Technical IAW ECU Live Mapping & Tech Info

[madnez]

Nope, I made one myself If you have figured out the map addresses making up the xdf file is a piece of cake

Yeah if i had it would be Any way you would share your xdf ? I need to change WOT fuelling, fuel enrichment and rev limiter.

 

[woj]

Yeah if i had it would be Any way you would share your xdf ? I need to change WOT fuelling, fuel enrichment and rev limiter.

It is not my intention, not for quite a while at least. Reason being, I do not want to find tons of "mapped" chips on ebay the following day.

But if you are after these simple things, you can try getting your hands on something called ECM (a tuning program) that should have some basic maps pre-defined for this ECU, or at least 18F. Also I am not sure what do you mean by fuel enrichment here, coolant dependant?

 

[madnez]

It is not my intention, not for quite a while at least. Reason being, I do not want to find tons of "mapped" chips on ebay the following day.

But if you are after these simple things, you can try getting your hands on something called ECM (a tuning program) that should have some basic maps pre-defined for this ECU, or at least 18F. Also I am not sure what do you mean by fuel enrichment here, coolant dependant?

I understand and yeah coolant temp fuel enrichment.

 

[craigb244]

Woj,

Im in need of some information that I hope you might have.

the punto 75 ecu runs batch injection as you mentioned.
do you happen to know when and how many times injection happens during each engine rotation.

or maybe you can tell me if the timing is the same as the spi. if so I will be able to find the information myself by painting the timing wheel and using a strobe connected to the injector.

 

[woj]

Woj,

Im in need of some information that I hope you might have.

the punto 75 ecu runs batch injection as you mentioned.
do you happen to know when and how many times injection happens during each engine rotation.

or maybe you can tell me if the timing is the same as the spi. if so I will be able to find the information myself by painting the timing wheel and using a strobe connected to the injector.

Really funny you ask that, we just had a very extensive discussion about that on the PL forum with lots of speculations.

Bottom line is:

The batch-injection in 8F/18F ECU happens once per crankshaft rotation.

When exactly it is difficult to tell, but a well educated guess is at TDC. At least that's what the Punto manual says for the half-sequential 18FD ECU. Here (half sequential) I guess every pair goes on every crankshaft rotation, so all-in-all there are two injections per crank rotation. But this part is yet to be confirmed.

In the new MPI Seis and similar engines with full sequential injection I think it is clear...

 

[craigb244]

Really funny you ask that, we just had a very extensive discussion about that on the PL forum with lots of speculations.

Bottom line is:

The batch-injection in 8F/18F ECU happens once per crankshaft rotation.

When exactly it is difficult to tell, but a well educated guess is at TDC. At least that's what the Punto manual says for the half-sequential 18FD ECU. Here (half sequential) I guess every pair goes on every crankshaft rotation, so all-in-all there are two injections per crank rotation. But this part is yet to be confirmed.

In the new MPI Seis and similar engines with full sequential injection I think it is clear...

thanks,

The reason I ask is because ive had idle mixture problems with my car.
my injectors are about twice the size as standard (not really the problem)
I used to run 2 banks (back one is 1+4, bank two is 2+4)
each back fired twice per engine cycle (720degrees)

the problem was the injector pulse width is so low. so the injectors where outputting more of a spit of fuel then an actual stream. With this set-up even the 75 ejectors caused idle problems.


So the other day I thought... punto 75 was batch injection... if fiat could make it run smooth and pass emissions then so can I.

changed the settings to batch injection with 2 injections per engine cycle.
this also ran poor because the pulse width is the same.

so I changed it to one injection per cycle. pulse widths went from 1.1ms to 1.6ms engine idle instantly went smooth and afr went from a VERY jumpy 11 to 13 afr to a perfect and steady 14.7



I was going to mess with the injection timing to see if it can get even better.



Im guessing the standard cento ecu runs the same timing as the 75 because people say they wire punto 75 injectors onto it and it runs fine.

one way to test the injector is to connect a led to the first coil pack and to the injector wire. if the injector led flashes at half the speed of the coil then the batch fire will be once per engine cycle(720degrees)

might even be able to work out the angle by using the led

 

[woj]

The reason I ask is because ive had idle mixture problems with my car.
my injectors are about twice the size as standard (not really the problem)
I used to run 2 banks (back one is 1+4, bank two is 2+4)
each back fired twice per engine cycle (720degrees)

So exactly like in the 18FD ECU for 1.2 16v-s.

the problem was the injector pulse width is so low. so the injectors where outputting more of a spit of fuel then an actual stream. With this set-up even the 75 ejectors caused idle problems.

