[Twenty95]

Ok apparently the problem would be solved, Gribouilly tried by removing the Vr sensor and the Vr conditioner and by putting a hall sensor, he succeeded in having the two crank and cam signals, the trigger angle at 246 °. So there is a big problem with the vr conditioners, I myself tested two models with the simulator and could not work.

I ordered a vr sensor so I will be able to test it on the car on my side tomorrow, hoping that it will solve the ignition advance problem.

[jevgienij]

I have a 1.2 8V Cento at the moment but the trigger wheel and general setup is the same. I built two different conditioners. One on MAX9926 and one on LM258 op-amp. Both are working great. The car starts and runs both on -114 and 246 trigger angle. In the end I'm not even gonna use the MAX9926 version. I'm building everything on a THT board so I prefer LM258 better. It's cheap, available everywhere and it just works...

I also made a plug&play board to fit the Seicento MPI ECU case: https://www.youtube.com/watch?v=PXGRcTwU5JE

[Twenty95]

I tried with a Speeduino v0.4.3 with a V0.4.4, a dsc vr conditioner, a dual vr, chinese arduino, "real" arduino, maybe we have a combination of thing that causes a bug, I don't know, but apparently it works with a hall sensor and without the vr conditioner, it's logical, I don't know if Gribouilly tested the DRV8825 but it didn't work either.

[JanKovalski]

DRV8825 works fine with the 4 wire stepper motor.
https://youtu.be/Ew5TzRGZy90
I have a potentiometer connected as the coolant sensor here.

[Twenty95]

Maybe but on none of the cars it works so far, these projects are a curse.

[PSIG]

There are so many odd results, that it is difficult to know if there is a primary issue (wiring, assembly, settings, etc) that is responsible for some or all of them. If faced with diagnosing this vehicle, I would back-up and begin basic testing of the Mega, the PCB, verification of settings (ALL of them), then wiring, etc. Two sets of eyes are better than one.

Not to confuse or add to the pile, but have you verified some basic stuff like which coil pins are for what? It is a passive ("dumb") coil, apparently with power, ground, and shield connections. Is the wiring correct? If not, the coils may not function correctly, and you may be sending large noise pulses through the vehicle, disrupting other functions. In a related question, have you tested for running EMR noise with an AM radio or o-scope?

Another possible issue is your ignition coil driver wiring, which may need the following changes to the Speeduino and added resistors to each driver channel. Again, these are not necessarily suggestions for primary causes, but possibly additional or contributing issues.

David

One example layout for firing quad waste-spark from one TC4424 driver channel. Ignore the cylinder assignments, as this was for a 16-spark-plug engine, but the concepts are the same. As the small resistors are to suppress harmonic ringing (rapid on/off oscillation) from each 211 input to others, they should be independent as-shown, rather that one resistor to feed all of them. R1-R4 can be 10 to 18 ohm 1/4W by the datasheets. Check the coil driver is on a good heat sink, and staying reasonably cool. :

[Chris Wolfson]

What bothers me with these problems: Both engines where working with very reliable injection systems. All components can be re-used. So why are there suddenly such problems, that the Fiat never had. I convert old engines to injection and ignition. If you take reliable parts (I prefer Fiat, Alfa, Lancia parts, from Bosch and Magneti Marelli, as they are perfect, cheap and available), with a base trim an engine should start instantly.
It is a little bit arrogant to expect a Speeduino powered, DIYS tuned t-Jet engine to run better than it did with the Motronic Fiat developed. Just my two Cent, after 40 years of engine tuning.

So, if your Arduino-Speeduino hardware works, what you can check on the bench and you use Fiat parts in Fiat locations, all that can go wrong are your wires... and the basic tune.
What you need is some time, a multi meter, a workshop manual of your engine (components) and a computer to check the tune.
From experience I would check the part of the install you hate to do the most, first. Usually this is the region where you went for short cuts and made some mistakes.

Measuring the resistance of wires and looking for logical values is so easy and tells you what may be wrong. Usually open, ground and a value you find in the manual are all options you have. Open has to end somewhere, check it...
Do not jump from one topic to others, do everything once and complete. When all done start with the next thing.

You have to solve such problems systematically, don´t hope for the single, great idea that fixes anything.

[Twenty95]

Unfortunately we did not have the Ecu with the engines.
I did the assembly with a hall sensor, I'm testing this afternoon.

[Twenty95]

It starts with a hall sensor and 246 ° as it should.

[PSIG]

... It is a little bit arrogant to expect a Speeduino powered, DIYS tuned t-Jet engine to run better than it did with the Motronic Fiat developed. Just my two Cent, after 40 years of engine tuning.

I'd say that depends on the context of "better". As the original hardware and tune may be for different purposes (emissions, regulations, emphasis on economy or whatever), a tunable ECM can provide that performance the original did not. It's not arrogant, it's what we do.