Jeff96-993

Those are all great questions, and I've been reading this thread to learn something, as others have, so answers are always very appreciated. I've never heard of individual cylinder knock control, nor how it gets implemented on a distributor system ignition. But I'm very interested in hearing info that's not readily available, and I haven't seen it discussed here on Rennlist. But when it is, I'll be reading through it for a better understanding.

trophy

I never said that none of them, There are quite a number that do but the large majority don't. Where does this come from?, I have a number of friends that have worked on these systems via Motec, Pectel and Bosch motorsport and would be sure they would love to get into a technical conversation with you, if you could leave your cynicism at the door.

I am not making any contributions, other than trying to filter the crap you continue to type... And more so supporting ingenuity in members of this forum, something you cannot do without **** canning every step of the way...


Loren, you keep telling us you are the expert, why don't YOU tell us the answers instead of creating the list... If you answered the se I am sure the members of this forum would be more open to your posts.

Your a smart guy, and know more than pretty much anyone on the forum regarding Motronic and the ignition systems of Porsche's, you just suck at communicating.....

A little FYI, I have an aftermarket ECU in my car, and the Stock Knock sensors are used and the system is programmed in the same way as the stock system.... I like the protection the knock control gives...

trophy

Individual knock control is relatively simple, and the cars ECU has an algorithm to deal with.

Our cars use 2 Bosch knock sensors, one for each back of cylinders, therefore each sensor reads the knock of 3 cylinders.

The ECU in my car (aftermarket) calculates which cylinder is knocking by comparing the time of the knock with the current position of the engine from the flywheel sensor and hall sensor in the distributor.

I assume the stock Motronic system would do exactly the same thing, Loren could confirm.

In simple terms, The ECU then retards the ignition in increments till the knocking stops. Once there is no knock the system will eventually bring back advance again till the max advance as per the ignition table in the ECU, and the cycle begins again should knock occur.

Jeff96-993

That was always my understanding, so my question was around the individual cylinder. Each bank has a sensor, and each sensor just tells the system to retard the timing... I wasn't sure where the "individual" part of it came in, so was asking for clarification on that point.

And then I couldn't figure out how to implement a solution like that on a distributor engine. I'd imagine it was much easier on a sensor engine with more mapping capability.

But I'd never heard it done there either, so maybe it was a pointless question or I have a lot more to learn.

trophy

Never any pointless questions, just pointless answers. . Or maybe pointless ignition

Individual knock sensing can happen a number of ways:

1: By having a sensor on each cylinder. Then you know when an individual cylinder is knocking.
2: By having 1, 2, 4 sensors and then using the time of the knock and compare this to what cylinder is on the ignition stroke. the ecu can then determine what cylinder is knocking.

So the individual cylinder knock is either actual via individual knock sensors or calculated as it is on our cars

The distributors in our cars are for just that, distributing the spark to the correct cylinder in firing order. It does none of the timing, that is the job of the ECU. However the rotor and the length of the electrode does have an upper and lower limit to the timing (max amount of retard and advance) The ecu works within these parameters to provide the ignition timing for the engine. The ECU fires the ignition module when it needs spark, the timing of this gives the ignition timing for the engine, as long as it happens when the electrode (sure there is a proper name for this part)is contacting the respective terminal inside the distributor, the spark plug fires.

If the engine receives a signal from the knock sensor then it will retard this cylinder by 3 degrees, if it still knocks will retard by another 3 degrees up to a max of 9 degrees of retard.

Lorenfb

"you keep telling us you are the expert"

No, don't think so! Just challenge the claims made by some and raise issues.
Isn't that one of the values of a this forum? It's really too sad that some,
because of the disagreement, take it personally and have to respond by
a personal attack. The challenge is claim being made and not the person
making the claim. Rennlist members benefit from receiving all views on a topic.

"I am not making any contributions, other than trying to filter the crap you continue to type"

Rather than do personal attacks, post what you feel is "crap" and corroborate it!

"I've never heard of individual cylinder knock control, nor how it gets implemented on a distributor system ignition."

