On Mon, 2 Apr 2007, Sandwich Maker wrote: > " From: Tom Jennings <tomj@xxxxxxx> > " > " The only thing I've read "bad" about Howell and others is from > " people who tried to set them up without enough sensor feedback, > " "open loop", and feedback is the whole POINT of FI. Except for > " those few, I've not heard a bad thing about Howell. > > i disagree about feedback. it's the only practical way to zero in on > perfect mixture under widely and wildly changing conditions, but i've > always thought that to a degree it was a crutch for systems that > didn't have the proper sensors to read the environment going in. I was specifically referring to non-OEMs adding EFI to an engine that's either older, modified, rebuilt, customized, or otherwise not-factory-unit-to-unit consistent. Saab designed a limited feedback FI system for some particular engine coming off an assembly line with probably good consistency unit to unit. That's not what hobbiests are doing. Isn't it funny though, that a carb basically has one "sensor", an air-velocity/partially-mass measuring device aka venturi, and with that one crappy old thing meters gas pretty damn well over a surprising range of variables. (OK power valves etc count.) Carbs are lousy at: engine (coolant) temperature compensation, economy/partial throttle cruise vs. power "modes", and altitude. Chokes and vacuum or progressive-mechanical secondaries pick the low-hanging-fruit of the first two, and there's some 80's carbs that attempt altitude but we don't hear about those any more! > this was btw an > analog system, based afaik on bendix patents that may have been on the > '57 rebel system. It would be very interesting to re-visit those older analog systems with the electronic gunk we have now! Too bad the hardware is all museum quality and availability now. Analog servo stuff is something I'm good at (useless skill). _______________________________________________ Amc-list mailing list Amc-list@xxxxxxxxxxxx http://www.amc-list.com/mailman/listinfo/amc-list