组hp8722d的数量是什么意思

组数（GPIB命令是“NUMG”，因为我看到你昨天发布在你的另一个帖子中）是针对正在使用扫描平均功能的情况。
例如，假设您想要读取其数据在8次扫描中平均的测量迹线。
您可以将分析仪置于扫描保持模式，将平均因子设置为8并打开平均值。
然后你可以做Number of Groups = 8，并在8次扫描结束时读取（平均）跟踪。
或者你可以做Number Of Groups =，读取部分平均的轨迹，并重复直到发生了8次扫描。
但我认为你想做5次非平均扫描并获得每次扫描的跟踪数据。
但是，分析仪只能为每条迹线缓冲一次一次扫描的数据（在平均值为N的平均情况下，它只缓冲前面的N次扫描数据的平均值）
对于每个痕迹）。
因此，您需要将“OPC ?; SING”和“OUTPFORM”放在执行5次的循环中，并将每个跟踪读入程序代码中的单独内存缓冲区。



以下为原文

Number Of Groups (GPIB command is 'NUMG' as I see you posted in your other thread yesterday) is geared toward situations where the sweep averaging feature is being used. For example, say you want to read a measurement trace whose data is averaged over 8 sweeps. You would put the analyzer into sweep Hold mode, set the averaging factor to 8 and turn averaging on. Then you could do Number of Groups = 8, and read the (averaged) trace when the 8 sweeps finish. Or you could do Number Of Groups = , read the partially averaged trace, and repeat until the 8 total sweeps have occurred.

But I think you want to do 5 unaveraged sweeps and get the trace data for each one of those sweeps. However, the analyzer only ever buffers just one-sweep-at-a-time worth of data for each trace (in the averaging case with average factor N, it just buffers the average of the preceding up-to-N sweeps worth of data for each trace). So you will want to put your "OPC?;SING" and "OUTPFORM" in a loop that executes 5 times, and read each trace into a separate memory buffer in your program's code.

> {quote：title = bhokkan写道：} {quote}> Number Of Groups（GPIB命令是'NUMG'，因为我看到你昨天发布在你的另一个帖子中）是针对正在使用扫描平均功能的情况。
例如，假设您想要读取其数据在8次扫描中平均的测量迹线。
您可以将分析仪置于扫描保持模式，将平均因子设置为8并打开平均值。
然后你可以做Number of Groups = 8，并在8次扫描结束时读取（平均）跟踪。
或者你可以做Number Of Groups =，读取部分平均的轨迹，并重复直到发生了8次扫描。
我从来没有尝试过像这样阅读平均数据，我发现在8720D的一般使用中很烦人的是平均值是如何工作的。
在我看来，如果你将平均值设置为16，那么在16次扫描之后，跟踪不稳定。
将平均值设置为N不会导致最后N组数据的简单均值。
我想我在某处看到了平均值在8720D中呈指数加权，所以我假设其频率较高的表兄8722D。



以下为原文

> {quote:title=bhokkan wrote:}{quote}
> Number Of Groups (GPIB command is 'NUMG' as I see you posted in your other thread yesterday) is geared toward situations where the sweep averaging feature is being used. For example, say you want to read a measurement trace whose data is averaged over 8 sweeps. You would put the analyzer into sweep Hold mode, set the averaging factor to 8 and turn averaging on. Then you could do Number of Groups = 8, and read the (averaged) trace when the 8 sweeps finish. Or you could do Number Of Groups = , read the partially averaged trace, and repeat until the 8 total sweeps have occurred.

I've never tried reading averaged data like this, something I have found annoying in general use of the 8720D is how the averaging works. It appears to me that if you set the average at say 16, then after 16 sweeps, the trace is not stable. Setting the average to N, does not result in a simple mean being taken of the last N sets of data. I think I read somewhere the averaging is exponentially weighted in the 8720D, and so I assume in its higher frequency cousin, the 8722D.

所有Agilent VNA中的扫描平均值遵循以下规则：New_Data_Displayed =（New_Data / N）+ Old_displayed_data *（N-1）/ N这形成IIR滤波器并且内存效率非常高。
为了显示最后N次测量的平均值，我认为必须有N个内存缓冲区;
这里只需要2个缓冲区。
如果确切地进行N次扫描，则结果与最后N个数据点的平均值相同。



以下为原文

Sweep averaging in all Agilent VNA follows the rule:

New_Data_Displayed=(New_Data/N)+Old_displayed_data*(N-1)/N

This forms an IIR filter and is very memory efficient. To show the results a the average of the last N measurements, I think one must N memory buffers; here only 2 buffers are needed. If one does exactly N sweeps, then the result is the the same as the average of the last N data points.