While surfing the library database (what most of you ought to be doing instead of cruising those alt.binary groups) I came upon a tasty little process of considerable academic interest.
One of Dan Haney's little pets is the Mitsunobu Reaction, and finally I got around to looking it up. It deals with the conversion of alcohols to azides, which can then be conveniently reduced to amines with a variety of reagents (NaBH4 seems to be the most popular, but Al-Hg would surely do the trick).
This is fine for producing heavy mixed aryl/alkyl systems, like MDA, but extremely hazardous for producing, say, methylamine, as the intermediate Methyl Azide is exceedindly unstable (ie - it tends to explode) and to make matters worse, alkyl azides generally boil at ridiculously high temperatures.
Personally, I'd rather make HCN without a bodysuit, but then again...
Some inventive and entrepeneurial chaps at the University of Newcastle-upon-Tyne in the UK came up with a one-pot process that goes straight from an alcohol to the amine hydochloride salt. You can't get more direct than that for some of those more interesting compounds. Unfortunately, this is not a viable way (as far as I can tell) to produce MDMA, or any other methylaminated compound, as it does not tack a methyl group on with it. The advantage in using this to prepare MDA is that it will not produce any di- or tri- aminated product, as an Al-Hg reduction (or any other similar mode of preparation) will do. As well, there is no need to supply ammonia either in salt or aqueous form, though would be the least of worries to me in the matter.
The overview follows, with a reference to chase at the end.
Cognate Preparation, MDA from MDP-2-Pol 
1 Mole of MDP-2-Pol(1) (17.9g) is dissolved in 50mL THF in a 1L RB flask with magnetic stirring. 120mL of 1M Hydrazoic Acid (2) in Benzene (caution: highly toxic) is added, followed by 22.2g Di-Isopropyl Azido-carboxylate in 50mL THF.
Next, a solution of 57.7g Triphenylphosphine in 300mL THF is added *carefully and slowly*. This will raise the reaction temperature, potentially considerably. Maintain temperature around 30-40C. After the temperature has returned to room temperature for 1 hour, heat the reaction mixture to 50C for 3 hours using a temp. regulated heating mantle. Add 10mL of water at this point, then continue heating for an additional 3 hours.
Strip the solvents off under vacuum and suspend the residue (liquid freebase + impurities) in 100mL of methylene chloride. Shake this in a separatory funnel with 100mL of 2M HCl (caution!) to convert the amine freebase to the HCl salt. Draw off the methylene chloride layer (should be on bottom) and wash the acidic aqueous layer further with 2 x 50mL of methylene chloride.
Pour the aqueous layer out onto glass trays heated by steam (improvise) to evaporate off the water, leaving the crystalline hydrochloride salt behind. The yield, I would estimate, should be around 65%, or .065 moles.
*DISCLAIMER* Remember, this is for academic purposes only.