• 状态 TI Thinks Resolved
  • 已锁定 已锁定
  • 个回复 11 个回复
  • 答案 1 答案
  • 名订阅用户 4 名订阅用户
  • 次查看 2331 次查看
  • 用户 0 名在线用户
选项
选项
相关

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ISO7742: ISO7742 CMTI疑问

user4653817
Expert 1015 points
Part Number: ISO7742

TI专家，你们好：

最近使用使用贵司的ISO7742，对CMTI这个指标存在以下几点疑问，还请帮忙解答下，谢谢。

   

  1、CMTI这个值是根据输出为什么状态下判断是Fail，比如输入为高时，输出是跌到多少V？持续时间多久判断是Fail？

  2、CMTI测试系统图输出CL这个值对测试结果影响很大，规格书测试图是要求15pF，输出管脚对GND+测试板寄生电容+示波器探头寄生电容是如何保证小于此值？

  3、CMTI这个干扰源上升时间、和电压是这么定的，比如 电压100V，时间1ns，或者电压2000V时间20ns,这个斜率都是100kV/us，但是测试结果是不一样的，电压时间这个是怎么考虑呢？另外这个斜率不是理想的上升沿或者下降沿，示波器是怎么取这个可PASS的斜率，是该段的最大斜率还是平均或者？

  • 0
    TI__Guru**** 295050 points
    user4653817 说：
    CMTI这个值是根据输出为什么状态下判断是Fail，比如输入为高时，输出是跌到多少V？持续时间多久判断是Fail？

    您好，由于 数字隔离器输出的是数字信号，所以输出如果由于由于共模瞬变而导致的状态变化被判断为CMTI fail。

    user4653817 说：
    CMTI测试系统图输出CL这个值对测试结果影响很大，规格书测试图是要求15pF，输出管脚对GND+测试板寄生电容+示波器探头寄生电容是如何保证小于此值？

    是的。大多数无源示波器探头具有10pF 至20pF 的电容，这可能会搭建具有串联电阻和其他连接阻抗的 RC 网络，从而在测量期间衰减隔离器原始输出信号的上升沿和下降沿。所以我们尽量使用具有最小探头电容(<=15pF 或有源探头)的示波器探头，并且尽量不要和其他具有阻抗的电路接在一起，基本上才能保证和数据手册值相符。

  • 0 回复 Kailyn Chen
    Expert 1015 points

    1、第一个问题好像是没有答复；如下左图，是可以看的输出是由明显翻转，可以判断是Fail，右下图，输出为常高，但存在短时间跌落，且幅值没到0V，那是否算Fail?

    2、第二个问题也答复只是尽量保证电容小，实际测试如果CMTI沿很快，如果总得CL小，示波器探头自身也容易受干扰，CL大，输出得波形是比较干净的，这个怎么去权衡呢？

    3、第三个问题也请答复下

  • 0 回复 user4653817
    TI__Guru**** 295050 points

    您好，第三个问题我这边会去再确认一下。

  • 0 回复 Kailyn Chen
    Expert 1015 points

    三个问题都没答复呀

  • 0 回复 user4653817
    TI__Guru**** 295050 points

    都会进一步确认。

  • 0 回复 user4653817
    TI__Guru**** 295050 points

    您好，关于您的三个问题，经确认如下：

    第一个问题的两个波形，第一个根据高低转换可以判断为fail。

    但是第二个由于低电平的持续时间低于10ns，不能判断为错误信号，所以可被认为pass。

    1. CMTI pass/fail
      1. The CMTI specs you see in the datasheet is the pass level for the device. levels above this could cause the device to fail.
      2.  for the first waveform your understanding is correct. If the output has changed from LOW to HIGH due to CMTI, then this is a fail.
      3. for the second waveform since the output voltage does not drop to low logic threshold and only lasts for less than 10ns, it is not read as a false low signal. Therefore, this is a pass
    2. The CMTI test need 15pF load. How to balance the small and big Cload? If using the big Cload, the output could be smooth, but if using active probe which could be met the 15pF, but the probe is easy to crosstalk.
      1. The 15pF load that is added in the CMTI test circuit in datasheet is an optional test load. Customer may not need to add this load to their test. It is OK to use any probe for testing. 
    3. How to decide the CMTI interference source rise time, the voltage is so defined, such as voltage 100V, time 1ns, Or voltage 2000V time 20ns, this slope is all 100kV/us, but the test results are different, voltage time this is how to consider? In addition, this slope is not the ideal ramp or ramp, how does the oscilloscope take this PASS-capable slope, is it the maximum slope of the segment or the average? 
      1. We use VCM = 1200V for CMTI testing, as stated in the datasheet. Customer can use any voltage up to 1200V.
        1. The maximum slope of the CMTI generator output should be less than the datasheet specification for the device to pass
  • 0 回复 user4653817
    TI__Guru**** 295050 points

    您好，针对您的三个问题，您还有什么疑问吗？

  • 0 回复 Kailyn Chen
    Expert 1015 points

    第一个问题：跌落10ns不算 FAIL,那多长时间算FAIL？
    第二个问题：贵司测试平台整个系统的CL是多少PF?
    第三个问题：其中内部2个问题都没有答复，1、沿的电压和时间不同是会输出不同测试结果的，斜率不算稳定的，是按最快沿还是平均沿来？

  • 0 回复 user4653817
    TI__Guru**** 295050 points

    好的，您的这三个问题再进一步确认中。

  • 0 回复 Kailyn Chen
    Expert 1015 points

    请问确认怎么样？

  • 0 回复 user4653817
    TI__Guru**** 295050 points

    您好，关于您的几个问题，请参考：

    1. The logic low pulse only lasts less than 10ns and read as pass signal, how long does the low signal last at least to read as fail signal？ 
      1. The 10ns pulse you see is an output from the ISO7742 device. Whether this is a pass or fail is determined by the device that is reading the signal. For example, if the bandwidth of the input channel of the reading device is 10MHz, then it will not be able to recognize a 10ns pulse and therefore it would be a pass. Similarly, if the bandwidth is 200MHz, then it will be able to recognize the 10ns pulse and therefore it would be a fail.
    2. He would like to confirm how much load of our test platform? 
      1. Did you mean to say how much load we have in TI test setup? if yes, then it would be 15pF load.
      2. Did you mean to ask how much load to apply on customer test setup? if yes, then it would be up to the customer on how much load they have in their system.
    3. The slew rate of CMTI interference source is not stable,  such as voltage 100V, time 1ns, Or voltage 2000V time 20ns, this slope is all 100kV/us, but the test results are different, we decided it by average slope or the fastest slope? 
      1. Any transients durations less than 3ns are typically not recognizable and all voltage - time combinations where the time is greater than 3ns can cause the device to fail.
      2. please note that 100kV/us is a typical pass level. Due to manufacturing differences between units, some device can fail at 100kV/us while some others can fail at 120kV/us, while some can also fail at 150kV/us. The actual fail level of the device will be greater than 85kV/us but they can be anywhere above 85kV/us. 
      3. To summarize, not all devices will fail at 100kV/us.