spray arc disaster.

running a lincoln power mig 350. the wire is 0.045 Mil spec 70S2 hardwire. the gun is a magnum 400. shielding gas is 98%argon with 2%Oxygen at 42cfh. I'm trying to run a testing coupon of 3/8" thickness with a 30 degree bevel (60 degree total for both sides) with a 1/8" root gap. using 1/4" backing plate. I have the machine on setting "21" for 0.045 argon mix welding. the procedure per the PQR says that I should be using 3/4" plate with the same bevel and a 3/16" root gap running 27 volts and 300 amps. Cannot seem to get it to work what so ever. I've already tried 4 tests and each one has linear cracks down both sides of the root which carry thru the entire weld sequence. I've tried preheating the plates and postheating the plates. I've tried a ton of different amperages and wire speeds to no avail. it seems like the spray just wants to suck into the sides instead of getting down into the root on my initial pass.

Does anybody know exactly what settings or parameters I should be running for these conditions? I've always run fluxcored wire at my old job and this spray arc hardwire mig is really pissing me off. also, any idea why i keep getting cracks/porosity? it's carbon steel so I cannot understand why this is so difficult especially running in the flat position! I've run fluxcore 0.045 vertical up tests that were over an inch thick and a foot long with absolutely no problems and they passed RT, PT, and VT testing so I can weld believe it or not this is just a totally new procedure for me. any advice or suggestions are much appreciated because my boss is about to kick me out the door even tho he's told me to use 3 different shielding gases already so he has no clue.

Yahoo/answers doesn't really have a category for this so if anybody knows where I can find the info that'd be great too.

-------------------------------------------------
my gut tells me you are experiencing center line cracking from excess heat input and fast travel speed creating an alignment of solidification grain boundaries down the center line of each bead.

i would recommend using the bottom end voltage settings to achieve spray to decrease heat input..and cheat toward the high side on wire speed to prevent the cracking.