LOW PASS FILTER BOARD The words most frequently heard from homebrewer's are" I hate winding toroids!" Well friend if you want to homebrew, then you will have to learn! One Winding through a toroid IS One Turn! I have found one of the best sources for low pass filters data (aside from simulating one in LT SPICE) is the work of W3NQN who published an article in QST on LPF's. His design specifically addresses the 2nd harmonic. Below is an excerpt from that article that shows the component values. The bonus is that where possible he has used standard capacitors which make life a lot easier. Following the table, I will demonstrate how to build a working 40 Meter LPF. The article can be found on the Internet: QST February 1999. So the data for 40 Meters indicates we will need to build three coils, two of which are the same. L2,L6 are 1.34 Uhy and L4 is 1.22 Uhy. Since we are running about 20 watts through the LPF it is a good idea to use cores sizes that will handle that power level and also use cores that will work in that frequency range. The RED cores ( 2) are well suited for use at 7 MHz and the core size needs to be a minimum of the T50 size (about 1/2 inch). But if you want to be safe in the event you will run 50 Watts, then the T68 is a better choice. The core size may also dictate getting closer to the design value in terms of the number of turns. The key parameter in the number of turns is the Al value provide by Amidon. The Al value for the T502 = 49 and the Al value for the T682 = 57. There is a simple formula to calculate the number of turns. Start first by taking the Inductance value and dividing that by the Al value. Next you take the square root of that result. Finally multiply that by 100 and you have the number of turns. Using the T502 core here is the step by step for coils L2,L6: 1.34/49 = 0.0273. Now the square root = 0.1653. Now X 100 = 16.54 Turns. Repeating with the T682 core: 1.34/57 = 0.0235 and the square root = 0.153 and X 100 = 15.3 Turns. So now lets look at coil L4 with the T682: 1.22/57 = 0.0214 and the square root = 0.1462 x100 would give 14.6 turns. For the T502: 1.22/49 = 0.0249 and the square root = 0.15779 x100 = 15.8 turns
As stated earlier the coil selection is based on power level and what gets you closest to an integer number of turns. In this case, I would use the T502 core and round up the L2,L6 to 17 Turns and the L4 to 16 Turns. Rounding the T682 would literally give you the same coil for all three. Now lets repeat that exercise using the T506 cores where the Al value is 40. Repeating our same calculations would yield L2,L6 = 18.3 Turns and L4 = 17.46 Turns. Rounding would give 18 turns and 17 Turns. In practice you will find that L4 is usually one turn less. I chose the T506 cores  if only for a practical reason it was midnight and those were the only cores I had in the parts bin at the time!
40 Meter Toroid Winding Data A calculation for T506 cores L2,L6 and L4 for 20 Meters would yield L2,L6 = 12.9 turns and L4 = 12.32 turns. Rounding these would give L2,L6 = 13 turns and L4 = 12 turns. Like I said One turn Less! Thus both LPF's used the T506 cores and use something around #22 wire to carry the current.
20 Meter Toroid Winding Data

