Honda / Acura K20a K24a Engine Forum banner

1261 - 1265 of 1265 Posts

·
Registered
Joined
·
489 Posts
Good question. The meaning of "knock count" here is KPro Manager SW counts the events, which are rated by ECU as an Knock Event. The parameter in KPro Manager is called "Knock Count".

Background:
The Honda Anti-Knock-System is very simple. It consists out of a "ear" the knock sensor, the wiring to the hardware listening to the knock sensor signal and the SW which compares the signal to a engine speed and load related map of maximum signal amplitude. If the SW see an specific for engine speed and load exceeding amplitude it interprets it as knock. So the only thing preventing this system from listening a false knock is the specification of the knock sensor. If this part has not the specified frequency sensitivity (resonance tuned), then wrong interpretation is most likely. So use only OEM sensors or the exact copy of it. Otherwise when the knock source get changed with turbo, header, higher VE, different timings and so on the signal amplitude get changed while the thresholds get not. In that case you get a mismatch of both and wrong interpretation is the consequence. Anyway, any change of HW or SW on the engine affects the reliability of that system.

The knock events we saw were independent from the thermal knock, which is a pressure oscillation in the chamber with supersonic speed and huge pressure amplitudes of up to 1000+ bar peak. It get caused by self ignition of the mixture which is not below the flame front but in front of it. In a rough simplification, think of two waves running to each other and heat each other with some massive reflection, super-positioning and a lot of fun to combust much more faster. If such a 100-1000+ bar peak pressure hits the wall and lose some of it's energy at the wall during reflection it makes a noise like you hear when you collapse these air chamber containing foils which are used for sheltering products during shipping. I mean the small (around 8-10 mm ID) circular air chamber foils. If you press them the foils get stretched until it breaks and make that little hard explosion noise. This sounds a bit like knocking.

With the above mentioned Background now you can imagine false knock is a signal of the knock sensor, which let it exceed the threshold, but is not caused by combustion, it is maybe caused by an oscillating plate, a loose timing chain, a hard cam lobe or what so ever. False knock can have any source which in amplitude and frequency looks similar to the knock amplitude and frequency. The frequency we speak about has 6.7-6.5 kHz with the 86-90 mm bore size. A rattling timing chain is a good candidate to have such frequency and high enough amplitude in its noise spectra.

I am not sure about the English term "spark knock". If it means thermal knock or knock caused within the combustion, then I would be ok with that. But it is not very likely to have a knock event at the spark plug, almost impossible. Knock happens always in the so called end gas, the mixture in beyond the flame front, never within the flame front. This does not mean spark plugs don't get affected by knock events, they do, the pressure waves running in all directions up and down destroy everything on there way and where the peak touches the wall it increases highly the heat transfer by factors. So melting, pitting is all included. But this has nothing to do with the knock source, only the consequences.

I saw many data about knock, pre-ignition, running knock, glowing knock and all that kind of in engine damages, papers and measurements during my career. These are the most dangerous and best investigated but less understood combustion events when it comes to IC engines. Some of them like thermal knock can be handled easily with some know how and with an good detection system. The issue is, no single detection system of any OEM is safe enough to interpret knock right after the source system was changed. There is no general rule, it is always an adapted listening with simple up to complex mathematical analyse systems which say yes or no to a signal. So, never expect an modified system with the stock knock detection system to be safe. You need always to hear yourself if knock signal is a yes or no and tune the knock interpreation yourself. My issue as tuner is, none of the customers pays for this as this takes a longer road down and need onsite work with special equipment, from simple DIY DET cans, which work very good if you know what to listen to, up to 10-100k detection systems, which measure the in-cylinder conditions (via pressure, light reflection, frequency spectra analysis and so on).

I worked 9 years for an OEM engine development on test bench from single cylinder up to full sized engine, field applications and in the pre-development process of the combustion process design. No single detection system makes an engine safe, because all of them are reacting systems, none of them are predictive. Honda tried to do so with there K.Control (KPro parameter) calculation, which uses ign. timings for longer period of times when detecting knock, its a sort of fuel sensing system. But it works very conservatively. Nothing good for racing or weekend worriers. It can calm the engine down so that you mean you drive a half size engine when maximum conditions are reached.

If you run an engine like Brosome does it with stock knock logic activated, you cut the power output by up to 25 %, reduce the mpg massively and get a weird and less cross sounding engine finally. Therefore we investigated the if those knock events will get a yes or no. Our strategy after the analysis is frequent observation of spark plugs, retarded ign. timing map, aggressiv IAT bounded ign. timing retard and some other stuff and with stock knock logic off. Unfortunately the driver at the dyno wasn't able to differentiate knock and noise with the DÈT can. As I said, you need to know what to listen too. Honda engines makes a lot of noise, it need will and experience to differentiate knock from noise. That guy wasn't willing to listen, he was more on his phone as on the DET can. Some noise I was able to hear via driver laptop and Skype, at least his many phone calls (not with me) :cool:. So no data from that side, therefore the conservative ign. timing tuning.
Very thorough answer thanks for clarifying 👍
 

·
Arouse the DAMPFHAMMER!
Joined
·
6,377 Posts
Very thorough answer thanks for clarifying
Thanks. I know it was a bit long answer, but that topic is exiting and complex and need some background details, which I hope helped to understand the essence of knock handling regarding Honda K-series with KPro. If not, I will try my best to answer upcoming questions.
 

·
Registered
Joined
·
24 Posts
Hi guys,

A bit of news for those who had followed the dynos of my car to say that I unfortunately blew up the K20 on a run ... 2>1 instead of 3, big over-revving and that was not forgiven!

I think there is not a lot of recoverable, to be confirmed when dismantling ...

I'm going on a K24a Jdm 200hp project (Stock head but with 50vtc and K20a oil pump or maybe even the intake cam if recoverable!) With the same bolt-on, except the RBC which will be "ported" by me and change the injectors of 550cc for 850cc? (E85)

More soon!
 

·
Arouse the DAMPFHAMMER!
Joined
·
6,377 Posts
A bit of news for those who had followed the dynos of my car to say that I unfortunately blew up the K20 on a run ... 2>1 instead of 3, big over-revving and that was not forgiven!
Sorry to read that man!

...a K24a Jdm 200hp project...injectors of 550cc for 850cc? (E85)
Nice project, I suggest you 850 ccm injectors.
 

·
Registered
Joined
·
180 Posts
I wish the shop that did that original tune gave me a dyno print out (which probably tells us something right there), but ya it just felt gutless down low, I was doing my everyday shifting at like 4500ish just to get down the road. I did read way back in the stock k24 block thread someone had the same issue with it just feeling very lacking down low on those cams. But from what I've been able to find on the 3.2's as far as other people's results with them and now my experience, they really do seem to come alive with more compression. I'd have gone a tad higher comp if I could source E85 near me, but 11:1 is what the shop I went to advised for pump 93 safety.
I know Newport News at Patrick Henry Mall BP has E85. 1st gas station after you get off 64 W & Jefferson Ave. I am going to ask my P-Town peeps where they have seen E85 at.
 
1261 - 1265 of 1265 Posts
Top