Geoffrey

One additional question, is there a table for injector end angle or injection timing in the 964 ECU?

DougB

Maybe just using different terms but...

The Motronic Idle, part throttle and WOT fuel maps show fuel trim, not base pulsewidth or target lambda. The fuel trim can be expressed as a percentage or a factor that kinda looks like lambda:

+10% or 1.10
-10% or .90

Some of the Motronic editor software I've looked at simply shows you the raw 8 bit values (128 = 0 trim)

For the idle MAP the values show you how much trim was applied to the base pulse-width to bring the mixture to lambda=1. I'm pretty sure the low and mid areas of the part-throttle map also are shooting for lambda=1.

I'm also interested in the injection angle thing too.

-doug
-doug

Geoffrey

Can someone show a 3D fuel table?

Colin 90 C2

I know this is a little off topic, but It's kinda related.

For all you Motronic gurus:

Is it possible to take the program off a 993 chip and translate it onto a 964 chip?


Thanks,
Colin

Geoffrey

In the US, the 993s use a different ECU than the 964s so no

Red rooster

Colin,
The 993 DME in all its variants uses a different hardware platform to the 964 family of DMEs so code is not directly transferable.
What are you trying to acheive ?

Geoffrey,
I have been meaning to look for the " end of injection " . It was one of the factors that may have relevance in the Motec discussion.

All the best

Geoff

Geoffrey

The other item I'm curious about would be a 2d dwell table since the stock coil requires 3.2ms of dwell to fully charge and that means that at RPMs over 6000 the coil is not being fully charged.

Lorenfb

"One additional question, is there a table for injector end angle or injection timing in the 964 ECU"

It's a constant, and always occurs at the same relative timing point for each
cylinder when running in sequential mode (Hall sensor connected).
There's basically little value in changing the injection point, i.e. a variable one, versus
an optimum point determined by sequential injection.

I haven't yet seen data supporting a variable injection point timing as a function of an input,
e.g. loads, RPMs, as being beneficial. It's not in this version of Bosch's DME, though. It's easy
to verify for those interested by just using a scope triggered by the Hall sensor and viewing each
injector pulse as the RPM/load changes.

"The other item I'm curious about would be a 2d dwell table since the stock coil requires 3.2ms of dwell to fully charge and that means that at RPMs over 6000 the coil is not being fully charged"

It's all just basically a function of the coil inductance
Energy = L X I X I / 2
Where I is the peak coil current, L is the inductance

V (battery voltage) = L X dI/dT or

Time (for peak coil current) = L X I / V

For L = 1mH I = 8 amps V = 13 volts
Energy = 32 mjoules (min spark energy), Time = 1 X 8 / 13 = .62 ms

For 2mh coil, E = 64 mjoules, Time = 1.24 ms (@ 6000 RPMs - a spark every 3.3 ms), i.e. enough dwell time

Check here for typical Bosch coil parameters (964 is a TCI ignition):

www.systemsc.com/diagnostic.htm

Lorenfb

Dwell angle?????????????????
The dwell graph is the generic one shown in the Bosch Automotive Manual.
Why complicate things?

1. Battery Voltage = constant (for all practical purposes) = 13 - 14 volts
2. Current (max) is defined by the coils series resistance.
3. The 964 ignition module is just a switch which turns on
shortly (1.0 - 1.5 ms) after the last spark to start the current
ramp for the next spark.
4. The coil current reaches a peak value equal to V (battery) / R (coil)
5. The coil current is sustained at the peak value until the next spark.

OR

The ignition module is ALWAYS switched on Tms (dwell) = I X L / V before the next spark,
i.e very SIMPLE. Thus, the SAME spark energy is ALWAYS developed.

There's no need for ANY complicated dwell calculations, e.g. RPM or voltage.

Geoffrey

Loren,

I'm not sure it is as simple as that. In practice you can put a current probe on the coil wire which will show the coil charge. This is how I identify what the proper coil dwell time is for an inductive ignition when using MoTeC. This allows me to ensure the coil is fully charged, but not over charged which is damaging. The 964 Coil has a dwell time of about 3.2ms meaning that it requires the ignitior (ignition module) to be turned on for 3.2ms for the coil to be fully charged. Voltage variations change the time slightly, but as you pointed out, it should be 13-14 volts in a normally functioning system. At the higher RPMs, there is not enough time to charge and fire the coil with the coil having reach full saturation. This occurs at any RPM over 6000. 60000 / 6000 = 10ms and there are 3 spark events in 10ms with the coil requiring 3.2ms meaning there is just barely enough time at 6000rpm to charge and fire the coil. Above that there is less time available and it won't fully charge the coil.

Laurence Gibbs

lets see if i have this correct when the AFM outputs a voltage via it's variable resistor this voltage is used to look up a load factor in the afm map? This in tern with the other "modifiers" then alters the fuel trim for the required map (idle, partial and wot) to set injector pulse width? Presumably the load factors are created by the characteristics of the AFM itself? So to move to MAF you would need to know the relationship between the load factors and the MAF??

Red rooster

Jason,
Just read your stuff. Boy , you have been busy !
Just so I understand, you seem to be saying that fuel injector duration is set primarily by the part load map and then that map is modified/multiplied as required by the idle /full load maps ?
That is wrong , so I have obviously misunderstood !

All the best,

Geoff

Lorenfb

"I'm not sure it is as simple as that. In practice you can put a current probe on the coil wire which will show the coil charge."
- Geoffrey -

But it is, that simple!!!!!!!!!!

No guessing allowed. Read and understand the equations I posted. As I said, it's all a function
of the coil being used. Then get a scope and view the coil waveforms. Then measure
the coil resistance and inductance. Then do some calculations using the equations posted.

So, what about the engines that revv to over 15,000 RPMs, you're implying
they don't have enough spark energy? Come on! Given the right coil design,
ideal spark energy can be developed in a millisecond, 18,000 RPMs (6 cyl)!

SimonExtreme

I am always scared of commenting in the presence of people who know so much more than I would ever know but......................................

Geoffrey
I can accept your figures but how important is it for the coil to be fully charged. Is the amount that it isn't charged by enough to cause problems. From the research I have done, I feel it's not an issue in a properly working system as it will produce enough charge.

Loren
Correct me if I am wrong, but I believe that the 964 uses an inductive discharge system but high reving engines use a different type (capacitive discharge?). Everything I have ever read suggests that you start to have issues with inductive discharge at about 6000rpm and you need a different system over about 8000rpm.

Lorenfb

"Correct me if I am wrong, but I believe that the 964 uses an inductive discharge system but high reving engines use a different type (capacitive discharge?). Everything I have ever read suggests that you start to have issues with inductive discharge at about 6000rpm and you need a different system over about 8000rpm"

That's right TCI - transistor coil ignition (inductive discharge) is used on the 964.

As I said, it's ALL in how the ignition system is designed, e.g. coil selection on the TCI type.

The CDI system has a charging time issue, i.e. the capacitor must be charged just like the
coil, which is more problematic than the TCI type.

Read here for more info on ignition systems: www.systemsc.com/technical.htm