• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 390 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

AMC3302-Q1: -Voltage fluctuations on the grounded magnetic bead

qixi Zhai
Prodigy 10 points
Part Number: AMC3302-Q1
Other Parts Discussed in Thread: AMC3302

Tool/software:

Hello, when we tested the 3302 module, we found that the jitter on the measurement side will cause a high-frequency voltage jitter of 3MHz ±7.8V about 8us at both ends of the GNDI bead(FB3):
1. How might this jitter affect the output?
2. Is it possible to cause the output to be full biased?

  • 0
    TI__Guru 56172 points

    Hi, 

    +/-7.8V seems quite high, do you have a screenshot from the oscilloscope you can share showing this? 

    I would expect the majority of this to be attenuated as the bandwidth of the isolated amplifier is only 340kHz. 

  • 0 in reply to Alexander Smith
    Prodigy 10 points

    Hi，Smith：

      I am happy to recive your answer.The oscilloscope screenshot seems like this:

    After asking questions on the E2E forum, we continued to test at the same time, and found that when the bus voltage fluctuates, high-frequency jitter also occurs on the beads, and the amplitude is not low, and the current through RSHUNT（only 0.1mR） is actually not large, but the isolation amplifier has a negative almost fully biased output, which is also strange.

    We look forward to hearing from you.

    Best wishes

  • 0 in reply to qixi Zhai
    TI__Guru 56172 points

    Hi, 

    So this screenshot shows about 10us/div. 

    What is happening on Vbus at this time that such a significant +200V, then -500V swing occurs within a matter of 10us?

    I also assume that the high voltage probe is bandwidth limited, so perhaps this swing is much quicker that what is captured. 

    Is this for some sort of transient or shut-down event on the DC bus? 

    How close is the AMC3302 to this power stage and is there any reference to the DC bus voltage? 

    Strange that we see such a voltage variation, but very little variation in the current? 

    I'm thinking is it possible that a current transient across the shunt's inductance could cause a high frequency over-shoot to occur.  

    Can you please share the layout of the AMC3302? 

  • 0 in reply to Alexander Smith
    Prodigy 10 points

    Hello, sorry for not going into detail before:
    1. I    n the case of no load, we created a short-circuit event on the BUS at the input stage, and the oscilloscope screenshot shows the dynamic response of the BUS on the output side, and the semiconductor switch in parallel with the BUS on the input side was quickly turned off at this time, and the module circuit where the AMC3302 chip being measured was not turned off.
    2. The normal voltage of the input stage BUS was 500V at that time.
    3. Due to the multi-channel setup, the main current of the short circuit does not theoretically pass through the channel in which the AMC3302 chip is located.

  • +1 in reply to qixi Zhai
    TI__Guru 56172 points

    Hi, 

    No worries, thank you for the additional detail. I think I understand better now. So the current from this event does not pass through the shunt, however the AMC3302 is referred to this high side voltage that sees the variation. 

    Where is this current measurement occurring in the system? Is this for DC/DC converter?  

    Two things I think may be happening and a couple debug steps to try. 

    1. Pin 8 and Pin 2 are not hard tied. There is inductance between the two pins due to the vias down to the lower plane that connects them. During this voltage fluctuation, Pin 8 and Pin 2 may see this change in voltage at different times. For correct operation, these two pins should see equal voltage at all times. Try to short Pin 2 and Pin 8 with an additional wire, directly soldered to the device pins. Re-test and see if the issue is resolved. If this is the case, please see the layout recommendation in this application note: https://www.ti.com/lit/an/sbaa515a/sbaa515a.pdf

    2. The second possibility would be input common-mode voltage violation due to the inductance of the shunt and necessary compensation from the input RC filter. I think this is less likely, but as a debug step if #1 does not fix the issue, you could try shorting INP = INN = HGND. If this fixes the issue, then we would need to understand the inductance of the shunt and layout, which creates a zero to occur at a particular frequency. This could be compensated by adding a pole from the differential RC filter at the same frequency. This is discussed in this user's guide: https://www.ti.com/lit/ug/sbau473a/sbau473a.pdf

  • 0 in reply to Alexander Smith
    Prodigy 10 points

    Hi,

    According to the test, actually is the second. Thanks for your reply. 

  • 0 in reply to qixi Zhai
    TI__Guru 56172 points

    Hi, 

    Great to hear that this was able to help resolve the issue!