This is strange, are you sure it was not due to something else, for example injection angle?

So the other day I thought... punto 75 was batch injection... if fiat could make it run smooth and pass emissions then so can I.

changed the settings to batch injection with 2 injections per engine cycle.
this also ran poor because the pulse width is the same.

so I changed it to one injection per cycle. pulse widths went from 1.1ms to 1.6ms engine idle instantly went smooth and afr went from a VERY jumpy 11 to 13 afr to a perfect and steady 14.7



I was going to mess with the injection timing to see if it can get even better.



Im guessing the standard cento ecu runs the same timing as the 75 because people say they wire punto 75 injectors onto it and it runs fine.

one way to test the injector is to connect a led to the first coil pack and to the injector wire. if the injector led flashes at half the speed of the coil then the batch fire will be once per engine cycle(720degrees)

might even be able to work out the angle by using the led

This is still even stranger and more confusing for me. One knowledgeable person on the PL forum said that according to his study of the ECU he is convinced that the batch firing happens once per engine cycle, which would be in line with your findings. Since there was a general disbelief about that one other guy actually recorded the engine noise with his noisy injectors on idle and measured the timing with a pro audio program, conclusion is once per crankshaft rotation. But now I am thinking that he may as well have been measuring the sparks noise peaks instead of injections.

As for the rest. No, Cinq timings are not the same as Punto 75 timings. But I guess they are close enough for the autocalibration to do the rest if indeed it works for some people as you report.

Then, I have 250ccm injectors (almost twice the stock size), original batch firing system of the 18F ECU. My idle injection timings are around 1.2ms IIRC. This suggests to me (because of your results) again that the injection could be indeed once per engine cycle, not crankshaft rotation.

Finally, I was getting the impression from studying the 18F program, that the once/twice per engine cycle could be varied depending on the engine state/load/RPMs. But I never dived into that.

Hence my total confusion now. If you are going to do any more experimentation, please post the results

 

[craigb244]

the injector angle is just standard... as are the injector spray pattern.

the problem is the fact our injectors take 1ms to fully open.
I tested this by running injector test on megasquirt and setting it to inject for 1ms.
and looking at the injection pattern showed more of a random spray then an proper squirt (which is aimed at the valve head) some this spray goes onto the walls of the ports which is what causes the problems.

now you should know the ecu is programmed with the injector opening time. this is a time where the ecu fires the injector but assumes that no fuel is injected.

so an injection time of 1.1ms is an full injection time of 1millionth of a second


another thing to remember is...
fuel should NEVER be injected on a closing intake valve.
running batch does not leave a large window for injection (its doable hence why i would like to know the timing).
by injecting twice per crank rotation you are effecivly making this window smaller and also effecting 2 pistons rather then 1.

My theory would be that it is set to inject all the fuel for a couple of degrees after TDC.
obviously without a cam sensor the ecu wont know what piston 1 or 4 are doing but it wont matter... it will just inject when it sees TDC on the crank wheel then wait for 720 degrees before the next injection.

doing it this way would mean that it will be injecting all the fuel on closed valves.. 3 pistons will be on their exhaust/compression/power strokes and the other (be it piston 1 or 4) will be on its intake but on a opening or opened valve.



When you say the cento spi timing is not the same as the 75 are talking about injection timing? Either way im guessing that the injections per engine cycle will be the same... so ill see how many squirts the a standard cento ecu makes and report back



oh and about switching between 1 and 2 pulses depending on rpm/load... its not at all needed. once the engine is over a couple of thousand rpm then it wont matter if you used 1 squirt or 10 squirts.

really you want to run the least amount of pulses you can. this is because the amount and quality of the fuel during opening is now reliable.

in a perfect world the injectors would be either on or off with no opening time.

