On a starting note. My dish has an f/D of 0.43 and 7.6m in diameter. And after two years and two winters of changing feeds from 23cm to 70cm, and back again, an effort was made to combine both a 70cm & 23cm. The first attempt was using the very popular W21MU feed, but all efforts to correctly, and fully illuminate the 0.43 f/D dish using this feed was a complete failure. Cold sky to sun-noise at a solar flux of approx. 70 to 71 produced a consistent degradation of nearly 2.5db below the VE4MA scalar ring feed using polarizer stubs. All measurements and angles on the IMU feed were checked and triple checked, but nothing would regain the losses. This fact was mentioned several times during the 70cm and above EME net and it was suggested that Russ, N7ART be contacted for a possible solution. Russ very willingly mailed the article presented here, and that first phone call eventually led to many more phone calls until everything was working as indicated in the article. Not being an RF design engineer I picked Russ's wealth of knowledge until I ran out of questions. The end result, thanks to Russ, is that I am now very pleased with the results I am now getting. My cold sky to sun-noise is where I expect it be, and both 70cm & 23cm are operational at the flick of a switch. The first installation of this diagonal dipole feed used RG-142 in place of the semi-rigid teflon both RX & TX balun lines in conjunction with the 5/32 OD tubing. This combination still produced a sun-noise measurement approx. 1.2db below the original VE4MA installation. After some deliberation and calculations it was discovered the RG-142, although only 1/2 electrical wavelength long induced a loss of at least 0.10db. This loss represented a decrease in S/N ratio of 1. 1 db ....... nearly exactly what I was missing. Along came the construction of an air-line for the RX port. All dimension regarding length above dipole balun mount, dipole lengths, dipole placements, exacting alignment and placements, were retained from the original article written by Russ. The dipoles themselves MUST be constructed and assembled with exacting measurements, as would be expected, to obtain near perfect balance between dipole pairs. This assembly uses an air dielectric on the RX dipole and a square dipole balun mount. The square balun mount, at least for me, allowed for a more exact adjustment of each dipole half relative to the waveguide back plate. Each 1/4" dipole half should be constructed to be near identical as possible. The assembly of the waveguide, dipoles, back-plate, etc, etc, MUST be very exacting and true, wether the angles are right angles, diagonal, or otherwise. The dipoles must be at exact right angles to the baluns, dipole pairs must be at exact right angles to each other and the dipoles must be placed truly diagonally, comer to comer within the wave guide and centered through the 3/4" opening in the waveguide back plate. For this installation 18 gauge galvanized material used with a 1/4" overlap for the waveguide and 3/32 pop-rivets installed from inside waveguide, spaced approx. 1 " apart. The overlap is located halfway up one side of waveguide. 10/32 polarizer machine screw nuts are soldered to the waveguide using just enough heat to melt the solder and not warp the material. A 10/32 Tap was used to secure 10/32 brass nut in place until solder had cooled. Adjustment of dipoles relative to back-plate is made by slight adjustment of the dipole balun mounting plate using either, or each, of the four comer mounting bolts. When completed, return losses should be approaching -30db. And the dipole pair should, and must. produce, very, very equal output levels. The complete alignment is very well covered in the original publication by Russ, so little will be said here. Except to say that a pair of very well matched right-hand and left-hand circular polarized helical antennas should be used, and I recommend constructing a balanced dipole such as those used in the feed. RF shielding relating to unwanted reflection of RF energy should also be applied. To this concern and necessity I gathered all the heavy blankets, quilts, etc I could find, and using a staple gun, stapled them completely around the area in the recreation room being use as a test range, this worked better than originally expected. The RG-142 is still being used at the TX port and unless power at the TX feed port exceeds 1KW under CW conditions I see no need to change it. It did directly handle 850 watts under test with no heating or adverse effects. The construction of this waveguide on first glance may seem rather complex, but really is no more so than anyother feed. The results have been more than gratifying, and for a dual band feed, or for that matter a single band 23cm feed, on a dish with an f/D of 0.47 or deeper, as well as small dishes where blockage may be a concern I would strongly recommend giving this waveguide serious consideration. One final comment, the phase point appears to be 1.25" to 1.5" inside the waveguide. The 27.5" diameter 70cm back-plane is located the same as on the IMU feed. 2.8" back from front edge of feed. On a starting note. My dish has an f/D of 0.43 and 7.6m in diameter. And after two years and two winters of changing feeds from 23cm to 70cm, and back again, an effort was made to combine both a 70cm & 23cm. The first attempt was using the very popular W21MU feed, but all efforts to correctly, and fully illuminate the 0.43 f/D dish using this feed was a complete failure. Cold sky to sun-noise at a solar flux of approx. 70 to 71 produced a consistent degradation of nearly 2.5db below the VE4MA scalar ring feed using polarizer stubs. All measurements and angles on the IMU feed were checked and triple checked, but nothing would regain the losses. This fact was mentioned several times during the 70cm and above EME net and it was suggested that Russ, N7ART be contacted for a possible solution. Russ very willingly mailed the article presented here, and that first phone call eventually led to many more phone calls until everything was working as indicated in the article. Not being an RF design engineer I picked Russ's wealth of knowledge until I ran out of question. The end result, thanks to Russ, is that I am now very pleased with the results I am now getting. My cold sky to sun-noise is where I expect it be, and both 70cm & 23cm are operational at the flick of a switch. The first installation of this diagonal dipole feed used RG-142 in place of the semi-rigid 14lon both RX & TX balun lines in conjunction with the 5/32 OD tubing. This combination still produced a sun-noise measurement approx. 1.2db below the original VE4MA installation. After some deliberation and calculations it was discovered the RG-142, although only 1/2 electrical wavelength long induced a loss of at least 0.10db. This loss represented a decrease in S/N ratio of 1. 1 db ....... nearly exactly what I was missing. Along came the construction of an air-line for the RX port. All dimension regarding length above dipole balun mount, dipole lengths, dipole placements, exacting alignment and placements, were retained from the original article written by Russ.