Peter

Nom,

There is only one 'anti-Mycroft' argument here, the rest is just debate. I'd wouldn't like to think that my comments were taken as 'anti-Mycroft'.

Jon,

Yes, I agree with your sentiments totally.

Peter.

nom

Hurrah, we're back on intelligent debate
OK, yes, the 'school' thing was really 'normal school' as opposed to A-levels - I certainly covered this stuff (or the theory behind it anyway) at A or S level or whatnot. And Peter, I didn't actually mean you, there's a few others who are a little less thoughtful...

Anyhoo, I think the 'fooling' the sensor thing seems quite possible - the noise may still be there, but disguised. So, 'fluff fluff fluff fluff' sounds fine to the sensor, 'pumph pumph pumph pumph' means there's knock, but what the hell does 'boof boof boof boof' mean? Well, I'll ignore it then, says the sensor. (Please inset whatever noises you think the engine makes rather than mine). But, once the 'boof'ing gets loud enough (i.e. end of effective addition of 'octane booster') then the sensor wakes up to it & hey-presto, the knock is back.
Assuming for a moment that the idea that boosters don't actually boost but instead 'cover up', does this 'cover up' actually help anyway? Or at least partly...
Sorry for the being simple. It is after 4pm on a Monday Mr Brain is still somewhere on Sunday morning... Maybe someone more awake would like to slash my idea (intelligently) to pieces?

Mycroft

Syed Shah, no I can't say I do recall you from there, but as all I seemed to get was flamed by so many, if you were one of them then please don't follow Nitros example and start abusing me here! To answer your question, that is what I'm trying to discover, but my thoughts go to this, if I sell a full blown Octane booster that needs to be mixed at about 8%BV then it is going to be bulky, very bulky, if however you sell the old 'magic bullet' then it is small and convenient, now call me a cynic but if you can use something that lubricates and cools the valves and adds a little to the octane and masks (a less contentious term than fools, I agree) the knock, then is this a real power filip or is it a 'get around', I suspect the latter, my belief is that this is indeed what we have here.

Nitro, first yeah a test engine is the way to do some testing, it is however a bit 'old tech'. the question i have been trying to answer and getting for the most part valuable feedback from is whether an Octane Booster is worth its name.

Modern testing equipment is a boon to us, it is quick and cheap and this means I don't have to book and PAY for a test run, a handheld device that is as accurate as a room full of gear from 10years ago means that I can post info on here without costing me my Job or emptying my wallet.

I find it difficult to grasp all that you have posted but you seem to be having a problem getting your head around this.

If you wish to pay for a BS test, then PM me and I can arrange this with the Lab. guys. that would be a great service to the club.

#Knock = pre ignition# Wrong! For want of a better term Knock is a post-ignition, pre-ignition cannot be detected by the Knock sensor, please re-read previous post by me. The knock you here and the sensors detect happens AFTER the initial spark and is a result of end-gas detonation.

Peter you have been exemplary in your questioning and I like to have any reasoned 'case against' you have done this and it has merely made me double check my findings. That is exactly what I want.

Nom good question, my next 'thing' is are the apparent power gains just the advancing of the ignition, as the knock now at a lower less 'resounding' frequency we don't notice it and if this is the case just how far can we safely go without damaging our cars?

So as to not confuse people I will again state that if you are using any OB then please do not stop on account of anything posted here as yet, this is not a conclusion.

Mycroft

I have found them at last! one of our test units!

http://www.ridgenet.net/~hideseng/

Syed Shah

No, you were not 'flamed' by me, in fact quite the opposite at times.

Anyway, as regards to your answer, well I look forward to your continuing research especially as to why they bother doing this sort of thing in the first place, also, although I am not an expert on cars, I do know that a knock-sensor is merely a microphone to listen for 'knock', this could not be fooled by changing the frequency of the 'flame front'.

