I just thought I'd throw this out there, in case it might help someone else.
NOTE: It worked for me. Your mileage may vary.
When I built my QDX r2, I measured out the stated 3 lengths of 20 cm of 0.33mm wire for the trifilar toroid T2. At that stage in the construction, after cutting wire lengths as given in the instructions, there was just barely enough 0.33mm wire left to make each length 20 cm (310 cm provided - 251 cm used prior to the point of winding T2 = 59 cm left for T2). After twisting the wire together using Hans' described technique (which worked well), I ended up with a twisted length of wire that was only long enough for about 7 turns through the toroid. Fortunately, I was able to find some similar wire in my junk drawer from a junked transformer or something; so I was able to try again, immediately. For my second attempt, I think I measured each of the 3 lengths at 35 cm. This time, the twisted wire was more than long enough to give me the required 10 turns.
When I built my QDX r3, I remembered that after measuring out the stated length of enameled wire and winding each of the toroids for the QDX r2, I consistently had quite a bit of excess (scrap) wire left. So, when I wound the QDX r3 toroids, I subtracted 4 cm from each of the lengths given in the instructions. Even then, I still ended up with about 4-5 cm of scrap wire each time. The result of this was that by the time I got to the trifilar toroid T2, I still had enough wire left to cut each of the 3 lengths to 30 cm, rather than the stated 20 cm, which was more than enough to get the 10 turns through the toroid.
So, in my QDX building experience, I was able to use about 4 cm less 0.33mm wire for each toroid, which left me with enough remaining wire to cut the 3 lengths for T2 at 30 cm each (rather than 20 cm) -- which resulted in a trifilar wire long enough for 10 turns through T2.
Again, it worked for me. YMMV.
