|
Correct. As I have it drawn now, there are two relays handling each signal in parallel, each driven by it's own transistor. Everything is redundant there. Next, make all relays a NO type, but have one ECU set be NC, so that even with no power to said relay board, one of the computer's signals still gets through. Is it truly 100% redundant? Probably wouldn't be considered as such...it does have a failsafe mode though.
Regarding the board itself, just make sure it has more than a screw in each corner holding it down so that it doesn't vibe itself to pieces. Again, not hard.
Connectors are a potential place of concern certainly, but I'm going to use MIL-DTL-38999 locking connectors like those from Amphenol and whatnot. If they're good enough for the Space Shuttle, they're good enough for me.
On Fri, Feb 18, 2011 at 9:41 AM, Ernest Christley <echristley@nc.rr.com> wrote:
Dustin Lobner wrote:
Sounds good. I'm planning on a pair or maybe even 3 MS3 units with a relay board going between them. Basically, said board would distribute any and all sensor signals so that all MS3s would see all the signals, but the injectors and coils would be run by only one MS3. I don't yet know exactly what I'm going to do here, but the MS3 (with the expansion board) can run 8 injectors. I'm thinking 2 or 3 per rotor, so that if any one injector goes down, it keeps running OK. The MS3 does the whole injector staging thing well, so that might work.
The goal is to have independent systems. You're description takes two or three computers that have been used in a variety of situations, and thus are fairly well shaken down. Then you're going to route them all through a one off relay board. The systems aren't really independent if they all rely on a single communication path or resource.
|
|