使用AgilentN9010ASpectrum分析仪测量IIP3和P1dB压缩点需要什么规格组合器

有些人可以尝试回答



     以下为原文

  some can try to answer it

通常，任何SA都可以测量IM产品，并且可以从测量的IM产品计算IP3。
你需要使用一些组合器（你使用的是什么类型并不重要）但你可能需要在组合器之前在每个源腿中使用相同频率的隔离器，以确保两个源不交叉调制
并生产IM3。
您可以通过将两个源连接到组合器并将输出发送到SA来进行检查。
测量2个音调并将所需驱动电平的IM dBc电平记录到放大器中。
然后连接放大器并测量IM3（以dBc为单位）并输出SA上的电源。
更改SA衰减器值并查看IM3产品，确保衰减器更改时不会改变。
如果是，则表明IM3是由SA中的混频器而不是放大器引起的。
继续添加更多SA衰减器，直到IM3的值不变。
如果它落入噪声，那么你需要降低RBW才能看到它。
（在测试源生成的IM3以查看它是否由SA或源引起时，您还应该进行此参与者测试。
找到放大器IM3后，取最近的大信号输出功率，将输出功率和IM3 dBc的一半加到输出功率上，得到输出截点（OIP3）：例如 - 如果输出功率输出
每个音调为+10 dBm，IM3为60 dBc，则OPI3为+10 + 60/2 = +40 dBm。



     以下为原文

  In general any SA can measure the IM products and IP3 can be computed from the measured IM products.  You would need to use some combiner (doesn't matter too much what kind you use) but you might need to use an isolator at the same frequency in each source leg before the combiner to make sure the two sources don't cross-modulate and produce IM3.  You can check this by connecting the two sources to the combiner and sending the output to the SA.  Measure the 2 tones and record the IM dBc level for the desired drive level into the amplifier.

Then connect your amplifier and meaure the IM3 (in dBc) and output power on the SA. Change the SA attenuator value and look at the IM3 product and make sure it doesn't change when the attenuator changes. If it does it is an indication that the IM3 is caused by the mixer in the SA and not the amplifier.  Keep adding more SA attenuator until the value of the IM3 doesn't change.  If it drops into the noise then you need to lower the RBW to see it.  (You should also do this attenautor test when testing the source generated IM3 to see if it is caused by SA or sources).

After you find the amplifier IM3, take the output power of the nearest large signal and add half the difference of the output power and IM3 dBc to the output power to get the output interecept point (OIP3): example -- if the power out in each tone is +10 dBm, and the IM3 is 60 dBc, then the OPI3 is +10 + 60/2 = +40 dBm.

> {quote：title = nimbargi写道：} {quote}>在我们的机构，我们有EXA信号分析仪安捷伦频谱分析仪N9010A（9 KHz - 3.6 GHz）。
我的问题是，使用Agilent N9010A频谱分析仪，我们能够测量三阶输入截取点（IIP3）和P1 dB压缩点。
错过了第二点：P1 dB压缩：它有点困难，因为源可能不会像设置所说的那样完全改变功率（例如，你从+ 10 ot +11 dBm改变，也许源功率上升1.2 dB或
.8 dB，这被称为光源的“功率线性度”。压缩通常使用VNA测量。但是老式的方法是采用10或20 dB的焊盘（20更好用于匹配
VNA结果）并将其放置在信号源和放大器之间，然后将信号发送到SA进行测量。记录该值，然后将焊盘切换到放大器的输出并再次记录。不同之处在于输入电平的压缩
你正在应用。逐步增加源功率并在输入和输出之间交换衰减器，直到在某个源功率设置下，当你将衰减器移到输出端时，增益会下降1 dB。这就是输入功率
导致1 dB压缩。取下焊盘并测量输出功率t
o在输出端找到P1 dB。
慢，但相当准确。
当您选择压缩方法为“退避”或“x / y方法”时，我们在PNA“增益压缩应用程序”中基本上做同样的事情，这个结果通常非常接近正常方法（偏离线性增益）。



     以下为原文

  > {quote:title=nimbargi wrote:}{quote}
> In our institution, we have EXA Signal Analyzer Agilent Spectrum Analyzer N9010A (9 KHz - 3.6 GHz). My query is, using this Agilent N9010A spectrum analyzer , are we able to measure third order input intercept point (IIP3) and P1 dB compression point.

Missed the second point: P1 dB compression:  it is a little more difficult because the source may not change power exactly as the setting says (for example, you change from+10 ot +11 dBm, maybe the source power goes up 1.2 dB or .8 dB, this is called the "power linearity" of the source.  Compression is usually measured using a VNA.  But an old fashioned way to do this is to take a 10 or 20 dB pad (20 is better to use to match the VNA results) and place it between the source and the amplifier, then send the signal to the SA to measure.  Record the value and then swich the pad to the output of the amplifier and record again.

The difference is the compression at the input level you are applying.  In steps, increase the source power and swap the attenuator between the input and ouput until at some source power setting, the gain goes down by exactly 1 dB when you move the attenuator to the output.  This is the input power that causes 1 dB compression.  Remove the pad and measure the output power to find the P1 dB at the output.

Slow, but reasonably accurate.  We do essentiallyt the same thing in the PNA "Gain Compression Application" when you choose the compression method as "backoff" or "x/y method" and this result is usually quite close to the normal method (deviation from linear gain).