A question of power
Registered User
Joined: Apr 2003
Posts: 14,273
Likes: 2
From: Chester
Originally posted by bmarshall
Not quite. The ideal is adjusting the ignition up to, but not at the point where knock happens and keeping it there. Then you're getting the best performance you can. The F20C's knock sensor is pretty basic like most cars. The signal is filtered from it, to try figure out if the noise is knock. But this isn't perfect - you can still get some knock that the sensor doesn't rate as knock. Also, it only happens after knock has started. If you're running low octane fuel, the car constantly advances cycles between knocking, and backing off the ignition timing a lot. It doesn't remember what the fuel is - just cycles like that. Which means you can knock a lot. So while the engine will generally not knock for long (unless it's in the window where the knock sensor doesn't pick it up), it will knock quite a lot.
What you really want is ion-sensing ignition, then you can run the timing right on the edge of knock, without getting knock. But I don't expect that on a road car any time soon.
-Brian.
Not quite. The ideal is adjusting the ignition up to, but not at the point where knock happens and keeping it there. Then you're getting the best performance you can. The F20C's knock sensor is pretty basic like most cars. The signal is filtered from it, to try figure out if the noise is knock. But this isn't perfect - you can still get some knock that the sensor doesn't rate as knock. Also, it only happens after knock has started. If you're running low octane fuel, the car constantly advances cycles between knocking, and backing off the ignition timing a lot. It doesn't remember what the fuel is - just cycles like that. Which means you can knock a lot. So while the engine will generally not knock for long (unless it's in the window where the knock sensor doesn't pick it up), it will knock quite a lot.
What you really want is ion-sensing ignition, then you can run the timing right on the edge of knock, without getting knock. But I don't expect that on a road car any time soon.
-Brian.
98 Ron fuel will knock as hard as 95 Ron it's just at the compression level during the burn that affects when this occurs. This means that 98 Ron can have further advanced timing than the 95Ron fuel (hence power and economy increases) but will have the same knock level since the knock sensor is controlling to the same level.
Joined: May 2001
Posts: 3,332
Likes: 2
From: On a fencepost
[QUOTE]Originally posted by fluffyninja
[B]
But this will still happen with 98Ron fuel as it will with 95 Ron. The detection threshold will be the same (i.e once the noise amplitude reaches a threshold level the ECU retards the timing) as there is no change in ECU or knock sensor obviously between fuel types.
[B]
But this will still happen with 98Ron fuel as it will with 95 Ron. The detection threshold will be the same (i.e once the noise amplitude reaches a threshold level the ECU retards the timing) as there is no change in ECU or knock sensor obviously between fuel types.
Registered User
Joined: Apr 2003
Posts: 14,273
Likes: 2
From: Chester
[QUOTE]Originally posted by bmarshall
[B]
In theory, yes, in practice nope. Remember that the ECU has only advances up to the limit set in the map. But retards a lot when it sees knock. Well, the standard maps are quite conservative on timing so on 98 fuel there's still a margin with the ignition as advanced as the ECU will make it. So you don't reach the point of knock.
While most people tune fueling using a V-AFC or similar, there's actually a lot of power to be made from more advanced ignition timing -
[B]
In theory, yes, in practice nope. Remember that the ECU has only advances up to the limit set in the map. But retards a lot when it sees knock. Well, the standard maps are quite conservative on timing so on 98 fuel there's still a margin with the ignition as advanced as the ECU will make it. So you don't reach the point of knock.
While most people tune fueling using a V-AFC or similar, there's actually a lot of power to be made from more advanced ignition timing -
Joined: May 2001
Posts: 3,332
Likes: 2
From: On a fencepost
Originally posted by fluffyninja
So, if I understand you right you're saying the map is set to not allow the knock sensor to retard enough to prevent knock with 95RON fuel? Is this just an increased risk or does it happen with all engines?
As for the Optimax how can it get a 98RON rating then if it's more likely to knock? The RON rating is based on the knock characteritics as compaed to a reference fuel?
So, if I understand you right you're saying the map is set to not allow the knock sensor to retard enough to prevent knock with 95RON fuel? Is this just an increased risk or does it happen with all engines?
As for the Optimax how can it get a 98RON rating then if it's more likely to knock? The RON rating is based on the knock characteritics as compaed to a reference fuel?
It happens on all F20C's I've logged if you run on low octane fuel.
