Variable Ignition Timing Based on Knock Sensors

Status
Not open for further replies.
Joined
Jul 8, 2012
Messages
5,176
Location
Dickson, TN.
So, my question is really, how long is the cycle.

Let's say it's late July in Tennessee, almost 100° outside. I'm driving my Tacoma down a highway at 55 mph in high gear. The road grade starts to go uphill and I tip into the throttle a little bit, just to maintain my speed, but not enough for the automatic to downshift. I hear some light pinging, which is pretty normal. I've spiked cylinder pressures by cracking the throttle open even more at a not-very-high RPM.

SO, let's say I had been burning 93-octane fuel, but, on my last stop, I filled up with 87-octane pee water. Presumably, the knock sensors would detect any knock (how sensitive they are, I'd also like to know - will the sensors pick up pinging, or, would it have to be an actual knock event?) and pull the timing back a little bit.

OK, let's say, 300 miles down the road, I fill up with 93 again. How long does it take my ECU to re-advance my ignition timing and give me back my few extra HP & LB/FT? If I floor it and redline my engine through the first 3 gears as I'm merging back onto the highway, does the ECU then detect the fact that there was no knock, and advance timing again? Or, is there a timeframe that it has to get through with no knock, in order for the timing to be re-advanced?
 
There is a whole set of "War and Peace" like novels to be written on the subject ................... so imo its too manufacturer specific, model specific, and engineering approach specific to make a "generalized" statement.
 
Last edited:
In the case of GM vehicles the PCM's fuel trim and timing "maps" are always the same. This means the vehicle will always attempt max timing advance/fuel trim every time - the knock sensors pull timing back but next full throttle or other load event will have the car again attempting full advance.....the knock sensors control how much advance is allowed. This is the short version of a long and complicated strategy. Hopefully this will explain it a little.
 
Originally Posted By: LvR
There is a whole set of "War and Peace" like novels to be written on the subject ................... so imo its too manufacturer specific, model specific, and engineering approach specific to make a "generalized" statement.


Exactly, wildly platform specific.

Our sig car's knock sensors are so sensitive that you never hear knock even when premium is not used. Our newer fleet trucks are GM and I've never heard any audible knocking there either.
 
I have noticed in TN that our 87 pings way more than in other states
I would say what gas we buy is not the same state to state .

After multiple road trips from TN to Florida and other places I ping the most
in my home state,no matter what brand . I think we are not getting what we expect
at the pumps
 
Knock sensors are notorious for being ineffective, I wouldn't rely on one to keep detonation at bay. Think about the job they are asked to do ... decide which engine noise is knock, and which is some other normal engine noise/vibration. Nearly impossible. I would keep a can of Octane Boost in the truck instead, using it when you are stuck with a grade lower than you need. Much safer procedure.
 
Originally Posted By: Johnny2Bad
Knock sensors are notorious for being ineffective, I wouldn't rely on one to keep detonation at bay. Think about the job they are asked to do ... decide which engine noise is knock, and which is some other normal engine noise/vibration. Nearly impossible. I would keep a can of Octane Boost in the truck instead, using it when you are stuck with a grade lower than you need. Much safer procedure.


Que? What tripe. Knock sensors have been installed for decades because they work. Do you really thing manufacturers would take the time to design and tune them on a system if they were ineffective?

If you actually understood how they worked, and how the algorithms in the ECUs very effectively mask out the mechanical noise you refer to, you might have a better understanding of just how effective they really are. Once you've put a scope/logger on a vehicle and seen the ignition retard individually on single cylinders to cope with anomalies you'd be blown away at how "effective" they are.

Don't let the facts get in the way of a good conspiracy theory though. Much better to add some "miracle in a can" to the tank. Oddly enough, if your miracle in a can is any good (and again you were actually watching the ignition) you'd see the ECU slowly advance the ignition to actually take advantage of the extra octane your elixir bestows upon your chariot.
 
Originally Posted By: john_pifer
Presumably, the knock sensors would detect any knock (how sensitive they are, I'd also like to know - will the sensors pick up pinging, or, would it have to be an actual knock event?)...


Doubt that's an "officially" defined distinction, which it would need to be before the question could be answered by anyone.
 
Originally Posted By: Johnny2Bad
Knock sensors are notorious for being ineffective, I wouldn't rely on one to keep detonation at bay. Think about the job they are asked to do ... decide which engine noise is knock, and which is some other normal engine noise/vibration. Nearly impossible.............................


Well - read my first post again and then realize this is in fact the only instrument/sensor found on most current engines that allow the ECU to actually cope with the garbage fuel found in some places ............................... without it EVERYBODY would ALWAYS have to drive on a pre determined fixed (read expensive) RON/MON fuel in order to allow ALL ECUS to manage combustion effectively.
 