The simple way it's done is to use a cam sensor (Hall Sensor in the 964/993)
and store a count (flywheel sensor) relative to the Hall Sensor occurrence.
If the Hall signal is missing, all cylinders get retarded the same amount.
It can also be done on an engine with a distributor by sensing the first time
a spark occurs on a known cylinder and 'remembering' where it occurred.

"However the rotor and the length of the electrode does have an upper and lower limit to the timing (max amount of retard and advance)"

That's correct. It's the width of the rotor tip and its alignment to the appropriate
distributor cap electrode that allows for correct spark distribution. If the rotor
is not aligned properly or the rotor tip too narrow, the wrong cylinder can receive
the spark over the timing range which is not the case for a conventional non-electronic
timing ignition where the rotor 'follows' the ignition point.

"Why Bosch installed a second dedicated microprocessor in the 964 DME
ECM for knock control if just for poor fuel grades when in the past they
used a simple jumper and/or a rotary switch for different fuel markets."

Porsche/Bosch considered knock control critical to further enhance the
performance of the 911 engine while maintaining long term reliability.
Given the complexity of analyzing a knock signal which is in the time
domain and for reliable results then converting it to the frequency domain,
i.e. a Fourier transform for digital analysis, a dedicated processor was required.
The 993 used a much faster microprocessor and didn't require a dedicated
one for knock control.

"Why was a twin plug approach AND knock control used starting with the
964 with its much higher CR than the early 911 3.2."

Porsche used more aggressive advance 'curves' in the later engines and
further compounded the potential detonation problem by increased CRs
which necessitated twin plugs (no ideal center plug location) and knock control.
Knock control is to an ignition system as is a wide band O2 control to a fuel
system. Both provide a closed-loop feedback system to maximize torque
under varying engine conditions while maintaining engine reliability.

trophy

Great information Loren...

Juha G

Good information, thanks Loren!

Knock control is definately good, cannot deny that. But is it obligatory? I say no it's not.
Like Steven posted, a lot of the best racing teams in the world don't use knock control. And then there are tens of thousands of club racers all around the world pushing their race tuned engines to the limits without knock control...

My tuner's opinion is that knock control is not worth the effort. I take his word on this as he has built and tuned racing engines for living for 20 or so years.

If you are concerned about the missing knock control, just wait untill next Summer when I have tracked my high compression ratio, air cooled, SUPERCHARGED 993 with Motec and no knock control. If the pistons and the rings can take that, then I'd say it's quite safe to assume you don't need knock control in a similar NA engine.

Despite the fear of sounding like a broken record, I want to make it clear that this "project" is not about making profit but only a much loved hobby for my friend and I personally am just trying to contribute to the 993 community.

Tonight we will work on the correction factors with the aid of wide band lambda out of my project BMW. So stay tuned.

evoderby

Were you bullied in school?

This thread is sounding nothing like a performance chip sell. Juha is very upfront in what 'his' ecu does and doesn't do:

i.e. it offers flexibility in mapping in a way the standard motronic doesn't (at least not without significant time spent on reverse engineering), it does so without the need of adapting wire looms and replacing sensors asociated with other aftermarket ecu's, it's plug and play at a fraction of the cost asociated with the likes of Motec.

That's it, and at this price who's ever going to make money on this??? Again this has nothing to do with a performance chip 'sell' as you call it.

You are right in stating Juha's ecu features no knock control, which in itself is a very useful feature. In fact I use aftermarket individual adaptive knock control myself to great effect in further pushing performance.

However without it tuning for 98Ron MBT, and pulling ignition 2 degrees back from there in combination with IAT thresholds to further pull back timing should inlet temps become dangerously high, gives a very reliable path to significantly improve performance over standard Porsche (non rs) mapping.

How many NA 9m motec engines have grenaded due to knock? The answer is zero. Could they make even more power with individual adaptive knock control? Definately so!

Your examples of 962 and 959 brings turbo boosting up to 3bar into the equation (remember 959 had to be homologated for group B). A rather different playing field than NA 85HP/liter applications!!

