Reading knock on SAFC II? (Post your values.)

[MK3Brent]

Tomorrow I'm going to re-calibrate my AFC2 for knk reading.


What are your maximum values showing on your AFC?
Did you notice a reduction when A)increasing fuel octane or B)decreasing boost pressure?

I'll post my numbers tomorrow, but wanted to create this thread before I forgot.

 

[MK3Brent]

16psi GT4088 6:30PM Max KNK value = 186
16psi GT4088 1:15AM Max KNK value = 177

The second pull at 1:15AM registered at 1964 hz in airflow, I did not remember to record the first pull .


So far, I'd say these numbers make sense.
With the decrease in temperature, the less my knock count was. (ruling out 'they're just BS numbers.')

What does 186-177 mean? I don't know, but I'm going to keep recording data.

Anyone else?

 

[TobyCat]

I've wondered the same thing, and asked this a while back.

I have also noticed that evening driving vs. mid day driving, the knock values are quite different. On a hard pull you can definitely watch the ECU adjust timing as the knock values get high then fall back.

I'm going to say it records it as 'knocks per minute' ?

 

[MK3Brent]

I think the AFC takes in any audible ping or knock spike (voltage?) from the sensor and displays it as raw data. The stock toyota computer sees this same data of course, and makes timing changes accordingly. I believe the real question is: "What does the stock computer determine as knock, and at what knock level will it retard my ignition?"

Also,

Does anyone know for sure that our stock ECU advances ignition UNTIL it sees knock, and then retards? (or is this some rumor.)

 

[Poodles]

more than likely does after an ECU reset, but I would think it would do it to compensate for bad gas or different octane...

 

[Shytheed Dumas]

I think there's one good way to make sense out of these values, which would be with the use of a control chart. If you're not familiar with these, it uses the natural variation in data points to calculate statistical 'control limits', and any value falling outside of the upper limit would be due to a 'special cause', which would most likely be legitimate knock.

I left my KS unwired to my safc, but if you can collect some data points we can take a look at the data. It wouldn't be a perfect data set, but if you could collect the max KNK value from 30 or 32 driving sessions, then we can look at some numbers. It would be good to keep a log with things like approximate ambient temps, octane, day/night, etc, because a lot more can be done in these cases to explain the things that contribute to higher numbers.

 

[NashMan]

the only way i can really think you can find out were it nocks and re tards is on a dyno with diggy timing light and have maft pro installed so you can see your timing target your timing and weight tell it smack's of zero this is really rough way of doing but you will get an idea

but really the stock ecu timing map is all way's changing

and in the end you really wana find out what a bad nock is

 

[Idealsupra]

honestly....there is no point...when apexi themselves cant tell you what the values on THEIR device mean... theres no point in even worrying about it. it will pick up ANY audible noise INCLUDING electrical static.

i dont even hook up the knock function when i install afc2s anymore...its just utterly useless.

 

[MK3Brent]

So getting "0" when cruising and then numbers at full load is utterly useless?

 

[Ric]

my knock wire isnt hooked up and it picks up knocks lol

 

[Idealsupra]

So getting "0" when cruising and then numbers at full load is utterly useless?

yeah pretty much lol. if there was a scale for the values then i would think otherwise but apexi doesnt even know what they mean. imo its just something to make you become paranoid about your engine

 

[Ric]

it's a safety feature. By unsheilding your knock wire to attach it to the SAFC2 knock wire, you just successfully retarded your ignition

 

[Idealsupra]

it's a safety feature. By unsheilding your knock wire to attach it to the SAFC2 knock wire, you just successfully retarded your ignition

:biglaugh:

 

[NashMan]

ummmm at the ecu it's not shield far as i remmber cause there is no 12 volts or = to distrbing voltage wire what so ever nere the pin's


and yes there is a point to the nock monitor if you change your base timing on the same run of roads the nock will change it's habbit of cycling numbers out ti back teh way ti was back to old same habbit

now the next day it will chnage again since the air is diffent and a diffent temp out side ect

so there nock 101 on the sfac


so in a way if the number is really flipping hi some thing is freaking wroung


like in my car i have never passed 150 ish ever far as i remmber

 

[Shytheed Dumas]

^^^^ which is why understanding the natural variation of knock numbers over time is the only way to make sense of the numbers. The only way to understand them is to collect data in a meaningful manner.

I'm tempted to wire mine up and start logging the numbers myself after reading through these threads...

 

[MK3Brent]

By unsheilding your knock wire to attach it to the SAFC2 knock wire, you just successfully retarded your ignition

Sincerely doubt it.

 

[MK3Brent]

It really grinds my gears when people are so dismissive on something they haven't even tried... it's always hearsay.

If these numbers are completely random, then I guess the ECU is getting nothing but "random" numbers too. This wire is sending the same possibly variant numbers and making sense out of them. I'd like to as well.

So again post your values if you have them to share.

 

[Idealsupra]

It really grinds my gears when people are so dismissive on something they haven't even tried... it's always hearsay.

If these numbers are completely random, then I guess the ECU is getting nothing but "random" numbers too. This wire is sending the same possibly variant numbers and making sense out of them. I'd like to as well.

So again post your values if you have them to share.

ask yourself this... is the wire going from the afc to the KS wire shielded?

 

[TobyCat]

The values are definitely *not* random. They do increase with RPM and with ambient temperature. If it's cold in the morning I get around 60 and it never breaks 100 on my way to school. On the way back, and around 25 degrees Celsius, it starts at about 90 ish and quickly sits around the 150-200 mark.

You can also watch your ECU retarding timing based on the numbers. If you do a hard pull you'll see the numbers climb very high (upwards of 700) then drop drastically to ~140 again and repeat this till you run out of rpm.

So, what the units are is a mystery sure, but weather or not it's BS seems very certain that it isn't.

From what I *guess* is it's a running average of knock sensor spikes over a minute. I think this would best be resolved by contacting an engineer @ apex.

 

[MK3Brent]

Thank-you for posting real world data. (the whole goal of this thread.)

What's really neat to watch is the knk numbers spike at full boost, then drop once the ECU pulls timing.

If you do the digital view of 2-channels (Graphic display
【Function】 Pause and replay)

You are able to record up to 30 seconds. I find this the best way to analyze knk vs. rpm.

The memory time is
as follows
【1Channel】...... 60sec
【2Channel】...... 30sec
【3Channel】...... 20sec
【4Channel】...... 15sec