tuning programmable engine management systems

[scarceller]

Knock control should only be a safety net, a well tuned engine should not hit the knock limit while running even at WOT. If a EFI system hits knock it normally simply pulls ignition out and this yields to less than best max Torque. A good Tuner looks for knock to stay away from it, they should not tune a engine intentionally to hit the knock limit.

Also worth mention is that a tuner looks at several factors for IGN. For example most EFI systems have ign compensation tables for IAT and CHT, as either of these increase the maps pull ign out as well. Knowing howto properly set these allows a engine to run more ign adv if IATs are cold, like in cold Winter weather. It is very dangerous to tune a engine in cold weather and not understand these concepts, this is a major mistake made by first time tuners.

[John at J&S]

Originally Posted by Lorenfb

Right. But the difficult part is getting the proper algorithm designed, which is a very
difficult task and is specific to each engine. One would need to characterize the
actual knock/pinging/detonation signal from engine noise and convert it digitally,
(a digital filter), e.g. using the Fast Fourier Transform.

Yes, the difficult part is getting the proper algorithm designed. You are describing a "brute force" approach that must be mapped for each engine type.

Of course, there are other ways to do this. I came up with an elegantly simple method over twenty years ago.

Tippy:

It sounds reasonable that an air-cooled engine would be noisier than one cooled by water, but that is not the case.

The SafeGuard has no problem working on VW, Porsche, or Corvair.

Partial list of some on pelicanparts that are using it:

spuggy
Nize
Steve@Rennsport
Cloggie
smurfbus
sjf
Louie928

[brcorp]

Subscribed

[Tippy]

I run low-15 AFR's at cruise easily. It's not until I see low to mid-16's before I get lean surge. Might be I get away with it due to pretty advanced timing at part throttle.

[jpnovak]

Originally Posted by Tippy

Might be I get away with it due to pretty advanced timing at part throttle.

Yes, If you can not advance timing, your call will be unhappy under lean conditions.

[EPorsche]

Subscribed

[Tippy]

Yes, If you can not advance timing, your call will be unhappy under lean conditions.

After reading Sal run 50+ degrees at part throttle, low loads and "psalt" stating the same for good MPG's, I'm going to try this soon and see how it goes.

[gsmith660]

Originally Posted by EPorsche

Subscribed

Hey eporsche you need to answer your PM's

[scarceller]

This is an excellent article written by OEM engine management expert (Allen W. Cline) in Detonation and Pre-Ignition I thought folks may enjoy the detailed read:
Engine Basics: Detonation and Pre-Ignition by Allen W. Cline

This article explains Detonation and Pre-Ignition in very clear easy to understand terms. Pay very close attention to his points about Cyl Temps throughout the article because in our air cooled engines these tend to rise extremely quickly. You'd be very surprised to learn that CHTs rise very quickly without being at WOT. The biggest offender is lugging a 911 engine, where lugging is a term meaning that you are applying moderate to heavy load at low RPMs like 4th gear at less than 2500RPMs. I have a CHT monitoring system in my 3.2L and you can see CHTs rise amazingly fast under this condition especially in HOT 90F+ weather. Do this to our motors then pull a WOT run and your in big trouble! I never run my car below 3000RPMs in hot weather.

[joeblow]

Originally Posted by scarceller

This is an excellent article written by OEM engine management expert (Allen W. Cline) in Detonation and Pre-Ignition I thought folks may enjoy the detailed read:
Engine Basics: Detonation and Pre-Ignition by Allen W. Cline

This article explains Detonation and Pre-Ignition in very clear easy to understand terms. Pay very close attention to his points about Cyl Temps throughout the article because in our air cooled engines these tend to rise extremely quickly. You'd be very surprised to learn that CHTs rise very quickly without being at WOT. The biggest offender is lugging a 911 engine, where lugging is a term meaning that you are applying moderate to heavy load at low RPMs like 4th gear at less than 2500RPMs. I have a CHT monitoring system in my 3.2L and you can see CHTs rise amazingly fast under this condition especially in HOT 90F+ weather. Do this to our motors then pull a WOT run and your in big trouble! I never run my car below 3000RPMs in hot weather.

Yep this is a common misconception that by running at lower RPMs you are putting less load on the engine and creating less heat. The opposite is true especially on aircooled engines with a direct drive fan. More RPMS = more air cooling, more oil flow through the cooler and lower load per firing event in the cyls.

