Discuss all things tuning in this section. News, products, problems and results.
Is contructing a database of engine knock characteristics by engine (ex. 4g63 or B18C5) something that this tuning community would see value in? How difficult would it be to obtain accurate data? What would be the limitations of this data as a reference?
I think that would be a great Idea. Its just a matter of ppl who have tuned various engines to chime in and share where the various engines tend to knock. That would be based on ign timing fuel mixtures etc.
I think the idea is great, however the application of it in reality is probably risky. The reason for this is that there are so many aspects that can effect the knock threshold. Fuel quality would be obvious, but even changes such as a less restrictive exhaust, or altered cam timing can make a difference. Then there is the issue that even with two identical stock engines, often there is a difference in what ignition timing the engines will take. I've noticed this over the years with the wide number of Mitsubishi 4G63's I've tuned. Some engines for no apparent reason will not accept the same timing without knock.
While the idea is great, quality audio knock detection is still your best option unfortunately. I'd hate to set up a database like this and have someone damage an engine because it wouldn't accept typical ignition advance.
I imagined that this idea had probably occured to other tuners in the past and that there was a good reason no such database was evident. In my mind, the concept was to give tuners a course refernce, but now I understand the variabilty is too great and the potential for someone to use the data as an exact model without following standard ignition tuning protocol could easily result in engine damage. Thank you for the insight!
I think it's a good idea but I agree there are reasons why it may not work. If it is used as a guide then fine but if it is used as a definitive knowledge database then it could cause problems. I think with things like this that a HUGE disclaimer at the beginning is always best saying that it is only a guide and not a definitve knowledge database is the only way to limit potential damage by people who rush into things without thinking about it.
The variables would just numbers too greatly to consider, as Andre has said there's a difference in knock levels between two similar engines. You can even have a difference in knock frequency in the same engine from just raising the boost.
If you were to try and compile a list and did have a disclaimer that it should be used as a guide no matter what you do there will always be someone who'll try and short cut their tuning, use the information and do damage to their engine. To add to that the type of society we live in now you could end up facing legal proceedings or being sued means you'd be laying a mine field for yourself
"If you were to try and compile a list and did have a disclaimer that it should be used as a guide no matter what you do there will always be someone who'll try and short cut their tuning, use the information and do damage to their engine. To add to that the type of society we live in now you could end up facing legal proceedings or being sued means you'd be laying a mine field for yourself"
I disagree with this assumption simply because the disclaimer is the legal "out" for usage of such a thing. It is common practise to put within disclaimers what jurisdiction the law binding the use of the document is. I serioulsy doubt someone in Siberia or Far East Russia would try to sue me for posting a brief guide on how to tune a pre-Gen3 Delco. Also if it is written in plain English (assuming of course we are talking Common Law jurisdictions) and says something like "You use this at your own risk and no responsibility for damages encured by you will be taken by the authors nor can responsibility be infered upon the authors for your usage of this guide." It gives the author a clear legal "out" for the sometimes rather obvious stupidity of others because they chose, afterall choice is an important concept in legal matters, to use the guide.
If everything on the Internet gave people a legal basis for litigation there would be nothing on it at all because we would all be scared of posting anything. It is the exact same for the HPA courses someone watches a course but kills their engine it isnt HPA's fault it broke but rather the person who was doing the tune or the engine itself was going to break anyway. Simple reality is you use it, you break something using it, you can't blame others for it.
Database of good bandpass settings for knock detection in different engines would be a good idea with additional comb filter frequencies that can be identified as known noise from certain parts of engine.
For piston engines, there is a way to calculate the probable knock frequency http://www.phormula.com/KnockCalculator.aspx
For rotary engines, I've read at least two different frequencies, 3.5kHz and 2.5Khz. I think this is based similarly on side of ther rotor face as is the knock calculation based on piston diameter.
I wouldn't really try to calculate the frequency on a rotary, if that engine knocks your apex seals are probably gone by then.
Anyways, I agree that there are too many variables to produce such a database. I assume that a lot of guys would just punch in the numbers and hope for the best without determining if it is correct or not.
Speaking of different fuels... I've had a 2JZ on the dyno last week and between our 100 octane gas and the german 102 octane, I was able to run around 5-8 degrees more timing on the same boost levels with that fuel.
So the database number for a 2JZ would be next to useless ;)
Following on form what jalava has posted, a database of engines and their corresponding knocking sounds would be handy, but probably not something that many people would be keen on contributing to..
KGPerformance - just out of curiosity, what kind of power increase did you see with the additional 5-8 degrees (all else - boost - remaining the same)
I'll look at the dyno plots tomorrow ;)
I should add, that Aral 102 octane fuel has a minimum of 10% ETBE content (maximum of 15% allowed and has a research octane rating of 118).
Nice. When using that value in knock control, should that value be used or the should the second harmonic be used? IE, a 4.00"( 101mm ) bore/piston has a base harmoniuc of 5.7KHz and a second harmonic of 11.4KHz.
G'day Keith. The amplitude of the knock signal will be higher at the base frequency, but so will the amplitude of the rest of the engine sounds. In engines with really noisy valvetrains the second harmonic can often offer a better signal to noise ratio, allowing you to amplify things more and get clearer audio / signal level of genuine knock events. Its not something you can usually know without trying, though.
A DB with raw audio files (pre-bandpass filter) containing detonation events would be very useful for training supervised algorithms.