"Partial" water changes only exist to help lower the possible shock of a sudden change to fish. Thus, in a fishless cycle, partial really has no purpose. Instead, you want to get the most effect on the water chemistry for your work, so water changes during fishless cycling might as well be large.
Now just -how- large might depend on convenience. In my case, I have an external cannister filter which has an intake tube that reaches down near the substrate to take in water to be filtered. During a water change I can just leave the filter running, as long as I don't allow the water surface to drop below the filter intake. That's very convenient for me so that's how I determine when to stop removing water. If I had to turn off my filter I would just change water all the way down to the substrate, until the siphon broke on its own.
The water chemistry reason you'd like to maximize your water change is that the cycling process has created enough nitric acid to use up most of your buffering agents (buffering capacity can be determined by a KH test.) Once your buffers have all been used up neutralizing this acid, you get a rather sudden drop in pH (lower pH = more acidic water.. ie. more hydrogen protons floating around!) What the water change is doing is bringing in some fresh buffering ions (usually carbonates of calcium, magnesium etc.) to start grabbing those protons. Since your particular tap water only tests out at pH 7.4-7.8, you will want -more- of it going in, to help give you more buffer. Also, taking out water means taking out nitrate and nitric acid, so a large water change is a win-win for cycling.
As Si has suggested, your water has demonstrated (without you even needing to measure KH) that it is low enough in carbonate hardness (KH, the buffer measure) that you would benefit from supplementing it. Simple kitchen baking soda (pure sodium bicarbonate) (found in Alka Seltzer and other stomach antacids for obvious reasons) is super powerful and fast at buffering the pH of your water. If you dose at 1 teaspoon per 50L you will get a KH rise without getting any pH rise, if you dose at 3 teaspoons (= one tablespoon) per 50L then you'll get a big KH rise and a whole lot of pH rise. So I suggest you take the middle road and dose 2 teaspoons of baking soda per 50L (you'll have to do the math to figure out your amount to use) and this should help to raise your pH toward the 8.0 to 8.4 optimum that the autotrophic bacteria like. The sodium part of the sodium bicarbonate will also stay in the tank but we expect that to come out with the big water change at the end of fishless cycling.
Remember, baking soda treatment to raise your pH is only a good practice when there are not fish in the tank, because its so powerful and fast at making changes. Fish are very sensitive to mineral content (hardness) changes. In fact, its these that they are really sensitive to, not pH. pH is only a secondary indicator of mineral content changes, which is the real thing we are concerned about for fish. When fishless cycling is over, your pH of 7.4 to 7.8 baseline tap water is going to be fine for running your tank with fish in it. The fish won't be forcing the biofilter to produce nearly as much nitrate (and nitric acid) as is forced by you squirting in 5ppm of ammonia ever day! So the buffers and pH of the running tank with fish will be much more stable and less prone to pH drops.
~~waterdrop~~