However, should anyone feel the urge to use their 993 to tow their caravan up the mountain on a hot summer day in 5th, running bulgarian fuel....don't use Juha's ecu. If you are unhappy being stuck with the standard Motronic, it might be an option for you to consider.

Best polish??? That's an easy one.....swisswax!

Have fun,

Harald

Juha G

Good post Harald!

Loren reminds me of a guy who frequents on a local BMW Forum here. He takes part in racing and track driving topics, is very knowledgeable but has absolutely no communication skills. I thought he was a total jerk, but that was only until i met him. He ended up being a 60-year old ex pro racer who shaved 2seconds off my lap tie in just three laps sitting next to me. He was also the nicest guy in person.
So morale of the story is that don't judge people based on their internet communication skills...

Flying Finn

Maybe Loren is a Finn? Just saying...

And what's already been said, way to go keeping your cool Jussi.

Juha G

Hahahh!!!! Maybe...And maybe I'm getting old too...

So, today we managed to dial it in with the aid of the wideband lambda.
As Jeremy Clarkson would put it; "LISTEN TO THAT NOISE!!!!"


(The video may take some time to start working but you can download it to your computer to view it right away)


As you can see, there is quite a lot of fuel thrown in. That's because we are running alpha-N based ignition and fuel. i.e. the ECU cannot adjust to engine load, hence running too rich at WOT and no load on engine.

The proto is pretty much ready to be sent to Mike (Stealth) but we will do one more thing. And that is to add the MAP sensor. We will dial in some very basic mapping so that Mike can start the car and idle. Unfortunately we cannot do proper mapping as the car would need to be driven for that (to put load on the engine).
So Mike can either try the alpha-N based mapping (same as my 993 with Motec right now), or start from scratch with the MAP based tune.

evoderby

Some highly enjoyable comments on Finnish mentality @ 4:20 in the following clip where Mika Hakkinen teaches Top Gear's James May how to drive properly.


Some nice tongue in cheek references to Clarkson's 'loud' mentality as well.

Eric_Oz_S2

Just visiting here from the lowly 944 threads and curious as to what the benefits are of stand alone unless you are making serious engine changes. The 993 using a similar ECU to the 944S2 and 968 (M2.1 variant IIRC). I'm not sure why you wouldn't just retune the Motronic using an Ostrich EPROM emulator and TunerProRT. The Motronic knock control is very important to achieve decent power and engine safety, and standalones will never be quite as good due to the noise filtering algorithms that Porsche designed into the Motronic to suit these particular engines. I have noticed a few incorrect statements in here - like the Motronic's inability to change injector size. Actually, scaling injectors isn't terribly hard (at least on the 2.1 in the S2). Obviously if you go for very large injectors (and why would you unless going to forced induction) you will have an issue at idle as you will need to run semi-sequential maps to even out the fuel delivery. But for a NA engine with some hot cams and free flowing exhausts, the minor changes required can easily be adjusted in the WOT fuel maps. The Motronic is fairly clever, in that it uses the PT lambda adaptation values as an input into the WOT trim. So if the DME detects a lean condition at a certain RPM range in the PT maps, the WOT maps are compensated to some degree. This is a form of long term adaptation to account for engine wear, fuel types, etc. How do I know all this? Well, I have remapped my Motronic to suit a supercharger and tuned etc. I can say for sure that knock control is a necessity!

So unless you are looking at major power increases, I think you would be far better off just retuning the motronic.

Just my 2 cents.

evoderby

Hi Eric,

Good post, sounds like an interesting 944 you have.....I loved my 944 Ring tool, supercharging it would definately had made it even more of a blast.

Two short comments:

1)Motronic only operates narrow band lambda, which is basically an on/off switch around 14.7:1 and therefore is way too crude to completely rely on.

2)The above doesn't mean the Motronic can not successfully be remapped using emulation software. In my personal experience mapping this way is less intuitive than a well sorted out EMS and accompanying software, i.e. Aftermarket EMS mapping and future fine tuning can be done faster which saves money....especially when tuning in stages, adding cams, ported heads, itb's along the way. At the end of devlopment I wished I had bought aftermarket EMS to begin with.