__________________
Supra MKIV x 2
R6 *rip*
Soarer TT

Mycroft

Thats cool then.

The microphone will indeed pick-up all the frequencies, but the processor only wants to hear 6400hz, if it takes notice of other signals then it is not gonna work, if it took 'all comers' you could in theory retard the ignition by tapping on the cam cover or slamming the door on tickover!

There is good money in these things, their reasons for making the stuff is to make money.

I have a very jaundiced view of this World so just call me a cynic.

jrpidgeon

It would be interesting to test your theory Mycroft. If I get time I 'might' connect my Signal Generator to the input DET Sensor signal pin on the ECU and disconnect the DET sensor itself (Microphone). If I where to put the Power FC into Knock monitor mode and set the Signal generator to around ~6400hz I should see a Knock level on the Apexi hand controller screen. Using the signal generator, I would be able to see how wide banded the ECU detection signal is and how sensitive, also were the actual focal frequency is (your saying ~6400hz).
How does the ECU calculate the knock?, is it done by the amount of amplitude/signal strength (loudness) of the knock, or by a varied tap - IE tap.tap.tap, the more taps (knock) it hears the higher the knock volume?. (Just incase I have to use a pulse genetator also).

Mycroft

The Toyota unit uses phase locking to detect the 6400, what you propose would (I think) work.

To say the least I would be very interested in the result.

Ihave found this;

http://www.streetrodstuff.com/Articles/September_2000/Engine_Basics_I.php

Blow Dog

I would be most grateful if some of the foul-mouthed here would respect those that have obivously something valuable to give back to the community.

Very interesting reading BTW.

Cheers,

Cem

__________________
Forum Webmaster -
http://www.gtr.co.uk/

Peter

An unintentional swerve I'm sure what with all the 'action' today but what are your views on this? Are you familiar with the knock sensor used by Nissan on the GT-R or are you assuming it's similar to those fitted by Toyota to the Soarer, etc., i.e. next to useless?

Mycroft

I think I'm right in saying that the Nippon denso unit in my car is the same product as in your own GTR, the 'design' maybe different but the functionality is the same.

Did you read the link above?

somberg

Signal from a det sensor looks something like this:

http://www.picotech.com/auto/waveforms/knock_sensor.html

Basically it is a piezo sensor that generates a voltage in relation to the shockwaves generated in the block (in the case of the RB26DETT) following high pressures in a combustion chamber. Pinking in the end causes excessive pressures that in turn generate higher-than-average voltage spikes on a det sensor.
Piezo sensors come in two flavours: tuned and non-tuned. I think the RB26DETT are tuned sensors.

I'd be very interested in any test results, I'll probably be doing it myself also because the Motec ECU I'll be using doesn't have a direct det sensing capability and therefore I need to design addditional hardware/software.

For those interested, I have quite a bit more information and articles which I have dug up on the Web last week. If you're interested in these links, send me an email (andre@acq.nl or andre@inducom.nl)

Andre.

__________________
Bananaman..

Mycroft

What i do find interesting in the Picotech thing was the frequency, which to my eyes looks like 16Khz, now (if my music lessons serve me) the fifth harmonic of 6400hz is 16Khz (2½times) as to why they would use a fainter signal (amplitude/frequency) to detect knock is odd!

I coouldn't find out which sensor they were using or if by upping the sampling rate by 2½ times they 'shifted' the frequency notation on the scope.

We use National Instruments software Osc. and I will see if I can replicate this tomorrow (today!) by upping the SR.

Peter I have had a quick glance back and although not praising the knock sensors in our cars my quote was;
#The normal knock sensor on Japanese cars is actually quite basic, cant work much over 2000 rpm and as you say can even be fooled by internal noises# Next to useless is a bit strong, they do work.

somberg

Have a look at

http://www-s.ti.com/sc/psheets/spra039/spra039.pdf

Will give you a better idea of what's going on.

__________________
Bananaman..