Jump to content

Query about Knock Detection

Recommended Posts

Good afternoon Link Forum,

I am Daniel,

I am working in a Nissan S14 with SR20DETT forged engine with NS15 plug in ECU.

I tuning this car two years ago, and it was working well all this time.

The car is working at 1.4bar (relative pressure, 2.4 bar absolute, between 4000 and 7000 rpm), 

I am using det cans and also a Bosch Wideband Knock Sensor wired to the Link ECU to control knock.


If I use a very retrased ignition timing like -2 or -4 degrees at 2.4 bars I ear a very smoth noise with the det cans and also I see a very low noise profile with the knock sensor but the car achieves very poor dyno graph. Less ignition timing Log attached.

If I setup the ignition timing for 2 degrees at 2.4 bar I hear a noise like knock on my det cans and also the profile of the knock sensor changes and I can see more noise than -2 degrees. The car was running two years with this advance and it is working well. I add attached log standard ignition where you can see the noise profile with 2 degrees. Due to doubts related if it is real knock or not I change the 98 octane gasoline for Mol 102 Hi5 race gasoline and I hear on my det cans morelesss the same noise and also on the dattalogging I see similar spikes on the knock profile. Due to that I conclude that It is not real knock. I send attached the log with Mol with the name Standar ignition Mol. Due to I know that this fuel is very resistant to knock, in my experience with other cars, I work to achieve MBT with this fuel. I see that the noise on my det cans go to more noise but the car works well on dyno and I achieve MBT. I send attached a log with the name MBT advance with Mol.

Due to this results I begin to work again with 98 octane gasoline. I start with 2 degrees at 2.4 bar like the car was working for 2 years and I try to achieve MBT and I finally achieved the same result that I achieved with Mol. The noise that I hear with my det cans were the same as with Mol it sounds like knock is ocurring, and also the profile that I see from the knock sensor were very similar. I send attached the Maximum advance 98 oct log.

I was testing the car in very hard conditions to see if the engine is realiable. I made 100 kilometers without any problem.

My doubt about that is, How can I distingish real knock from noise on an engine?. Could the engine resist knock a lot of time without broke? The Mol gasoline is very knock resistant and I hear the same noise, this surprised me a lot. If I reduce a lot the advance the noise of the engine is very clean but the horsepower very low.

Less ignition timing.llgx Log Standard Ignition.rar Standard ignition Mol.rar Maximum Ignition Mol.rar Maximum advance 98 oct.rar

Link to comment
Share on other sites

Can you attach a copy of the tune.  What is the engine spec? forged or cast pistons?  What is the piston to bore clearance?  Hydraulic valve lash adjustment or mechanical?   camshafts?  Is the base timing accurate and is ignition timing stable and on target at high RPM?

Really need a PC log so that relevant data like RPM ROC and dwell are included. 

Link to comment
Share on other sites

Good morning Adamw, 

Thank you very much for your response,

The engine is a SR20DET from a 1995 Nissan 200SX.

- CP Carillo Forged Pistons

- Eagle Rods

- 8.5 : 1 Compression Ratio. (Stock Compression Ratio)

- OEM camshafts and cylinder head.

- OEM hydraulic valve lash (OEM hydraulic valve train).

- Garret 2871 R Turbo

- 555cc Nismo Injectors

- I don't have data about the piston clearance due to I don't assembly the engine but I can try to achieve the value.

 Is the base timing accurate and is ignition timing stable and on target at high RPM?

On the data base timming is accurate and ignition timing stable except when it detects knock in one cylinder and it reduces the ignition. On low rpm with the timing light it is also stable. Do I need to make some different test?

I add attached a PC log to check data. This is with the MBT ignition advance less 1 degree.


I add also attached the calibration file. Password: 123.

23-05-2024 19:40 I would like to add more PC Logs but the forum shows me that the files are too big.


