DCV采样期间的3458A DMM定时误差

大家好！我在DCV采样模式下使用3458A和选项001测量来自多达8000个样本的窦信号的属性。
我观察到，DMM有时会使采样延迟100 ns。
问题是：100 ns延迟的原因是什么，以及如何防止它？在探讨这个问题时，我将测量设置简化为以下组合，以省略其他来源的可能干扰。
（在原始设置中，DMM添加了自制的小电路，为其他设备的同步化提供光时钟信号）。
所以他是简单的测试设置：1）DMM输入连接到33220A输出。
2）DMM通过HPIB-USB适配器编程用于TRIG SGL 3）HPIB断开连接。
4）通过ext.trigger触发DMM 5）exttrigger电缆断开6）当采样结束时，连接HPIB并将样本从DMM读出到计算机。我在另一个房间的测量实验室外重复这个实验
，在另一层，与不同的DMM。
问题仍然存在.100 ns延迟的问题有时会发生，有时很少。
到目前为止，我已尝试并发现以下事项： - 问题不是由于生成信号的某些延迟。
（我已经使用来自不同发生器的DMM信号进行采样，而且我已经与PXI高速数字转换器并行采样相同的信号。生成的信号非常清晰） - 这是偶然的，在某些光圈时间组合下出现问题，其他
其他组合的一天或一周 - 其他用户也观察到了这个问题，但只有数百或数千个样本才能观察到。
这个问题的想法是，DMM在将数据保存到内存期间存在时序问题。
现在，我有时也会在前几百个样本中看到问题。有时看起来，在相同的样本数字处发生100ns延迟，有时这种延迟会沿着样本“浮动”并改变采样信号中的位置.-当我放置一些EMI滤波器时
在3458A电源插头之前，某些DCV模式孔径时间配置的问题消失了，但新出现在不同配置。
当我在DMM的触发输入上改变电容（例如，我使用更长的电缆进行触发输入）时，有时会发生同样的情况.-这个问题不仅发生在一个DMM上，我在3个DMM上看到了这个问题，两个相当
新的，一个人年纪大了.-上面提到的细节说我，电网有一些问题，但我不确定，因为我在实验室可以使用更多的电网，改变它们没有效果。对于长文本
并提前感谢您的回复



     以下为原文

  Hi all!

I am using 3458A with option 001 in DCV sampling mode to measure properties of sinus signals from hundereds up to 8000 samples. I have observed, that DMM makes sometimes delay 100 ns in sampling.

The question is: What can be the reason for the 100 ns delay and how to prevent from it?


During exploring this problem, I simplified measurement setup up to the following combination, to ommit possible interferences from other sources. (In original setup, DMM has added home made small circuit to provide optical clock signal for synchornization of other devices). So he is simple test setup:
  1) DMM input is connected to the 33220A output.
  2) DMM is programmed via HPIB-USB adaptor for TRIG SGL
  3) HPIB is disconnected.
  4) DMM is triggered via ext.trigger
  5) ext.trigger cable is disconnected
  6) when sampling is finished, HPIB is connected and samples read out from the DMM to the computer.
I repeated this experiment outside of measurement laboratory in the other room, on the other floor, with different DMM. The problem still exists.

The problem with 100 ns delay occurs sometimes less, sometimes very often. Up to now, I have tried and found following things:

- the problem is not due to some delays in generated signal. (I have sampled with DMM signals from different generators , moreover I have sampled same signal in parallel with PXI high   speed digitizer. Generated signal is really clear)
- it is ussual, that the problem occurs one day at some combinations of aperture time, other day or week on other combinations
- this problem was observed by other users too, but only after hundreds or thousands of samples. The idea for this problem was, that  DMM has problem with timing during saving data to memory. Now, I see the problem sometimes during first hundred of samples too.
- sometimes it seems, that 100ns delay occurs at same sample nummer, sometimes this delay "floats" along samples and change position in sampled signal.
- when I place some EMI filter before the 3458A power plug, the problem dissapears for some DCV mode aperture time configurations, but newly appears at different configurations. Same happens sometimes, when I change the capacitance (e.g. I use longer cable for trigger input) on the trigger input of the DMM.
- this problem occurs not only with one DMM, I have seen this problem on 3 DMM, two of the quite new, one was older.
- details mentioned above says me, that there is some problem in the power net but I am not sure, because I have access to more power nets in the laboratory and changing them has no effect.

Sorry for long text  and thanks in advance for response