So at 4500 at peak torque where the ECU allows the boost to ramp up to "just before" knock, it senses the first knock and it takes what 30 revolutions which at 4500rpm is 0.4s to back off timing then achieve the "restoration of timing" -during this period you don't lose all power (particularly if your overboost is hitting 1.2bar) so in practice it doesn't affect the on raod acceleration much ?

This is interesting because the other element being "pulled" at the moment of knock detection is the boost - now my experience shows that the Motronic can be caught out here and cannott reduce the boost quickly enough under some conditions which leads to the ignition cut out interrupting for a fraction of a second.

This is all moot for stock engines, but as soon as some hot shot messes with your ECU, this is what I believe is happening - going from your 408hp to your 440hp (and more importantly that big extra mid range torque) makes the Motronic work itself away from the factory limits and much closer to that knock threshold......?

Flaws in the above please ???

 

Felix,
You are confusing knock re-test with retard decay. The knock sytem loops round "retesting -reduce retard if no knock -retest etc,etc ". The timescales depend on the knock severity .

Knock is bad and is to be avoided at all times . The cylinder pressures involved are extreme and will very quickly destroy a piston/rod . I cannot imagine a scenario where an engine was allowed to knock due to excessive boost.
Ah well , not my engine /money I suppose.


Geoff

 

Geoff - I'm just trying to fathom out what the ECU is actually doing - I have described it in detail above - you seem to be saying that it is some pre programmed limit which determines when the boost backs off at peak torque then it follows that map/limit all the way to the red line ?
You don't have to worry about my engine - plenty of experience went into its programming, unfortunately they don't want to talk to customers about these sorts of things - they are more interested that their engines out perform all the competition -hence why I can't tell YOU the answer !

So if what you say is correct and the engine is pulling boost at peak torque becuase of the program then how can chassis dyno wizards get even close to factory type accuracy in the programming or do they just add % increases right accross the range and rely on the safety systems (knock control) to save the engine when their programs are compromised by "adverse" conditions.

 

TB,
I came into this thread when I read how knock was being used to optimise performance . I can assure you that is definitely not the case.

As to what your engine is doing , well that is in the hands of RSTuning and their mapping .I would only make the observation that knock in any form is bad news. If it were my engine I would not want to hear knock under any circumstances.

The 993tt system is rather obtuse in operation and at the same time OBDII had to be incorporated into Motronic.The guys at Bosch certainly had their hands full !

All the best for 2007

Geoff

 

Interesting change of direction in this thread.. In order to establish some baselines for this thread, we have to get some basic tuning guidelines and assume that our alleged engine has large enough IC, injectors, and high enough octane..

One will usually tune for peak torque, we will run up the RPM and can load the engine at a given RPM target. Timing is adjusted to net out peak torque numbers.. There is a certain point with timing that one will see a increase, then the peak number starts to fall off. One notes that number and goes to the next cell. Using a PST2 tester one can see if knock is present. With these engines we do not want to "have" knock.. I also must state that there comes a point that you can through a ton of boost at lets say your peak torque RPM and there will become no change in the output of power.. You will creap up on the boost numbers.. Lets say 1bar, 1.1, 1.2, 1.3 If you have dialed in the last boost increase on the dyno run and nothing has changed=no net torque increase. That is the number that you use as your boost threshold. Delivering more boost or more timing has no result on power/you actually lose power. One thing that we are forgetting is that if we start to move into "custom" engine builds, the variables will start to change.. Different camshaft grinds, intake porting and larger exhaust and intakes. Now with modern test equipment and data logging, we should see the echo or knock and not exceed these net numbers in the final mapping. And even if we are running more timing at a given RPM if we see a cylinder having knock being pulled a few degrees we realize that we have to back down the net value of timing.. The ECU does this while "tuning" the engine. Knock control is in place to save the engine. Knock should not be there. However, on uprated engines with less than perfect octane.. 90 and 91, there is knock that is being sensed and timing is retarding that cylinder. The echo of knock is always being monitored. The ECU will correct timing and if this does not cure it, Boost will be reduced.. This is one of the reasons why I hate EBC.. Because if there comes the situation where you have a bad tank of fuel, or god forbid a fuel pump on it's way out, the -10 to -12 degrees of pulled timing might not be on it's own to save your engine. Detonation IF present can and will destroy your engine in less than a second... The shock will detroy your ring lands, the piston will basically come apart.

