测试移相器DUT显示出高插入损耗

好像很奇怪。
有一些非常不寻常的事情可能导致这种情况，但我真的不相信它们中的任何一个。
有一点是，可能会有一些错误信号随着相移而变化（例如源和负载不匹配），因此校正正在做很多工作来纠正它，但不能完美地纠正它
因此它有一些未经纠正的“剩余”金额。
当您旋转移相器（比喻）时，您会发现移相器响应和残余误差排列正好相反的位置，并且误差校正算法会减去两者以产生明显的损失。
另一件事是可能有一些硬残余（如支线）正在做同样的事情。
以下是您可以尝试查看效果来自哪里的几件事。
从未校准的设置开始（确保取消对校准集的取消，或者只是从预设开始，在迹线上将其关闭不会在其他迹线上将其关闭）。
设置您的测量并完成各个阶段，看看是否在未修正的响应中出现任何异常。
如果它确实表现出异常，那么就像第二个问题（一些虚假信号取消你的信号），我们可以进一步调查。
如果它没有显示相同的奇怪行为，那么我们接下来会看错误纠正。
从简单的规范化开始（data-> mem和data / mem）。
再检查一遍。
我想这里也不会显而易见。
最后检查一个完整的2端口校准。
如果它出现在这里，那么我们可以追踪它;
一些问题：你的移相器有增益吗？
多少？
什么是你的FW转速，以及PNA上的选项。
“第二个PNA”也是同一型号吗？



     以下为原文

  seems strange.  There are a few very unusual things that might cause this, but I can't really believe any of them.

One thing is that there might be some error signal that changes with the phase shift (such as source and load mismatch) that is pretty big, and so the correction is doing a lot of work to correct it, but doesn't correct it perfectly and so it has some "residual" amount that is not corrected for. When you spin your phase shifter (figuratively) you find a position where the phase shifter response and the residual error line up exactly opposite and the error correction algorithm subtracts the two to give an apparent loss.

The other thing is there could be some hard residual (like a spur) that is doing the same thing. 

Here's a couple of things you can try to see where the effect is coming from. Start with an uncalibrated setup (make sure do UNSELECT the calset, or just start from preset, turning it off on a trace won't turn it off onthe other traces).  Setup your measurements and run through the phases and see if you get anything unusual on the uncorrected response.  If it does show something unusual, then it is like the second issue (some spurious signal canceling with your signal) and we can investigate further.  If it does not show the same strange behavior, then we look next at the error correciton.

Start with a simple normalization (data->mem and data/mem).  Check again. 

I guess it will not be apparent here either.

Finally check with a full 2 port cal.  If it occurs here then we can track it down; some questions:
does your phase shifter have gain? How much?
What is your FW rev, and options on the PNA.

Is the "second PNA" also the same model number?