Well on the third pass I nailed it pretty close to perfect (I think). I was doing the recommended method, run what you got, collect a lot of data, massage into a spreadsheet, calculate the errors, apply adjustments to the tables, lather rinse repeat. Being old fashioned I was doing this in row/column tables, not the graphical table editor. When I finally looked at that (damn kids new fangled toys) woah! ugly table! There was all these lumps and peaks... but there was a sense to the rest of it, visually, that looked like all the other tables I'd seen. So I visually (only!) edited the table to shape it minus the stupid lumps. Saved the original of course! Plugged in the mass of crap in the car, loaded the binary image, and drove off. Not bad... BLM was 140, ideal is 128, but that 140 was in every cell... so I actually pulled over, used the table tool to multiply the entire table by 1.10... two hours of driving and 5000 data points later, BLM was 125 to 130 in all slots (faintly leaner in the low rpm/high vacuum corner of the table). Now I gotta turn off closed-loop and finalize the table, but it's running pretty nice. I went back and looked more carefully at the tables in the originla Howell EPROM, and they're not that good. There's funny peaks here and there that to me, now, look like mistakes. _______________________________________________ Amc-list mailing list Amc-list@xxxxxxxxxxxx http://splatter.wps.com/cgi-bin/mailman/listinfo/amc-list