[Apex]

So I usually fill up with 93 octane when I am towing. However I have been noticing posts saying people use 93 octane in the tank before towing. I am guessing so the engine has adapted to the new fuel first and then the full benefits are realized when actually towing.

I have noticed the truck runs strong after a big towing trip. Or maybe I am just getting used to going trailer less?

 

[Carsmak]

Although much lighter than most items towed on here, I noticed when I took my friends bass boat to Lake Havasu & Lake Mojave it seemed like it drove better/stronger around Laughlin, Needles & Lake Havasu without the boat and better on the return trip. I'm guessing the computer needs time to change the "fuel mapping"? It's only a guess, but that may be the reasoning to run 1 tank of 93 before a trip.

 

[zroger73]

All modern vehicles use "knock sensors", which are basically microphones that are tuned to generate a signal when they hear "pinging" (detonation) long before the human ear can. Constantly and in real time, the computer listens for detonation. Within milliseconds of detecting it, the computer retards timing in an attempt to reduce the detonation to acceptable levels. To maintain performance and emissions, there are limits on how far the timing can be advanced or retarded. The computer can only retard timing so much at which point the human ear will hear the detonation if it is severe enough. Keep going long enough and you'll have engine damage. There is some "smoothing" done inside the computer, but a vehicle will typically respond to changes in fuel quality within a fraction of a second (essentially, "real time"). It doesn't take hours or miles of driving for this type of adaptation to occur. The computer is constantly altering the timing just enough to prevent detonation, which is why is basically responds to changes in fuel quality in real time - any perceived effects over a longer time will be psychological.

Now, some drivetrain parameters do take hours or miles of driving to adapt such as transmission shift timing and firmness and idle speed compensation, but ignition timing as it related to fuel quality isn't one of those items.

 

[csimo]

So I usually fill up with 93 octane when I am towing. However I have been noticing posts saying people use 93 octane in the tank before towing. I am guessing so the engine has adapted to the new fuel first and then the full benefits are realized when actually towing.

I have noticed the truck runs strong after a big towing trip. Or maybe I am just getting used to going trailer less?

Just a waste of money. There is only one base fuel map, and the ignition advance is determined (and limited) by the knock sensors and fuel map. In other words, if the engine load demands it the system will advance the ignition timing until either the limit of the fuel map or pre-detonation is detected. This happens hundreds of times per second. So the system doesn't know if you had 87, or 93 octane rated fuel in it even one second ago.

In all my testing at various loads and altitudes the ECU will max out ignition timing at about 90 octane rated fuel. So anything above 91 octane rated fuel is not going to help (technically you actually lose a small amount, but it's not enough to matter).

 

[northernlights]

So what can happen when this knock sensor is on the frits or totally goes out. Bad mpgs, CEL on, truck wont start,ect... what does one look for or at, Just wonderin.

 

[Dnick]

So what can happen when this knock sensor is on the frits or totally goes out. Bad mpgs, CEL on, truck wont start,ect... what does one look for or at, Just wonderin.

That's one of the sensors that can easily get knocked out by critters... the cable for it runs under the intake manifold & interfers with free passage through there for chipmunks, etc., so they just chew it out of the way. I had this happen a few year ago while in the Sierras for a few weeks.
What happens is you get dash lights telling you something's up (I don't recall if it's CEL, or something else). Easily diagnosed by dealer, and engine operation was not noticibly effected (local dealer said I could wait til I got home to get repaired (300 miles or so), which is what I did w/o event. Just FYI.

Oh and BTW, the replacement cable is now metal-braid shielded to prevent it from again becoming a snack for bored rodents! And I think it's now a bit longer too so that it doesn't present a 'trip hazard' for the critters (as it was in it's fairly tight/short form...IIRC).

 

[csimo]

So what can happen when this knock sensor is on the frits or totally goes out. Bad mpgs, CEL on, truck wont start,ect... what does one look for or at, Just wonderin.

CEL is set and fuel and ignition timing are set to limp mode... just the default map. Nothing will be damaged unless you somehow used some seriously inferior fuel. Reduced performance. Nothing serious... the failure is to the safe side of things.

 

[csimo]

So what can happen when this knock sensor is on the frits or totally goes out. Bad mpgs, CEL on, truck wont start,ect... what does one look for or at, Just wonderin.

CEL is set and fuel and ignition timing are set to ECU open-loop mode... just the default map. Nothing will be damaged unless you somehow used some seriously inferior fuel. Reduced performance. Nothing serious... the failure is to the safe side of things.

