ENA5071C绝对测量结果有些不一致

我需要有关ENA 5071C绝对测量的帮助。
我指的是直接访问绝对R1（n）R2（n）... A（n），B（n）测量值（n是应用刺激的端口）（测量菜单）。
我尝试使用该参数执行一些“特殊”校准，并与仪器的数学功能一起使用，而不是使用正常的S参数。不幸的是，我的算法没有像我预期的那样工作。
经过调查，我发现我认为绝对测量结果有些不一致。
即1）我只使用端口1和端口2（因此R1，A，R2，B是所涉及的测量）2）使用50欧姆电缆（1米）连接端口1和2，并在大约100 MHz（几个）进行测量
指出要轻松验证结果）3）我测量了R1（1），A（1），R2（1）和B（1），并将4测量标准化为第一个（R1（1））的值，这样
... -R1（1）/ R1（1）为1（显而易见） - 正如预期的那样B（1）/ R1（1）的模数接近1（信号通过电缆传播并进入端口2） - 如
预期A（1）/ R1（1）很小（表示反射波返回端口1），这意味着端口2相当匹配（负载匹配） - 因为我没想到R2（1）/ R1（
1）导致模数约为0.4而不是相当小（如A（1）/ R1（1））..如果R2（1）/ R1（1）与从波特2输出的波成比例......我可以
明白这至少可以不小...（这意味着端口不匹配）但在这种情况下我也期望A（1）/ R1（1）同样不小如w
使用其他端口并在所有可能的方向上发生相同的事情“（刺激位置）由于这种不一致现在很清楚为什么我的算法不能正常工作。
是否有人可以解释R2（1）的行为或建议文档和有关此绝对参数的更多信息？
在ENA 5071C手册中，对这些测量只有一个简短的提及，只有很少的解释。还有另一个谜......为什么上述4个参数在扫描到扫描的相位不同？
我可以验证相对相位是否稳定（归一化后相位从扫描到扫描是稳定的），但不清楚为什么它们不稳定它们每个的绝对相位......“可能是一个架构问题”我
告诉自己.....先谢谢大家......关心Aleberto69



     以下为原文

  I need help about Absolute Measurement with ENA 5071C.
I refer to the direct access to the absolute R1(n) R2(n) ... A(n), B(n) measurements ( n being the port where the stimulus is applied)  ( measurement menu).
I tried to perform some " special" calibrations using that parameters and together with the math capability of the instrument, instead of using normal S parameters.. Unfortunately my algorithms didn't work as I would have expected.
After investigations I found some inconsistency about what I suppose that  the absolute measurement  are.
Namely
1) I used just port 1 and port 2 ( so that R1,A,R2,B are the involved measurements)
2) Connect port 1 and 2 with a 50 ohm cable ( 1m) and perform measurement at about 100 MHz ( few points to easily verify the results)
3)I measured  R1(1), A(1),R2(1) and B(1) and  normalize the 4 measurement to the value of the first one ( R1(1))  

so that...
-R1(1)/R1(1) is 1  (obvious)
- as expected B(1)/R1(1) is in modulus close to 1 ( signal propagating through the cable and entering in port 2)
- as expected A(1)/R1(1) is small ( representing the reflected wave coming back into port 1) supposing this means that port 2 is rather matched (load match)
- as I didn't expect R2(1)/R1(1) result in modulus about 0.4 instead to be rather small ( as A(1)/R1(1))..

If R2(1)/R1(1) is proportional to the wave outgoing from the port2 ... I could understand that this could at least be not small...( It means that the port is not well matched) but in this case I also expect A(1)/R1(1) to be similarly not small as well ..

Identical things happens using other ports and in all the possible Direction" ( stimulus position)

With this inconsistency is it now clear why my algorithm is not working properly.

Is there somebody that could explain the behavior of R2(1) or suggest documentation and more information about this absolute parameter?

In the ENA 5071C manual there is just a brief mention to those measurement with very few explanation..


There is also another mystery...

Why the above 4 parameters  are not constant in phase from sweep to sweep?
I could verify that the relative phase is stable ( after normalization the phase is stable from sweep to sweep), but is not clear why it isn't stable the absolute phase of each of them...
"May be an architectural question" I told myself .....

Thank you in advance to everybody...
Regards
Aleberto69