Detonation

All about our superior engine management unit.
Locked
PBU
Posts:134
Joined:Mon Oct 27, 2014 12:42 am
Detonation

Post by PBU » Thu Jan 01, 2015 4:07 pm

Hi
Has anybody a log/picture of a detonation ?
I have no idea what to look for or how high the spikes are.

Grts
Patrick

Kurt
Posts:43
Joined:Tue Aug 26, 2014 8:31 am
Location:Ft Worth Texas

Re: Detonation

Post by Kurt » Thu Jan 01, 2015 5:40 pm

20Hz logging is useless for trying to see knock events from the Knock Sensor Value channel. You can see knock events with 20Hz logging through the Knock Action channels after it's all been set up correctly.

PBU
Posts:134
Joined:Mon Oct 27, 2014 12:42 am

Re: Detonation

Post by PBU » Thu Jan 01, 2015 7:15 pm

Kurt wrote:20Hz logging is useless for trying to see knock events from the Knock Sensor Value channel. You can see knock events with 20Hz logging through the Knock Action channels after it's all been set up correctly.
You mean 20hz is not fast enough. You can only check the result => the reaction in ignition or fuelling ?

User avatar
Jadzwin
Site Admin
Posts:1445
Joined:Tue Aug 12, 2014 11:27 pm

Re: Detonation

Post by Jadzwin » Thu Jan 01, 2015 10:29 pm

Kurt you are not right.

The logger is quite smart and it show the maximum value of knock level , voltage and all cylinders on what there was detonation between each log frame.
So even the log works 5Hz you still will see any knocking and you will get information what cylinder was involved.

PBU
Posts:134
Joined:Mon Oct 27, 2014 12:42 am

Re: Detonation

Post by PBU » Fri Jan 02, 2015 1:14 am

Jadzwin wrote:Kurt you are not right.

The logger is quite smart and it show the maximum value of knock level , voltage and all cylinders on what there was detonation between each log frame.
So even the log works 5Hz you still will see any knocking and you will get information what cylinder was involved.
Jadzwin
You have an example ?

Kurt
Posts:43
Joined:Tue Aug 26, 2014 8:31 am
Location:Ft Worth Texas

Re: Detonation

Post by Kurt » Fri Jan 02, 2015 8:16 am

Jadzwin wrote:Kurt you are not right.

The logger is quite smart and it show the maximum value of knock level , voltage and all cylinders on what there was detonation between each log frame.
So even the log works 5Hz you still will see any knocking and you will get information what cylinder was involved.

Very good info! Glad to be wrong about it. :mrgreen:

I've noticed that Knock Ignition Event will show 1 or more cylinders for a given data point. So, if the log shows a cylinder in Knock Ignition Event, will the Knock Level value at that point correspond to that cylinder?

Can you please explain what we're seeing with Knock Sensor Value and Knock Level? I'm certain (from bench testing) that the Knock Sensor Value channel will NOT always show a spike when a knock impulse is induced and knock ignition retard is activated. I did not pay attention to Knock Level, though.

I use Knock Sensor Value data to set Engine Noise table and just ensure through sound recordings that the data is 100% certain knock free. Is this channel showing only peak levels, average, or simply current value at each sample point?

User avatar
Jadzwin
Site Admin
Posts:1445
Joined:Tue Aug 12, 2014 11:27 pm

Re: Detonation

Post by Jadzwin » Fri Jan 02, 2015 7:25 pm

Hi Kurt!

I will check it after weekend again.
However knock level and knock sensor value should be the peak value between samples.
If knocking on multiple cylinders occurs between 2 log frames, you know only the peak value of knock level on one of the marked ignition events.

Howver the goal is to keep knock level equal to 0.

PBU
Posts:134
Joined:Mon Oct 27, 2014 12:42 am

Re: Detonation

Post by PBU » Fri Jan 02, 2015 9:32 pm

So is this knock or engine noise ?
(2.59V is knock output.)
I think 2800 rpm, 24.5°ign and 84 map is not really knock condition.

I have just installed the new soft so i think i have to adjust something , not ?
You see (bad) on the bottom the ignition retard taking 10° action. So before spike advance was 34°.

Image

User avatar
Jadzwin
Site Admin
Posts:1445
Joined:Tue Aug 12, 2014 11:27 pm

Re: Detonation

Post by Jadzwin » Fri Jan 02, 2015 11:30 pm

Engine Noise 2D table is used to calibrate noise level of normal engine operation on whole RPM range. If Knock sensor value exceeds Engine noise from table, it is interpreted as engine knocking. This difference is named Knock Level. Higher Knock Level means more severe engine knock.


So first you need to build your engine noise table based on non knocking engine based on knock sensor value. Then if the knock occurs the knock level will be > 0.

In next week we plan to release the 1.1 official version. There will be help in English for every EMU option so all information about knock sensing will be available.

User avatar
clem
Posts:83
Joined:Tue Nov 18, 2014 2:46 pm

Re: Detonation

Post by clem » Mon Apr 13, 2015 4:15 pm

Question for those who know:
Is it possible that hard engine mounts (very hard, but not welded stiff!) influence the knock sensor signal? I think with my Knock signal something is wrong, as it reaches up to 4,2V @ 7000 RPM redline. Engine is SR20DET with some mods @ 0,3 bar boost and standard knock sensor. when rolling down a hill with deceleration cut, value is at approx 0,7V and when idling at 1,4V. When engine starts up, the value spikes up to almost 5V.
I somehow lost a part of the log (accidentially closed the log :oops: ) and have just a few smaller revs on the safed log attached down there.

I have not connected an aplifier to the sensor signal yet, to listen to the sensor via headphones, but I surely will do that soon.
Which Knock sensor values should approximately be read on this engine?

THANKS!!
Attachments
20150412_1725.emulog.txt
(466.04KiB)Downloaded 65 times
Fuel Injector cleaning and flow measuring service needed? -> contact me :)

WHPZach
Posts:184
Joined:Tue Oct 07, 2014 11:13 pm

Re: Detonation

Post by WHPZach » Tue Apr 14, 2015 1:55 am

I've heard of cases involving stock vehicles where something as simple as a poorly fitted exhaust that hits or rattles against the chassis under load can cause a false knock reading. It's entirely possible that a metallic rattle from something attached to the engine produces enough noise to be picked up by the knock sensor. It's also worth inspecting your wiring and grounds just as a precaution.

Locked