well, it seems I wrote you a small book, but i think i covered all the bases.
-> is the piece grounded?
this is the DUH bullet. if you don't have a good ground (clamp on the spur, or clam on a metal table or vice that the spur is clamped to or clamped in) then you're going to have a hard time starting an arc.
->what welding mode are you using?
AC for aluminum, DCEN (direct current electrode negative) for everything else.
DCEP is something that you will never use.
also, you said you have a stick/tig machine. make sure the HF (high frequency) start is on (could be labeled TIG mode vs stick mode). if you're on scratch start(stick) that might explain some of the difficulty.
->what kind of helmet are you using?
some folks don't like autodarkening helmets. thats cool, but they help a LOT starting out. it takes some practice being able to see well enough through the dark lens to make sure you didnt move when you flipped the helmet down. save yourself the hassle and get an autodark that goes up to at least a 12 shade. you'll want the darker shades for welding on this thicker material.
->what's your arc length like?
thats the distance between the tip of the tungsten and the workpiece.
you want to be pretty close. the rule of thumb is keep it less than the thickness of the electrode. so 3/32" in this case. if you move too far away, the arc will wander and you'll be putting a lot of heat into the piece all over the place. that may be what's causing your brown dust. that leads me to believe that you're getting some tungsten deposit, meaning you're losing shielding, dipping the tungsten in the weld, or otherwise just getting the tungsten too hot.
I've found there isn't really a too close until you stick the tungsten in the puddle. not that close
-> how are you operating the foot pedal?
if youre waiting to get on the amps until after a good arc has started, I've found that pushing the pedal down to half amps or better will establish a nice arc much faster than pussyfooting around. after you get your arc started, you can back off if need be to get your bearings, ect. but make sure you start with some balls.
-> what cup are you using?
I'm assuming you're using a #7 with a standard collet body, which is usually what comes on the torch that comes with a new machine. if you can get the torch into the joint nice and close with your current cup, it will work just dandy. if you're having to extend the electrode much more than the recommended stickout to get into the joint, you might need to buy a gas lens collet body and a gas lens cup (probably a #7 or 8). the gas lens collet body has a couple of screens that diffuse the gas so it is less turbulent. what that boils down to is it will shield the work and the tungsten farther away from the cup opening for a given cup size than the standard collet body.
for any of these, I'd be looking for about 14-16 cfh 100% argon flow. bigger cup=more flow.
-> have you cleaned the workpiece?
it's good practice to grind/sand all mill scale/rust off and wipe down with a solvent to remove grease, oil, cutting wax/oil, ect.
a stainless wire brush is also good for brushing oxide off between starts. that bluish weld scale can make welds porous if you start on top of it again.
this might be part of your unstable arc problem, its hard for the arc to punch through mill scale/fire scale/black oxide, so it tends to dance around. you haven't said how your spur parts are made (smithed, cast, cold worked, ect.) I'm presuming they're either smithed or cold worked and not cast. cast is a whole different can of worms.
->where is your stickout at?
that's the distance from the end of the tungsten to the rim of the cup. I usually use the cup opening as my rule of thumb for stickout, so #6 cup = 6/16" = 3/8" so I try to keep about 3/8" stickout.
->what kind of tungsten are you using?
I'm really liking the lanthanated 3/32" tungstens. they start easy and they have a nice arc for both AC and DC
they have a gold paint band. ceriated (orange) thoriated(red- mildy radioactive, but the best performance) will both also work well for what you're doing. green (pure) is something you should stay away from at all costs. I've never used one, but jody says they start like crap, and I don't have reason to doubt him.
->what's your tungsten ground like?
you're looking for about a 20-25 degree point with fine longitudinal grinder marks and a slight flat spot ground on the end for the higher amperage you're going to need to weld 1/2" to 3/16"
->what's your amperage set at?
by the 1A/.001" rule of thumb, you should have your machine set to 188A for 3/16". this seems really hot to me, so I'd probably start down around 135-150A. you're welding steel, so if you have more than 100A even on really thick plate, eventually it will melt. after you get a stable arc, if it's taking forever to puddle, up your amperage.
thats all I can think of without seeing pictures or video.
some general tips for welding these dissimilar thicknesses, though:
-keep most of your heat on the thick shank so you don't undercut or blow out the side of the band.
-don't try to fill that whole bevel in one pass. get a good pass on the top side, wire brush, get a good pass on the bottom side, wire brush, then fill it in a little proud so you can file it down so there's no visible weld.(I'm assuming that's what you're going for)
-if you're going to be doing lots of these, make some kind of fixture that gives you good access to your weld joint and secures your pieces well. it will make life a LOT easier.
-if you're having trouble feeding rod fast enough to keep up with the puddle, move up to the next filler size. 1/8" seems a little heavy though, so you should be fine with your 3/32" filler. if this is just plain low alloy mild steel, ER70S2 is your best bet. if it's smithed, I'd be inclined to switch to 70S3 or 70S4. it will flow better with the impurities that can be introduced in smithing. 70S2 will definitely still work though.
hopefully this gets you going.