The Mitsubishi MUT III factory scanner shows knock learn - how much it is retarding from the 100% standard setting. Something we always look at with a GDI, normal engine shows nothing, as they are made to run on our fuel. My Volvo should run on a premium fuel, but I run it on the low octane, on a trip I put in premium, and fuel economy comes up a few kms later. I guess it remaps, but haven't bothered to watch ign advance on the scanner...a bit dangerous.
 
I agree with the above about each manufacturer being different. It all depends on ecu programming along with the sensitivity of the knock sensor. Sme more so than others. The ultimate goal is the ecu adjusts before there is knock. I went from a Ford modular 2valve to a Infiniti VQ engine. I swear I can't make that car knock. Even with 87 octane at low rpm high load. So the ecu is pulling time before there is a chance at a audible knock. You can get knock and not notice it in sound. So in Infiniti's scenario the car has timing being pulled before there is an audible knock. In my opinion the knock sensor is doing it's job and the ecu was programmed well. My Mustang would knock like [censored] with less than 93 and not adjust to it. I think it's best to not let knock happen. Use an octane booster that works if you need to
 
Saturn S-series was quick to pull and slow to give back.

Mine would drink a slug of oil after coasting down hill in gear, knock, and get a "flat spot" in acceleration that took ~200 miles to work out.

That car took very well to water decarbonization.
 
Originally Posted By: eljefino
Saturn S-series was quick to pull and slow to give back.

Mine would drink a slug of oil after coasting down hill in gear, knock, and get a "flat spot" in acceleration that took ~200 miles to work out.

That car took very well to water decarbonization.


Ha! That's interesting!

My WRX, being a 2016, and TGDI, and, being prone to LSPI, you would think it would have the most sophisticated anti-knock programming, that would keep it from knocking at all, right? Wrong. Must be a hard engine to keep from knocking, because it will knock from time to time. And I'm not talking the light pinging I get from the Taco on [censored] 87 fuel going uphill in hi gear in 100° heat. The WRX will knock hard enough that it actually momentarily loses power. This is if I'm in high gear at, say, 65 mph and I open the throttle to, say, 50-60% to pass someone. Classic LSPI. And I only use Top-Tier 93 in that car. Usually Shell. Must be hard to account for. But...then again...they didn't do a great job of EFI tuning in that car anyway.
 
Isn't 87 octane more volatile than 93?? Would the ECU actually advance timing for lower grade 87 due to it being more volatile?? And pull back timing for 93 being that it is less volatile?? Therefore the ECU would take more advantage of the 93 by pulling back timing and actually compensate for 87 being advancing the timing.
 
87 is more likely to ignite faster or too quickly if timing is not set up for it.... If timing is set up "slower" that's why vehicles recommend or require 93... Octane slows down the ignition process correct?? If that is indeed the case then it does effect volatility in a way does it not??
 
Originally Posted By: bbhero
87 is more likely to ignite faster or too quickly if timing is not set up for it.... If timing is set up "slower" that's why vehicles recommend or require 93... Octane slows down the ignition process correct?? If that is indeed the case then it does effect volatility in a way does it not??


Octane is not a measure of volatility (volatility is the speed at which something evaporates), Octane is a measure of knock resistance (the ability to resist spontaneous combustion, in the absence of an ignition source)
 
Originally Posted By: Olas
Originally Posted By: bbhero
87 is more likely to ignite faster or too quickly if timing is not set up for it.... If timing is set up "slower" that's why vehicles recommend or require 93... Octane slows down the ignition process correct?? If that is indeed the case then it does effect volatility in a way does it not??


Octane is not a measure of volatility (volatility is the speed at which something evaporates), Octane is a measure of knock resistance (the ability to resist spontaneous combustion, in the absence of an ignition source)
THIS is what most grease monkeys do not get.

However - I would like to add an important parameter to that statement ..................

It should read:

Octane is not a measure of volatility (volatility is the speed at which something evaporates), Octane is a measure of knock resistance (the ability to resist spontaneous combustion, in the absence of a spark ignition source).

Given enough pressure just about any fuel will spontaneously combust in ideal conditions (read possible knock condition if it happens at an unwanted position of the piston travel) and its exactly this phenomena that timing optimization needs to prevent. (Note that most engines actually make use of the fuel mixture's ability to combust at certain pressures created by and after the initial spark ignition point in order to improve flame speed propagation and burn efficiency in the combustion chamber)
 
Last edited:
Yes knock resistance.... AKI... But what causes the knock???? Ignition too quickly right??? Or no?? What does octane really do??? Slow down or create resistance to ignition... Does that not have a relationship to how fast something ignites?? It does I think.. maybe. Most regular people think of volatility in terms of how easliy something or some substance can catch fire or blow up. Water is not as volatile as gasoline. But in your definition of volatility it would be the other way around. Which would be correct. Strange in a way but true if taken to a very technical level.
 
Status
Not open for further replies.
Back
Top