-Sorry for bringing up an old thread for my first post but......I felt like I should chime in, just in case the op is still working on this.
-First off put the hairdryer down. Yes it is dangerous and yes using a hairdryer to evaporate saturated solvent will cause oxidization. Once this happens you will need zinc to recover the freebase. You can minimize loss by pouring a small amount of warm solvent over the oxidized dmt and proceed to re-freeze precip. This will pick up any freebase left mixed in with the goo because dmt-oxide is not soluble in naphtha/heptane. You can smoke the gooey oxidized junk. Just spread onto a favorite herb or onto steel wool/screen if using a 'machine' type device.
-Ok now for your methods. If you prefer a straight to base extraction that will work just fine. Your processes are good but your ratios (mhrb:naoh:water:solvent) are definitely negatively affecting the yield. Finding a good ratio that works best for you is personal preference and will take some time; but I would highly recommend starting with a ratio of 1:1:15 (mhrb:naoh:h20) . So for example let's say you want to process 250g of root bark. You would mix 250g shredded/powdered mhrb, 250g lye and 3750mL distilled water. I guarantee this change alone will almost double your yield. i.e.1%-1.3% (lower end % with stb, and higher % with a/b methods) yields are common and expected. I will mention that if you are using pre-powdered bark you may want to experiment with more h20 in the mix. Generally around 20% more water for pre-powdered mimosa.
-Alright the next item that will not only help increase yields, but also save your hard earned $$ is changing your solvent for pulls. Heptane is great...for recrystallizing; it's not so great for picking up freebase dmt. VM&P naphtha is much cheaper than heptane, it's easy to find at your local hardware store and it is superior to heptane as the solvent of choice for 'pulls'. Trust me on this one; you will use less solvent and pull more spice with naphtha than with heptane.
-To eliminate the problem of evaporating your solution before freeze precipitating and creating oxidized dmt by mistake you just need to use less solvent. For a one pound extraction using a 1 gallon jug, you will need no more than 40-50ml of naphtha per pull. On smaller extractions such as the 250g example above, I would recommend using 10-15ml of room temperature solvent per pull (don't use heat until the last pull or two as this will cause discoloration in final product). And I believe it was already stated to gently roll the naphtha within your container instead of shaking or using force; doing so will produce light yellow to white dmt freebase after freeze precipitating almost every time.
-Whew...that was long haha. I hope all this doesn't come off as arrogant, I'll be the first to say I have lots to learn about the process yet but I hope I answered some questions you had and I'd be happy to help further and/or post a detailed straight to base step-by-step extraction method that I've had success with in the past (1% yield avg). Take your time and find a tek you like then study it a little before you try it. Once you start to understand the process/reasoning behind each step you will find yourself ending up with some really nice spice.
-P.S.
Since it's my first post and I have no credibility here yet, here's are a few pics of an older stb extraction using a slightly modified Noman's tek to help validate some of my claims.
-Materials were 100g pre-powdered mhrb, 100g naoh, 1000ml distilled h20, naphtha for pulls and heptane for recrystallizing. Yielded 1.3g total spice white to light yellow in color... 3/4g of that was produced within 4 hours of beginning the process.
-Crystals forming through freeze-precip w/naphtha.
2012-08-08_14-00-15_901 by
nmk8743, on Flickr
-.75g white powder.
IMG_20120912_104257 by
nmk8743, on Flickr
-.60g light yellow powder.
IMG_20120809_133714 by
nmk8743, on Flickr
-First recrystallization of yellow powder.
IMG_20120827_165120 by
nmk8743, on Flickr
-First recrystallization of white powder.
2012-09-12_10-41-36_905 by
nmk8743, on Flickr
-Experimenting with submerged thread during recrystallization.
2012-09-11_10-40-09_525 by
nmk8743, on Flickr