 

[woj]

I am only a bit sceptic about this 1ms of opening time (but I trust all you wrote . For two reasons. One is that idle injection times for my SPI engine were 0.8ms (according to the ECU log, but then it is a different injector). Second one is that I have twice as big injectors as stock ones, what I did I basically globally scaled all injection times by 0.5 (0.59 to be exact, because this is how bigger the injectors really are) and my all fuel from idle to high load (apart from boost of course) scaled beatifully. If the timings I scaled contained 1ms of opening time my fueling would be horribly off somewhere, either idle or high load. My suspicion is that the opening time could be added elsewhere, but then some other parts of the fuelling procedure do not make sense...

Something for me to think about over the weekend

 

[craigb244]

on megasquirt the 1ms opening time is typed (along with battery corrections... low voltage means longer opening times) in the injector settings. this 1ms can then be forgotten about as it is added to the calculations the ecu makes.

the spi injector is a low impedance injector which can be much quicker at opening. also it is large and the pattern at which it sprays does not matter as the fuel gets mixed very well as its sucked past the butterfly

.. maybe your software does not show the opening time added to the pulse width.
so that 0.8ms might be 1.8ms?

 

[woj]

In 18FD (16V) ECU that I am in progress of studying, it is almost exactly like you say it is in MS. That is, a value dependent on the battery voltage (representing injector opening time) is added to injection after calculations. So this makes sense. And from what I can remember this opening time is not included in the time reported by the ECU through diagnostics (have to check again), so this would again be what you said. However, in 18F (8V) the whole battery voltage timing scaling is done quite differently (by factor multiplication rather than adding) and the scaled value is reported by the ECU. What remains to be checked is whether this is the final injection value, or perhaps something is still added to it after it is reported through diagnostics and before it is actually used to open inijectors (this part of the program is a bit hairy hence my uncertainty in all this that we started discussing here).

 

[Asteris]

oh and about switching between 1 and 2 pulses depending on rpm/load... its not at all needed. once the engine is over a couple of thousand rpm then it wont matter if you used 1 squirt or 10 squirts.

really you want to run the least amount of pulses you can. this is because the amount and quality of the fuel during opening is now reliable.

Craig,

If you set Enhanced Accel Enrichment at ON with lag compensation, MS uses extra injection pulses to feed the motor and keep the desired AFR while sudden acceleration.

http://www.msextra.com/doc/ms2extra/MS2-Extra_Tuning_Manual.html#EAE

 

[craigb244]

Thanks. have you got EAE set up? Tried it before but got confused lol

 

[Asteris]

Not yet, just managed to set standard AE, so the car doesn't stumble. I make the same mistake and jump to the newest firwmare as soon as it is out, and i forget to save the tunes...

I hope i'll manage to stay at 3.1.0 for a while and tune everything right.

I saw your setup and you don't use the IAC idle. Why is that? It works perfect for me...

 

[craigb244]

Sorry for hijacking we wojs thread.

I had idle control. but it was while it was still been developed so never worked right.
I blocked the valve off and have since wired the outputs for other things.

Never missed it. Idle is fine

 

[Munkul]

craig.... with all that talk about opening times you've just helped me realise why ive been maxxing out 240cc injectors before they really should be! Mine is running with 2 squirts per revolution, im gonna re-do it with 1 squirt and see how it goes...

Not sure why I didnt realise this myself before now, but cheers

 

[craigb244]

glad I could help
idle should improve as well if its anything like mine, you might have to tweak the req-fuel or scale the fuel map(s) to get the afr right after changing it as there will be a small change in the fueling by changing the injection style

 

[ReproBoy]

Did anybody compare data from original 27C512 with modded via hex editor? I did, I haven´t find any differences .. I really would like to upgrade my Lancia Y 1,2 16V for more fuel and limiter moved higher, but with 18F ECU I can´t reach it. Does anybody have here functional 64kbyte BIN file ? I have many files for 75MPI, but for 85 nothing

 

[woj]

Did anybody compare data from original 27C512 with modded via hex editor? I did, I haven´t find any differences .. I really would like to upgrade my Lancia Y 1,2 16V for more fuel and limiter moved higher, but with 18F ECU I can´t reach it. Does anybody have here functional 64kbyte BIN file ? I have many files for 75MPI, but for 85 nothing

What modded chip image do you talk about? If you bought something that said modded, compared to the original and there are no differences, well then, you got ripped off Have seen this before.

What do you mean by reaching something (what?) with 18F ECU?

Re BIN files, did you at least read the first point in this thread?

 

[Karolis]

Oh god, I found good stuff to read right now (by discovering this topic)