EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Molenaar on February 18, 2016, 09:41:04 pm
-
The old ignition of my Mercedes W108 250S from '67 is not in a really good condition any more, so I want to replace it with something better that allows timing adjustment. And why spend some $$$ if you can spend at least 10 times as much (in time) by building it yourself, especially since I'm not an EE by training? :D
As a follow up on this topic https://www.eevblog.com/forum/beginners/what-makes-an-automotive-mosfet-'automotive'/ (https://www.eevblog.com/forum/beginners/what-makes-an-automotive-mosfet-'automotive'/), instead of using a high voltage MOSFET I switched to an IGBT (Fairchild ISL9V5045S3ST_F085) for use with AEM 30-2852 coil. I don't think I have to include an IGBT driver, but I'm not completely sure. In the sketch I omitted the caps/inductors for the SMPS's.
Would the basic coil driver circuit work? Any basic optimisations? I was thinking of maybe adding a cap in parallel with the IGBT to ground, to create an LC oscillator with the low side of the ignition coil.
Would the ionizing current sense circuit work? The idea is to measure the current through the zener, isolate it using the IL300 (does anyone know an automotive rated linear optoisolator?) and use it to adaptively adjust ignition timing.
Any feedback would be very much welcome!
-
The idea is that this would go under the hood, so I would prefer automotive grade parts to withstand excess vibrations etc.
For normal flyback I would agree that a MOSFET would be preferable. For ignition, this should not make a difference, as long as you use non-air cored coils.
Worst case scenario (175 deg C) regarding the ISL9V5045S3ST_F085, it conducts 35 amps with 3.7 V at the gate. The coil I want to use has a 0.5 +- 10% primary resistance, so 25 - 28 A would be enough. Since 3.3 Vg is not defined in the IGBT datasheet, maybe I will add a driver nonetheless, to avoid ending up just short on current.
The ionizing current sense is meant to detect the time of ignition by applying 100 - 400 V DC on the spark plug. By measuring the current through it, the time of actual ignition and maximum pressure can be derived to optimally tune the engine. I think the circuit I came up with should work for that, but I'm not sure if I didn't overlook something.
-
"Automotive grade" doesn't mean a part is physically any different to a non-"Automotive" part. It's much more down to the testing and qualification that's done in the factory... actually testing parameters at 125C rather than inferring them based on a room temperature test, that sort of thing.
Vibration resistance is conferred by how a board and its enclosure are mechanically designed; make sure your board has enough mounting points, use anti-vibration washers on the screws, glue heavy capacitors in place, that sort of thing. Under vibration, components rarely fall apart, but they might come off the PCB entirely.
-
Would the ionizing current sense circuit work?
Did you have a chance to try this circuit? I would love to add ion sense into my setup.
-
Not yet. Due to corrosion problems this ECU project has been put on hold, but I will continue in the near future.
If you go ahead and integrate ion sensing, please report back!
-
You certainly don't want D4 in the circuit, it will prevent any kind of spark occurring (unless it happens to go open circuit when it inevitably burns out).
-
D4 was meant as an extra protection for the switch, to clamp the flyback spike. However, in theory D5 should be sufficient for that.
Still, since it is rated at 300v breakdown, it should not prevent sparks, correct?