O
Member
【Editorial Discussion】
At millimetre-wave, a waveguide thru that was acceptable a moment ago can move in magnitude or phase after a single remate. That is not automatically a failed calibration or a bad DUT. The flange joint is part of the measurement. Before you recable the whole bench or declare the standard unstable, it is worth writing down which part of the joint actually changed.
1) Alignment comes first. Pins that are bent, missing, or not seated, or a mating that is started off-axis, will shift the aperture and the contact. If the thru only moves after a remate that felt uneven, inspect the pins and the alignment before you touch the instrument. Remate on-axis, without rotating or side-loading the flange body, and see whether the thru returns.
2) The flange face is a second, separate question. A dent, a burr, a scratch, or debris on the contacting surface will change the joint even when the pins look fine. A visual check and a clean of both faces is cheaper than another calibration. If the thru moves only after a flange has been dropped, reworked, or left unprotected, the face is the more likely term. Record which flange you opened and whether both faces were inspected.
3) Joint hardware and repeatability belong in the same note. Uneven or incomplete tightening, a different screw sequence, or a shim or gasket that was not seated the same way will move a thru even when the faces are clean. Use the same hardware, the same sequence, and the intended tightness, then remate once or twice and write down the spread. A thru that only settles after the second or third remate is telling you about the joint, not about the DUT.
4) Do not treat a single remate as proof that the VNA or the standard has drifted. Repeat the same thru after a controlled remate, and compare it with a leave-it-connected stability check over the same time. If the connected thru is quiet and the remated thru walks, the joint is the story. If both walk, look at temperature, the instrument, or the rest of the path.
Open question: the last time a mmWave waveguide thru moved after one remate, what actually closed the gap — alignment pins, the flange face, or the joint hardware and tightening? Please share the band or flange family and how you proved it; no customer names or unverified numbers needed.
At millimetre-wave, a waveguide thru that was acceptable a moment ago can move in magnitude or phase after a single remate. That is not automatically a failed calibration or a bad DUT. The flange joint is part of the measurement. Before you recable the whole bench or declare the standard unstable, it is worth writing down which part of the joint actually changed.
1) Alignment comes first. Pins that are bent, missing, or not seated, or a mating that is started off-axis, will shift the aperture and the contact. If the thru only moves after a remate that felt uneven, inspect the pins and the alignment before you touch the instrument. Remate on-axis, without rotating or side-loading the flange body, and see whether the thru returns.
2) The flange face is a second, separate question. A dent, a burr, a scratch, or debris on the contacting surface will change the joint even when the pins look fine. A visual check and a clean of both faces is cheaper than another calibration. If the thru moves only after a flange has been dropped, reworked, or left unprotected, the face is the more likely term. Record which flange you opened and whether both faces were inspected.
3) Joint hardware and repeatability belong in the same note. Uneven or incomplete tightening, a different screw sequence, or a shim or gasket that was not seated the same way will move a thru even when the faces are clean. Use the same hardware, the same sequence, and the intended tightness, then remate once or twice and write down the spread. A thru that only settles after the second or third remate is telling you about the joint, not about the DUT.
4) Do not treat a single remate as proof that the VNA or the standard has drifted. Repeat the same thru after a controlled remate, and compare it with a leave-it-connected stability check over the same time. If the connected thru is quiet and the remated thru walks, the joint is the story. If both walk, look at temperature, the instrument, or the rest of the path.
Open question: the last time a mmWave waveguide thru moved after one remate, what actually closed the gap — alignment pins, the flange face, or the joint hardware and tightening? Please share the band or flange family and how you proved it; no customer names or unverified numbers needed.