RON ratings are produced by running the fuel in a research engine, just as you stated early on. But changes to the combustion chamber as one example, change what octane the same fuel is rated at. It's also possible for different fuels to be rated at the same RON, but burn differently. (Slight digression, but the MON is far more important to a modern high compression engine like the F20C than the RON.). Now Optimax is rated at over 98 Ron (There are different figures quoted, but around 98.6 seems a good average.). That's the figure from running in the research engine. Now Optimax is a different fuel to other 98 Ron fuels, with different burn charateristics. So when you run Optimax in an engine with a different combustion chamber design (etc... loads of possible variables) it's effective Ron in that engine will be slightly different. Compared with other fuels rated the same in the research engine, it could have a higher or lower effective Ron when run in a different engine.
-Brian.
Registered User
Joined: Apr 2003
Posts: 14,273
Likes: 2
From: Chester
What level ECU's do S2000's use then? (i.e 16bit, 32bit etc?)
The ones I'm used to working with are 32 bit and the control system constantly shifts the timing to hold it just before knock starts (we have a logging system at work and deliberately have to set the engine knocking to abuse the pistons and you can watch the signal from the ECU kicking in and out). This means that it doesn't really matter what fuel you put in (as long as it's petrol
) you'll just get a variation in performance without the engine being damaged
The ones I'm used to working with are 32 bit and the control system constantly shifts the timing to hold it just before knock starts (we have a logging system at work and deliberately have to set the engine knocking to abuse the pistons and you can watch the signal from the ECU kicking in and out). This means that it doesn't really matter what fuel you put in (as long as it's petrol
) you'll just get a variation in performance without the engine being damaged
Joined: May 2001
Posts: 3,332
Likes: 2
From: On a fencepost
Don't know the specifc processor details of the ECU. IIRC it's an OKI chip, which has the map's flashed on - so apart from a few people who've got hold of a Honda Development ECU, no-one knows much about the maps.
Note that I've not been saying that ECU's can't intelligently adjust the timing to acount for fuel and avoid knock, just that the F20C doesn't.
The ECU map's changed around My2002 - but I've never seen any firm evidence that the change-over exactly matched the model year change. MY2002+ ECU run a little bite more ignition advance than older cars, making a small amount more power. It's possible Honda changed some behaviour with regards the retard when you see knock. But I've not seen any work anyone's done on these later ECU's to find out. I've logged one, but not for long enough to be sure.
-Brian.
Note that I've not been saying that ECU's can't intelligently adjust the timing to acount for fuel and avoid knock, just that the F20C doesn't.
The ECU map's changed around My2002 - but I've never seen any firm evidence that the change-over exactly matched the model year change. MY2002+ ECU run a little bite more ignition advance than older cars, making a small amount more power. It's possible Honda changed some behaviour with regards the retard when you see knock. But I've not seen any work anyone's done on these later ECU's to find out. I've logged one, but not for long enough to be sure.
-Brian.
Registered User
Joined: Apr 2003
Posts: 14,273
Likes: 2
From: Chester
When was the ECU update and did it correspond with the Squirt Jet Recall?
Spoke top a few people at work about that recall and we reckon that they were having problems with piston cracks and wanted additional cooling of the pistons and the spark plug change could also affect this.
Maybe ECU change was somehow related? If there is a degree of knock this could be hurting the pistons and like some people on here there's going to be a few people who can't/won't use 98ROn
Spoke top a few people at work about that recall and we reckon that they were having problems with piston cracks and wanted additional cooling of the pistons and the spark plug change could also affect this.
Maybe ECU change was somehow related? If there is a degree of knock this could be hurting the pistons and like some people on here there's going to be a few people who can't/won't use 98ROn
Registered User
Joined: Mar 2003
Posts: 19,285
Likes: 0
From: SE VTecville
On a serious note, what exactly does the knock (as I understand it, a specific threshold vibration frequency of the engine) relate to in practical terms?
Who sets it, and how does it vary, and should it vary for different engines?
Kinda worrying if the 'familiar rattle' - by which, fj, I presume you mean the one at low revs (almost idle) - is related to 'minor detonations'
San.
Who sets it, and how does it vary, and should it vary for different engines?
Kinda worrying if the 'familiar rattle' - by which, fj, I presume you mean the one at low revs (almost idle) - is related to 'minor detonations'
San.