[M_D]

Good connections capable of carry the required current are definitely an essential part of success. Here are some other things I have found out about tumbled parts in 6061, 6063, and 2011 alloy.

Tumbling does something to the surface, which requires extra diligence in cleaning prep. Whether work hardening or surface contamination is the root of the problem, or both, I cannot conclusively say yet. I have seen a lot of inconsistency from commercial anodizers on these same tumbled parts. As a comparison between tumbled parts and fully machined parts, here is how we successfully prepare and the machined parts:

1) Spray rinse with hot tap water to remove all visible traces of water soluble machining coolant. Must be clean enough to pass water sheet test after step 2.
2) Rack.
3) Dip in 150º cleaner for 2-3 minutes max (or else will dull brightly machined surface).
4) Remove quickly and fully submerge in cool distilled water rinse tank to keep heated cleaner from drying on.
5) Rinse in cold tap water.
6) Anodize.
7) Dip in baking soda solution immediately after removing from tank.
8.) Thoroughly spray rinse in cold tap water.
9) Dye.
10) Cold water rinse.
11) Seal
12) Remove quickly and fully submerge in cool tap water rinse tank to keep heated sealer from drying on.

We do not etch the machined finish parts, as it destroys the finished required. The above procedure is between marginal and inadequate for tumbled parts, and most likely older and dirtier parts also, which would have a heavier native oxide than freshly machined parts.

This is how we are currently processing 6061 and 6063 tumbled parts to be dyed black (2011 is another story):

1) Rack.
2) Dip in 150º-165º cleaner for 12-15 minutes.
3) Remove quickly and fully submerge in cool distilled water rinse tank to keep heated cleaner from drying on.
4) Spray rinse.
5) Etch in room temp lye or stripper for 3-8 minutes, until parts brighten. Since the lye or stripper may be of different temperatures or concentration affecting how quickly it works, the main thing is to leave them in long enough to amply clean the surface.
6) Rinse in cold tap water tank.
7) Spray rinse.
8.) May dip in desmut, or may not. Followed by thorough double rinse. We have found that rinsing the desmut is very critical, or it will cause an irregular anodize with a smutty film. If the parts have any smut film from the lye it should be removed.
9) Anodize.
10) Same as above from here.

The main thing is the tumbled parts need to be aggressively cleaned so you are dealing with an even surface of the pure alloy if you are to stand any chance of consistency. Otherwise, you will chase you tail. Having the parts pass the water sheet test is will be no guarantee tumbled parts are clean enough, it only means there is no oil on them.

I am not sure what your tank size is, and your start and end tank temperatures. I suspect with your load size you may be seeing some effects of a tank size that is too small, and thus instable and high temperatures. I have found there is about a 2V drop for every 5-6º rise in tank temperature in the 65º - 80º range, so that could be part of the abnormal voltage curve, but since you don’t detail it this is just a guess. This is off subject since you are using aluminum racking, but I have found that Titanium racks also change the voltages substantially, dropping it from what you would expect from the LCD model.

So, I would look at connections, part cleanliness, excessive tank temperature and temperature rise.