200 sx test road PC Log.llgx 240523 Nissan 200SX SR20DET.pclx

Link to comment
Share on other sites

  • 3 weeks later...

Thank you very much for your contribution servicesoon. I never make this test. I will make it on the future.

I always have the doubts about how much time can an engine resist knock without broke and what test can I do to check signs of knock before the engine broke to solve it.

I suppose that the ideal timing for your test is tuning the car on dyno, go for a ride for example 20km or less and check the state of the spark plugs following the process that you say.

Link to comment
Share on other sites

on the link, a super clean run with no timing, vs a clean run with more timing, the noise level that you will see is different, that doesn't mean that it is knocking, usually knock you will see it as a big jump on the log. 

A engine will take some knock and is not going to break, depends on many factors how much it will take before showing signs of damage but I would not be too afraid of looking for knock on your engine. The only way to know that is not knocking is to make it knock, so you know how it sound on the knock ears and how the ecu will show it. Usually you try this at low rpms or/and low engine loads.

This is a timing table of a SR20DET from a S14 with VTC, not sure if it has cams, but if it does, they are small cause the VTC is working. Fully stock engine (like I said maybe cams, and maybe headgasket and head studs) tune on 98ron with a GTX2867 with the "big" garret vband rear housing. I did tune each cylinder, so some cylinders will have this timing some will run -1 or -2 degrees, the same with the fuel tables. 240kpa all the way. .82 lambda measure at the back of the exhaust so no pressure for the o2 sensor. Please do not try to run this ignition map but, is to show you how much could be archive without knock, so I would say that at 5degrees, is definitely not knocking if everything is OK on the engine.

ign table.jpg

Link to comment
Share on other sites

Thank you very much Rossobian, you help me a lot.

I chekced the noise in 2500 rpm and 120kpa and I can see that with more that 16 degrees it beggins pinging and if I set 25 degrees the car pinging a lot. In this part I can see the difference between knock or not I also feel on this part that the knock detection don't show spikes despite I hear the noise of pinging. But all the time that I exceed 5700 - 6000 rpm I hear this sound like pinging with more than 0 degrees of advance at 2.4 bar, the same with different spark plugs or different gasoline. 

On the road with this advance I can see continuos spikes on the knock signal in cylinder 3 and 2 but with the det cans I see a similar noise, I suppose that in this case more noise in the cylinders 3 and 2 are related with the position of the knock sensor. I changed the gain of this cylinders to 0.75 on cylinder 3 and 0.8 in cylinder 2 but the behaviour was similar. My lambda values are the same (the lambda is not at the end of the exhaust but it is after the turbo on the downpipe to avoid as possible backpressure effects.

Thank you very much for your ignition map.

Kind regards.

Link to comment
Share on other sites

Just a few examples of how the knock looks like, it could vary depends on the setup but, enough to show you. The one that is just one run, is on the dyno and you can see a little bit of knock, that's what happen when you push and push until you reach that threshold. In this case was a EJ25 which is famous for have very weak pistons, is making just over 400whp on pump gas, just a little bit of knock... no problem, the car is still alive, it did not left the shop knocking, but while I'm tuning I will reach knock, the only time I really do not want them to knock is when is a really high power engine, in that case I go to the point when it start to get nosier but no knock yet. The other picture is 1 pull to gears on the street, clean no knock, the only time it get noisy is when is on the limiter. I hope this can help you a little bit. 

knock example no knock.jpg

knock example.jpg

Link to comment
Share on other sites

Thank you very very much Rossobianconero, you help me a lot

I see that with a good calibration of the "normal noise" you need to see a very high spike on the knock readings.

In my case the noise that I hear on the det cans and also this spikes that I see on the cylinder 3 and 2 makes me doubts I cannot add the image attached due to I don't have enought space but it is related with the file that I add previously with the name maximum advance 98 oct. 

With you information I will spect more noise to consider it knock and I will push it more.

Thank you very much again.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...