ISO7841: Interference from 2G/3G/4G modem causing data corruption.

Part Number: ISO7841
Other Parts Discussed in Thread: ISO7741, ISO6741, ISO7341C

We are using the ISO7841DWW part to isolate an ADC which is continuously measuring 240V mains voltage. We have diagnosed some data corruption to the ISO7841 during transmission from our 2G/3G/4G modem. We have found that applying an RF burst with a carrier frequency of 800MHz results in the following behaviour.

ISO7841 input without RF

ISO7841 input with RF

ISO7841 output without RF

ISO7841 output with RF

As you can see in the last image, the signal edges are no longer stable and are showing significant jitter. Under these conditions the ADC data is corrupted.

We then replaced this part with a IL717-3E isolator and the issue was removed.

IL717-3E output with RF

See how the edges are stable.

We have tried various solutions (including the solutions given in this similar post e2e.ti.com/.../iso7841-carrier-frequency-and-gsm-immunity ) to remove this issue with the ISO7841. However we have had no luck. Can you provide information as to why the ISO7841 part is more susceptible to GSM interference than other isolators, and if there is any other methods which may improve the ISO7841's RF immunity. 

To confirm we have tried other TI isolators (such as the ISO7741) and they also suffered from the same issue.

Kind regards

  • Hi Robert,

    Thanks for reaching out and sorry to hear about the issue you are facing.

    ISO7841 output with RF

    I believe you meant to share a different picture to show jitter at ISO7841 output instead chose the previous image accidentally. Please do share the waveform showing jitter at ISO7841 output.

    ISO7841, along with ISO7741 and ISO6741, employ OOK architecture where a carrier of ~450MHz is used to modulate input signal and transmit across the barrier. External radiations with frequency close to ~450MHz or to its harmonics could interfere with ISO7841 communication if the radiation source is too close to ISO7841. This is more likely when the RF modem is sitting in the same PCB and very close to the isolator. To address this issue, we recommend that a stitching cap is used between GND1 and GND2 of ISO7841. You mentioned that you have already tried out the solution proposed, could you please share more details on what exactly has been tried out? A Y-Cap or a stitching cap formed using internal PCB layers? May I know the value of the capacitor?

    Please note that the older device families, like ISO7341C, use a different architecture which do not require such a high frequency carrier. Hence, we do not expect these devices to be affected by any nearby radiations.

    Please do share the requested information to further help you with the issue, thanks.


    Regards,
    Koteshwar Rao

  • Hi Koteshwar,

    Thanks for your prompt response.

    "I believe you meant to share a different picture to show jitter at ISO7841 output instead chose the previous image accidentally. Please do share the waveform showing jitter at ISO7841 output."

    Yes I have updated the screen shot.

    I will document the exact solutions we have tried shortly.

    "Please note that the older device families, like ISO7341C, use a different architecture which do not require such a high frequency carrier. Hence, we do not expect these devices to be affected by any nearby radiations."

    We require a clearance of 11mm and 5s isolation of 5.4KV. This is why we use the ISO7841DWW. Are you aware of any TI products which fit this requirement?

    Kind regards

     

     

  • Hi Robert,

    Thanks for your inputs. Unfortunately, the DWW package is only available in the ISO78xx family and device ISO7741. The other approach you could consider is to use two ISO7341C devices in series to achieve higher clearance. I understand this isn't the best solution as it requires additional power supply rails, I am just trying to give you additional options apart from fixing the issue with your current solution that is being evaluated.

    I will look forward for your inputs on the experiments you have tried out to fix the issue. Thanks.


    Regards,
    Koteshwar Rao

  • Hi Robert,

    I wanted to check back to see if you had a chance to document the solutions that you have already tried out. Please do share, thanks.


    Regards,
    Koteshwar Rao

  • Hi, we have fitted a 100pF Y6P capacitor across the gnd pins and this has made little difference.

    We have found that placing a cut of IFM16-030EB300X200 on top of the isolator effectively removes the issue in our test set up. However this is probably due to the location of our antenna. But it must be said that this EMI absorber on the part makes a large difference. The problem with this as a solution is that it is a conductive material (albeit with a very thin layer of insulation) which compromises our creepage area. We are reviewing our standards to see if there is a way we can use this material safely.

  • Hi Robert,

    Thanks for your inputs, this is useful to know.
    Yes, metal sheets do help absorb emissions in either direction and I agree they are not very easy to incorporate.

    Is it possible to share a photograph of device and Y-Cap showing how they are soldered onto device? This capacitor is a lot more effective for far field emissions suppression but for near field emissions, the noise coupling mechanism can be different / multiple. If possible, please do try increasing this cap value to 1nF or even 4.7nF and soldering them directly on device GND pins with shortest leads possible.

    I will look forward for your inputs, thanks.


    Regards,
    Koteshwar Rao

  • Hi Robert,

    Since the next steps are clearly defined, I will go ahead and mark this E2E thread as closed. Once you have test results, please do share us the same. Thanks.


    Regards,
    Koteshwar Rao