Early in my understanding I always questioned the why.. One important fact was that with this ECU if one tunes for 100 octane fuel. There isn't enough retard to save the engine if a lower octane was introduced, or if fuel was being starved.

Please understand that this is the tip of the iceburg, and when tuning these engine everything is happening in milliseconds. Operating the dyno and datalogging and watching torque-boost and RPM. Does require skill. I for one am vary happy to have the Motronics knock function in place!!!

Toby if you are getting knock, then fueling octane is not high enough for your timing values. Boost and thermal increases should have already been accounted for when your mapping was established.

On a high performance map which I would consider a 100 octane map, knock should not crop it's head on these engines. If present the ECU should/will adjust for the counts vs being the norm.. If it becomes the norm, the mapping is incorrect for the fuel grade being used.

 

Thanks for all the input guys

I only ever guessed at the "knock" thing - I thought that knock came in progressively and the ECU started doing its thing on the first sound of knock.

Of course I've never "heard" it in my engine (from what Kevin says if you hear it its too late anyway ) I just was looking for an explanation of the control of the overboost.

Reviewing other tuner's engines and their relative lack of mid range overboost (torque -and I don't don't mean chassis dyno stuff ) I truly still think RS do it differently and take it closer to the knock limit.
I will try and get a definitive answer - but thanks for the good feedback which I'm sure is correct for standard/most tuned motors and fits with how Porsche designed the Motronic to function in this area.

 

I too have seen the boost behaviour as TB describes it.

Last week my car at 2.5 deg ambient temp was 'only' boosting to just under 1 bar, after initially overboosting to 1.1. A few days later at an amb of 7.5 deg it was boosting just past 1 bar, and overboosting to 1.2 bar.

From memory in the hot summer it would boost to 1.1 - 1.2 bar ALL the time, with overboost up to 1.5 bar* seen.

I agree that it must be the Motronic control doing its thing, and only allowing the boost to reach the required level to 'make' the dyno numbers seen at RS.

I will make some videos as soon as I can of my boost gauge at diff amb temps and gears, etc. Will be fun

A Happy New Year to all


Martyn.

* This mad level of apparent overboost is thought to be a result of the time lag between the ecu signalling the wastegates to open and the gates actually physically opening. It lasts for a very short time before returning to 1.2 bar. Where as the o/b in cooler conditions seems to last a fraction longer.

Another interesting experiment would be to check the boost levels on the same day on the car with a stock intercooler opposed to the Secan.

 

TB,
I went back and read your original post ! I dont know if a quick review of the system would be useful but here goes.

BHP is proportional to air mass consumed.

When the air density is high ( cold air/high atmo pressure ) fewer cfm are needed to make the same mass flow so boost is lower for the same bhp.

When the air density is low ( hot intake air/low atmo pressure ) more cfm are needed to make the same mass flow ,so boost is higher for the same bhp.

So Motronic can vary boost to hold bhp levels sensibly constant within the constraints of the engine system thermal limits

The low density effect becomes very important when considering turbo aircraft engines ,as air density takes a real drop with altitude. In that application the boost level- compressor speed , has to be capped to prevent turbo self destruction !

A look at a turbo compressor cfm/pressure ratio/shaft speed plot will tell all .



I hope that makes sense.

Geoff

 

Thanks for the insight TB, one of the better 'motronic' over 'fixed boost' explanations I've read - as for the rest of this thread I'm totally out of my depth....still enjoy reading though!

On this note - do you know if 'RS tuning' ecu's retain motronic control and would alter boost depending on air temperature?

 

Kennyboy,

Yes !!

All the best for 2007

Geoff

 

Thanks Geoff, happy new year.

 

Does the PST2 display knock in realtime or does it display a knock count after 10,000 ignition events like the Bosch Hammer?