 

[Apex]

So basically what you are saying is the computer will advance timing until the knock sensor determines the slightest pre ignition, then The computer will retard the timing.

Makes sense to me but is there any other reason the computer would have to "get used to" the higher octane?

 

[zroger73]

So basically what you are saying is the computer will advance timing until the knock sensor determines the slightest pre ignition, then The computer will retard the timing.

Makes sense to me but is there any other reason the computer would have to "get used to" the higher octane?

In a vehicle designed to run on "regular" gasoline, the base timing is calibrated so that detonation does not occur when using this grade of fuel. Using "premium" gasoline won't improve performance.

In a vehicle designed to run on "premium" gasoline, the based timing is more advanced to take advantage of the slower burn and to produce more power. The use of "regular" gasoline can cause detonation which will cause the computer to retard the timing back to levels appropriate for lower octane gasoline which results in a loss of power.

The object is to avoid detonation altogether before it is detected and compensated for. If all vehicles advanced timing to "premium" gasoline levels and "regular" gasoline was normally used, detonation would frequently occur and the computer would have to compensate for it by retarding the timing. All these brief periods of detonation add up over time. On the other hand, if the vehicle was tuned to run on "regular", then there is nothing to be gained by using "premium" gasoline.

Most vehicles are designed to run on "regular". Using "premium" results in no improvements.

Some vehicles are designed to run on "premium". Of those, some are approved for use with "regular", although you won't get the advertised power or economy. Others may actually be damaged by using "regular" fuel and this damage may not be covered under warranty. Still others require that "regular" be used only in an emergency.

Honda has tuned the Ridgeline to utilize "regular" gasoline under most driving conditions. Under certain conditions, such as competitive driving or heavy towing, detonation can occur on "regular" gasoline causing the computer to retard timing which results in a slight loss of power. Using a higher octane gasoline will reduce detonation under these severe conditions. The computer won't have to retard timing so you'll have the "full" amount of power available.

 

[Apex]

Good description.

So does the frequency of knock timing adjustments have an affect on the long term maps? Similar to long term fuel trims? Does the pcm have a long term adaptive strategy with regard to ignition timing?

 

[zroger73]

Good description.

So does the frequency of knock timing adjustments have an affect on the long term maps? Similar to long term fuel trims? Does the pcm have a long term adaptive strategy with regard to ignition timing?

Generally, ignition timing retards rapidly (in a fraction of a second) when detonation is detected and advances slowly (over a second or more) when detonation is no longer detected. I am not aware of an ignition timing equivalent of short and long term fuel trims, although such may very well exist. Keep in mind that even if such exists, "short" and "long" to the computer may be very different than our perception of those terms. For example, a short term trim value can change several times per second. A long term trim value can changes over several seconds (NOT minutes or hours) depending on what the short term value is doing.

Remember, the computer is capable of polling sensor readings anywhere from dozens to thousands of times per second, depending on the priority of that particular sensor. For example, a coolant temperature sensor may only be polled a few times a second because 1) the coolant can't physically heat up or cool down any quicker than that and 2) there is a limit to processing power and bus width. There's no sense in flooding a data bus with thousands of coolant temperature readings each second when one every few seconds is more than sufficient.

Anyway, back on track. An engine adjusts to things like fuel quality and air density in what is essentially "real time". There's no such thing as replacing a tank of 89 octane fuel with 93 octane fuel and the engine taking several miles of driving to adjust. It will adjust the moment the higher octane fuel moves from the tank to the injectors. In a returnless fuel system like almost all vehicles have now, it may take a few minutes for the new fuel to move through the lines. In vehicles with a return-type fuel delivery system where the fuel is constantly moving from the tank to the fuel rail with the excess fuel returned to the tank, the adjustment will take places in only a few seconds.

Learned values for some parameters such as transmission performance and idle speed compensation are learned quickly and refined over time. Example: The computer commands the engine to idle at 650 RPM. It will trim the throttle as necessary to maintain this desired idle speed. If the A/C compressor kicks on, the RPM will suddenly nosedive and the computer will have to compensate. If it overcompensates, the engine will slightly overrev the desired speed. To minimize this, the computer learns over time what effects a particular load will have on engine speed under various conditions. The compensation becomes proactive instead of reactive. When your A/C compressor kicks on, the computer slightly revs the engine a small fraction of a second before the compressor clutch engages so that when it does engage the idle speed will land very close to the desired speed.

