sqrt[(a-1)/N * (F-1)]

where a is the number of conditions (a=3 for my case), N the total sample size (N=30 for my case for instance) and F the F-statistic value (F=0.495 for my case). Assuming that I have identified a minimum total sample size required for F=0.495 at a power of 0.8 (refer to attached image), what I unsure of, is that do you still input the original N of 30 into this formula when calculating the f adjusted value, or do you use the value of 42.5 (from the attached image)?

Any help is greatly appreciated!