与45dB衰减器测试相比为什么VNA漂移很多？

谁能帮我？
我们的型号是E5247A。
我想测试我的VNA随时间的稳定性。
VNA已经升温，至少3小时。
以下是我的setup-parameter。
1. 15dBm源功率2.1.5KHz分辨率BW 3.Span 60MHz，中心2.06GHz 4.6401扫描点5.Full 2端口Cal我做了什么：连接两个VNA端口和Bullet Adapter。
并存储S21（数据）到内存，显示内存数据。
开始时，内存和数据之间的差异大约为-80dB。
2小时后，差异增加到-50或-55dB（甚至-45dB），这意味着VNA漂移很多。
但我重复了相同的过程，但将两个VNA端口与45dB衰减器连接在一起。
并存储S21（数据）到内存，显示内存数据。
最初，内存和数据之间的差异大约是-110dB。
2小时后，差异增加到-90dB左右。
我在想，这是正式的吗？
与45dB衰减器测试相比，为什么VNA漂移很多？
提前致谢。

以上来自于谷歌翻译

   以下为原文

  Can any one help me?
Our model is E5247A. I want to test the stability of my VNA over time.  The VNA already wARMs up, at least 3hours. The following is my setup-parameter.
1. 15dBm source power
2.1.5KHz resolution BW
3.Span 60MHz, center at 2.06GHz
4.6401 sweep points
5.Full 2 port Cal

What i did :
connect two VNA port together with Bullet Adapter. And storage S21 (the data) to memory, show the memory-data. At the beginning, the difference between memory and data is around -80dB. 2hour later,  the difference increases to -50 or -55dB(even -45dB), meaning VNA drift a lot.

But I repeated the same process, but connect two VNA port together with 45dB attenuator. And storage S21 (the data) to memory, show the memory-data. At the beginning, the difference between memory and data is around -110dB. 2hour later,  the difference increases to around -90dB.

I was wondering, is this formal? Why VNA drift a lot with Bullet Adapter compared with 45dB attenuator test?

Thanks in advance.

孔德羲

2018-10-18 17:05:15

您使用了错误的数学来评估S21的稳定性。
S21稳定性应使用S21数据/存储器。
S11和S22应该使用数据存储器。
此外，将IF BW更改为1 kHz以进行测量。
1 kHz及更低将使分析仪处于更精确的模式。
S21的稳定性应为0.02至0.05dB。
您的测量对电缆的相变非常敏感，远远超过VNA的稳定性。
您的电缆很可能比VNA稳定10倍。

以上来自于谷歌翻译

以下为原文

You have used the wrong math to evaluate S21 stability. S21 stability should use S21 Data/Memory. S11 and S22 should use Data-Memory. Also, change the IF BW to 1kHz for your measurements. 1 kHz and lower will put the analyzer in a more accurate mode. S21 stability should be on the order of 0.02 to 0.05 dB.

Your measurement is very sensitive to phase change of the cable, much more so than VNA stability. Very likely your cable is 10 times less stable than the VNA.