 

[Apex]

So does the timing management logic dictate certain amount of response from the knock sensor? How often will the pcm "test" the limits of the ignition advance? The only way the computer can know it has achieved the maximum amount of ignition timing is if it recieves positive data from the knock sensor. What is the cadence the pcm prompts the knock sensor?

 

[zroger73]

So does the timing management logic dictate certain amount of response from the knock sensor? How often will the pcm "test" the limits of the ignition advance? The only way the computer can know it has achieved the maximum amount of ignition timing is if it recieves positive data from the knock sensor. What is the cadence the pcm prompts the knock sensor?

Now we're getting into propriety, unpublished, model-specific information. It is known generally how these systems work, but specific timings and degrees of advance/retard under specific conditions and polling rates are unknown. The engineers at Honda who designed the system know, but you won't get that information from them. Perhaps some aftermarket tuners and PCM programmers have reverse engineered the system. Unlike mechanical timing advance, modern ignition systems can alter timing as fast as the computer is programmed to adjust it and current PCMs are fast enough to adjust spark timing on a cylinder-by-cylinder faster beyond the mechanical limits of the engine if they are programmed to do so, but this is unnecessary.

 

[amindtat]

I could see running a tank of higher octane gas before a tow as a good practice just for the sake of avoiding dilution. How many people wait until they are on E to get gas? Do you fill up at 1/4? If so, there's about 5-6 gallons of regular diluting your higher octane gas. Do you use a gas pump with only one hose? There's about 1/3 of a gallon of whatever the person before you bought (most likely regular). Not a huge amount, but something I'm keenly aware of when filling my motorcycle that takes 93 octane.

 

[zroger73]

How many people wait until they are on E to get gas?

There's about 1/3 of a gallon of whatever the person before you bought (most likely regular).

Based on my observations and the number of cars I see on the side of the road that have run out of fuel, I'd say more than you think! I have NO sympathy for anyone who runs their vehicles out of fuel. This a completely avoidable event - there's just no valid excuse for it.

Great point about getting cheated out of a bit of octane and detergents while filling a motorcycle or portable gas can.

 

[Apex]

So the computer has no idea you just filled with high octane until it tests it by raising the timing above what has been previously limited by the knock sensor. Once it achieves a higher timing allowed by the higher octane fuel, how long does that last, and how long does that process take? I don't expect anyone to know this, but can you say for certain the process is less than burning through a tank of high octane? Also factoring dilution as mentioned above?

 

[48218]

As zroger said earlier, the 'base map' includes a timing advance curve for the designed fuel AKI (anti-knock index, or 'octane rating'). The ECU does not have a small-advance 'baseline' and then periodically try to advance timing to 'test' the limit. It has a large(r)-advance 'baseline' and retards timing in response to knock.

I can tell you that my Subaru, whose ECU I logged and watched a lot, and reprogrammed, would pull timing essentially instantly in response to knock, but it would not 're-advance' it quite as quickly. It would take a noticeable portion of a second, or more, to advance timing again, and it would make that advance smoothly rather than all at once.

It can be difficult to determine precisely what the ECU is programmed to do in this regard, particularly on a turbocharged vehicle like my Subaru, because the base map for ignition timing is a function of the combination of things. RPM, 'requested torque' (a misleading term), and calculated engine load would be primary factors for the base map. Air mass flow, knock sensor feedback, engine coolant temperature, and so forth drive advances or retardations from the base.

With all this in mind, I can tell you that I have regularly experienced what many owner's manuals refer to as 'light spark knock' in many, many vehicles of all different makes over the years. Knock sensors don't capture all knock events, and quite frankly ECU programming allows a certain amount of it. So, don't be too terribly alarmed if you hear a bit of light 'spark knock.'

 

[csimo]

Good description.

So does the frequency of knock timing adjustments have an affect on the long term maps? Similar to long term fuel trims? Does the pcm have a long term adaptive strategy with regard to ignition timing?

Quick answer is no. The transmission (TCU logic) has a bit of long term adaptive logic but not the ECU. As far as all the calculations go it doesn't know what happened one second ago (unless the airbags deploy and event recorder is involved).

 

[Apex]

So it seems to me that the computer adapts very quickly to the higher octane fuel. I guess people that buy a tank of high octane before they tow are wasting money.

Thanks to all.