N9020A振幅测量的奇怪现象

嗨 - 很可能，你看到的是两个MXA之间绝对幅度精度的差异。
在不知道被测信号的标称幅度和频率的情况下，很难说您所看到的是否符合规范。
例如，如果您在50 MHz处测量-20 dBm信号，则N9020A在20至30摄氏度范围内的绝对振幅规格为+/- 0.33 dB。
这意味着测量相同50 MHz信号的两个分析仪理论上可以显示高达0.66 dB的差异，并且都在规格范围内。
但要在这种极端情况下使用两台分析仪的可能性很小。
但是，如果此测量是在5 GHz进行的，则绝对幅度精度规格为+/- 0.33 dB + 1.5 dB，总计+/- 1.83 dB。
测量同一信号的两个MXA可能会显示+/- 3.66 dB的差异。
要在此频率下看到0.5 dB和0.7 dB之间的差异是完全合理的。
问候 -



     以下为原文

  Hi -

Very likely, what you are seeing is the difference in absolute amplitude accuracy between the two MXAs. It is difficult to say whether what you are seeing is within specification, or not, without knowing the nominal amplitude and frequency of the signals being measured. For example, if you are measuring a -20 dBm signal at 50 MHz, the absolute amplitude specification for the N9020A over a 20 to 30 degC range is +/- 0.33 dB. This means that two analyzers measuring this same 50 MHz signal could theoretically show a difference of as much as 0.66 dB and both be within specification. But to have two analyzers at such extremes is highly unlikely.

However, if this measurement were being made at 5 GHz, the absolute amplitude accuracy spec would be +/-0.33 dB + 1.5 dB for a total of +/-1.83 dB. Two MXAs measuring this same signal could show a difference of as much as +/-3.66 dB. To see a difference of between 0.5 dB and 0.7 dB at this frequency is entirely reasonable.

Regards -

谢谢，信号高达6GHz，所以看起来就像是解释。
如果我们使用参考信号来补偿参考信号的频率点需要多少接近频率响应的测试频率低于~0.2 dB（即总误差为~0.5dB）？



     以下为原文

  Thank you, the signals are up to 6GHz, so it looks like that's the explanation. 
If we use a reference signal to compensate how close do the frequency points of the reference signal need to be to the test frequencies for frequency response to be below ~0.2 dB (i.e total error of ~0.5dB)?

嗨 - 没有明确的价值，但一个好的经验法则是测量你的参考信号在测试频率的50 MHz范围内。
在如此窄的频率范围内频率响应的差异通常可以忽略不计。
问候 -



     以下为原文

  Hi -

There is no definite value, but a good rule-of-thumb would be to have your reference signal within 50 MHz of the test frequency to be measured. The difference in frequency response over such a narrow frequency range is usually negligible.

Regards -