ive also come to the same conclusion. I may not be able to catch the output signal from the sensor, but I may be able to catch the reaction of the ecu, i.e get a signal when the ecu decides to back off the advance. work in progress..
 


Quote: Originally posted by crono33 on 02 March 2004

there is a misconception here. knock sensors are there just as a safety feature, but they never prevented engines from getting knackered by severe detonation.
moreover, it takes a lot to make a NA engine to ping enough to get damaged, and this tells a lot about the quality of some aftermarket chips.
also, pinging can occur in just one cylinder, due to uneven mixture distribution, faulty injector, faulty spark plugs, faulty valves etc.
actually, when knock sensors detect knocking, it is often too late.
if knock sensors were so good, ECUs would simply r****d until detonation started to occur then back up a little. unfortunatly is not so, therefore the maps must be conbservative enough to allow for some crappy fuel or other thing to happen without destroying the engine. many aftermarket chips are notorius for destroying engines, expecially turbocharged ones.
they were probably written from a comfortable desk making a lot of assumtpions and little tests.
as i said, knock sensors are far from precise, and can get very confused by engine noise and vibration that occur at high rpm, where knocking can be really destructive. in fact, most of the tuning firms use a mike and headphones plus a skilled operator to detect pinging when setting a car on a rolling road.

another thing to keep in mind is that pinging can occur at different levels of severity, and that there is something else called pre-ignition, which can really destroy any engine in seconds and which has completely different causes.

so when engine damage of that type occurs, an investigation should be done.
broken engines is just one of the consequences of homemade tuning...
Some valid points, and generally along the correct lines, but a tad muddled up.

There are typically 4 not good ignition characteristics.

In increasing severity:


Pinking (not pinging)......a slight metallic noise you get. Usualy caused by ignition too far advanced and the piston rattling in the bore as it goes over TDC. Nothing major and the engine can live with i for longer than you think. Easiest solution is a higher octane or r****d the igntion.
Knock, a dull thud caused by 2 flame fronts colliding from accidental ignition of mixture away from the spark plug. Harder to diagnose as it can be hot sports, mixture variations, carbon depostis, lean spots, ignition timing. But best run a fuel that can resist knock, and RON is what we use to provide info on the fuels ability to hold back knock.
Pre-ignition, the auto ignition of the mixture before the spark is even present, so ignition timing isnt always, or rarely is teh cause. Almost always caused by a hot spot, either a plug that doesnt shed heat fast enough, a hot exhaust valve, carbon deposits that glow. But the heat is mostly caused by incorrect timing, so again, it can be solved by sorting ignition timing (conused? lol). But sorting mixture, timing and plugs usually sorts it.
Detonation, very similar to pre ignition, but wholitically more violent and instantaeous. Akin to an explosion rather than an abnormal burn. Difference is that it typically occurs after the spark has arrived


The sensors in engines are specific, each engine will give out a different frequency and the sensor is built to read that frequency. The knock sensor in the F4R will take action before any damaging problems start.

Possibly your piggy back ECU would have affected it since it affects input signals to the ECU.

Most RR operators who do look for knock use an automotive steathoscope since the engines they work on dont have a knock sensor, and you kneed to know the knock/det frequecy before yuo can fit your own knock sensor....so its easier just to listen.

There is alot of talk about making spark plugs aware to combustion abnomalities by monitiring the ionising voltages and resistances. By doing so, they can monitor each cylinder and each burn.
 


Thanks Ben. The info is very useful to try and work out what caused my original problem and make sure it doesnt occur again. after the repairs, we spent 4 hours fully mapping the unichip at all throttle openings. we were monitoring for any knock etc, so hopefully that will prevent any of this in the future. I only discovered after the mapping that the engine did in fact have a knock sensor which got me wondering about why it hadnt stopped the problem, but I think we have a plausible explanation for that (as discussed above). When the new pistons were put in, we looked for any hot spots that may be causing pre-ignition. There were one or two tiny imperfections that were smoothed out and we also changed the plugs to much harder NGK plugs used on the cup cars. The pistons were sent back to Renault for examination but we havent heard from them yet. Apparently the cup cars often have the same problems with piston rings cracking, so I am not too sure if this is not a design fault on the F4R having pistons too fragile for the job. For any of us running more powerful cars, maybe were tripping the pistons over their design tolerances. I remember now there were no pitting signs as such on the surface of the pistons, just the rings that were cracked on all 4 and two of them looked like a polo mint if you walked on it. Are there better quality pistons available out there for the F4R, just in case?
 


Top