X Load/Y RPMs = load per revolution

Good stuff!

[jsveb]

Very interesting article. Thanks

I am nothing but a rookie that wonders, but here goes...

On the turbo side of the pelican forum, there has recently been a thread about auto tune features in modern efi systems, and wether this is possible or not.

My question is regarding ignition timing and tuning.

Wouldn't it be possible to use a combustion pressure sensor, I.e. one of these:
https://www.avl.com/pressure-sensors-for-combustion-analysis

Based on the peak pressure and cam/crank position sensors, have the efi adjust ignition timing so peak pressure occurs at 14 degrees after tdc, optimum?

[scarceller]

Originally Posted by jsveb

Very interesting article. Thanks

I am nothing but a rookies that wonders, but here goes...

On the turbo side of the pelican forum, there has recently been a thread about auto tune features in modern efi systems, and wether this is possible or not.

My questions is regarding ignition timing and tuning.

Wouldn't it be possible to use a combustion pressure sensor, I.e. one of these:
https://www.avl.com/pressure-sensors-for-combustion-analysis

Based on the peak pressure and cam/crank position sensors, have the efi adjust ignitin timing so peak pressure occurs at 14 degrees after tdc, optimum?

In theory you could use autotune but it's not that simple. The very best way is to use a load dyno to determine Peak Torque for a given condition. It simply turn out that peak torque occurs around that 14deg peak pressure range.

The other much more dangerous issue is that peak torque can't always be done because detonation prohibits it.

In the past I've also mentioned how important it is to understand the effects IntakeAirTemp has on peak torque. A common mistake is to tune on a cold 30F day, dial in WOT ignition and call it a day. Huge mistake! The ignition values that can be used with cold IATs won't work once IATs get hot like 90+F! Good ECU systems have IAT and CHT Ignition comp tables, the stock DME has these built in. Knowing howto properly setup these tables can really help get some extra torque at cold temps. But you must dial back ignition as IATs and/or CHTs rise.

[jsveb]

Wouldn't the pressure sensor also be able to pick up detonation? Not only would we be able to see an abnormal rise in pressure, but it could also be identified as being "out if timing" when analyzed in relation to cam timing etc. together with regular knock sensing this would further protect the engine as well.

I am not a proponent for complete autotune, I just find this an interesting discussion enabling me to understand a little more based on your feedback - thanks.

[scarceller]

Originally Posted by jsveb

Wouldn't the pressure sensor also be able to pick up detonation? Not only would we be able to see an abnormal rise in pressure, but it could also be identified as being "out if timing" when analyzed in relation to cam timing etc. together with regular knock sensing this would further protect the engine as well.

I am not a proponent for complete autotune, I just find this an interesting discussion enabling me to understand a little more based on your feedback - thanks.

Yea I guess you could use these devices to help find Peak Pressure or Detonation whichever comes first. I do think this type of device would be very useful tool to help tune and help ID detonation. I suspect this type of sensor and it's related processing hardware is expensive. Plus keep in mind that the conditions from cyl to cyl differ and you could have detonation conditions in one cyl and not the others. Some cyl run leaner or richer than others, for example I've found that the RHS (cyl 4-6) in my 3.2L run leaner than the LHS, I suspect it has something to do with the distance intake air needs to travel through the intake to the RHS which has slightly longer travel? The difference is not much but it is noticable by like .1 to .2 AFR.

[jsveb]

FYI

Spark plug combustion pressure sensors for tuning - Engine - GT-R Life

[safe]

This would be cool to transfer to a 911:
Trionic - Wikipedia, the free encyclopedia

[jsveb]

Originally Posted by safe

This would be cool to transfer to a 911:
Trionic - Wikipedia, the free encyclopedia

I think I remember pelican BeepBeep mention this on the turbo forum earlier. Pretty cool.

[burgermeister]

Very neat article - thanks!

For the tuning crowd. why not just install a knock sensor? After all, that's what all the OEMs currently use.

There are a few reasonably priced aftermarket knock sensors available.

[Tippy]

I would love to have a knock sensor and MS II has the capability to use one but it takes an amplier that cost more than my entire ECU to make it work.

[burgermeister]

No affiliation, don't currently have one installed on my car, though I intend to do so and then get a SW chip. Price is not unreasonable, internet search suggest it does what he claims it does.

J&S Electronics SafeGuard Indivdual Cylinder Knock Control