jharvey wrote:The key item I'm looking at is how fast of an on time and how fast of a drop time do we need
hasn't this been beat to death enough ? it's a non-issue.
jharvey wrote:In either case the window for injecting is approximately half a cycle.
so I did interpret this correctly ! - So my question still stands .. why ? if you have a 720* cycle who says you can't inject over that whole cycle (idle or otherwise)?
No reason why you can't inject for the full cycle, including the compression cycle. Things like XTAU will start to become important, ect. If you inject outside the valve window, you can be quite a bit more sloppy in your tolerancing. Your tolerance at the min pulse width is key, and less important as you increase your period.
I guess you are right, it has been talked about a fair bit, I still haven't heard if the 5mS min pulse width is a resonable width. I guess it must be. I based that off fellows web page I once found where he posted his scope curves. It was an interesting page, you could see the injector needle hit the end of travel, and how it affected the inductance of the injector.
5ms is actually quite large. There are people running highly boosted engines with very big injectors who run pulse widths close to 1ms at idle. That's barely more than the opening time and with the accuracy of MS2 some do get reasonable idle behaviour.
Jean
Last edited by jbelanger on Fri Nov 06, 2009 2:19 am, edited 1 time in total.
How big are the injectors? The dead time accuracy is of critical importance with short times like that because the bit flowing fuel is only a small fraction of that time.
jbelanger wrote:There are people running highly boosted engines with very big injectors who run pulse widths close to 1ms at idle.
Great and thanks. So if we half that to .5mS allowing for some room for growth, then work with that 1% tolerance, we drop the rise time to 5uS, and the .1% tolerance drops to .5uS.
The ISL9V5036 and STGB35N35LZ IGBT's are in that range, as well as the VNP5N07 and VNP20N07 MOSFET's, so those chips are still good.
Fred truth be told the car drives maybe 3 or 4 times a year to have fun and upset people, i am a part owner in it and i do not wish to spend anymore time or money in it. I want to get my Subaru running on it..
On the topic of injector control.
Am i correct in thinking the plan is still to only support high Z injectors. say 6 ohms and up.
If this is the case i have to ask. why are we still trashing around with stuff here. the VNP5N07 has been tried proved and then tried and proven again. Why not just settle on this. it would also save a little bit of board rel-estate.. It is also Big enough that you can solder it by hand..
I would not use and IGBT for injector control. If you use a fully avalanche rated FET such as the VNP5N07 you only need that to drive one high Z injector. With the IGBT you would absolutely need a flyback circuit because the breakdown voltage of an IGBT is around 400V. An ignition coil will handle that but I don't know if an injector will. And even with a flyback circuit, I'd rather have a part that will have a breakdown voltage more in line with what the injector will handle.
