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.

ISOW7842: ISOW7842 Radiated emission failure

Part Number: ISOW7842
Other Parts Discussed in Thread: ISOW7841

Hi I am using ISOW7842 in my design  and due to which my product is failing radiated emission test. There is very high emission at around 66MHz.

Can someone help me out to reduce the radiation level  to pass the product.

  • That's a familiar story.

    I'm using ISOW7841 and am failing CISPR11 Level B at 65MHz, despite taking several steps recommended in the application note for the ISOW784x parts.

    I was failing even level A but adding common mode chokes to the input and output supplies to the isolator and ferrite beads to supply and ground connections as shown in the note has reduced the radiated noise considerably.

    The only way I've found to reduce the noise below level B is a conductive screen above (and maybe below) the ISOW7841. The screen doesn't have to be grounded or extend very far from the isolator to be effective. Unfortunately, incorporating this into production units will be tricky and the product is double insulated in a plastic enclosure.

    The sniffer probe suggests there is a stray field above the part. A horizontal loop produces a peak above the chip, a vertical loop produces a null above the chip with peaks either side. I am assuming this is from the power transfer device (transformer?). I am also going to try making some kind of magnet circuit around the chip with ferrite sheets to see if that works but a conductive screen looks favourite for now.

    If anybody has any other ideas, please chip in....

  • Hi Dileep, Alex,

    Thanks for the reaching out to us with your concern on meeting radiated emissions requirement with ISOW7841. Please do note that ISOW7841 alone meets CISPR22 Class B, what make emissions get bigger is the PCB size and cables connected directly or indirectly. If you could separate device pins with rest of the world through use of below listed ferrite beads, you would be able to minimize the impact of PCB and cables connected to ISOW7841. Please use below ferrite beads or equivalents with >1k impedance at 100MHz.

    Power Ferrite Bead – FBMH3225HM202NT (for VCC1, GND1, VISO & GNDISO)
    Signal Ferrite Bead – BLM18BD252SZ1D (for INx & OUTx)

    It is also important to note that your emissions test setup doesn't contribute to increased radiated emissions. Please refer to Section 7 of App Note SLLA368A for more details. Thanks.


    Regards,
    Koteshwar Rao
  • Hi Koteshwar,

    Thanks for your suggestions, I'll try those ferrites as soon as I can.

    I'm curious about the effectiveness of screening though. If I cover just the top of the ISOW with aluminium foil, emissions are significantly reduced.

    Can you suggest a mechanism for this? I want to avoid having a screen if at all possible so better filtering is attractive if it works....

    Unfortunately, ISOW7841 alone does not a product make....

    Cheers,

    Alex.

  • Hi Koteshwar,

    I tried with ferrite bead also but did not see any improvement in emission level. Even i tried with different combinations of decoupling capacitors.

    Can you suggest some other way to reduce the emission level.

  • Hello Koteshwar,

    I have revisited Section 7 and note that the test board for ISOW7841/7842 is sitting on top of the box lined with a screen.

    I think that provides an partial screen for the chip and will be reducing emissions compared to what would be seen with the board placed on the table.

    Our product has no conductive enclosure so the test set-up cannot be reproduced in our tests, without adding an screen somehow.

    Alex.

  • Hi Dileep,

    I suggest you try a conductive screen, just to see what effect it has. I found simply wrapping our plastic enclosure in aluminium foil (cooking foil) made a big difference, even without a connection to PE.

    Alex.
  • Hi Alex,

    Thanks for sharing more information on the experiments you have conducted.
    Regarding your question related to reasoning behind screen / aluminum foil improving emissions, please note that the aluminum foil cover over device would form a stitching capacitor between Side1 & Side2. Hence this would significantly improve emissions.

    Regarding your comment on Section 7 test setup image with block box, please note that the box is used to house battery, LDO and connecting cables. The box has inner Aluminum lining so that components inside box do not contribute radiated emissions. The white lining that you see on box right underneath ISOWEVM is not Aluminum foil, it is a label written on a paper pasted onto black box. The only place where you see Aluminum foil is used external to black box is in the center where there is an opening to pass power supply cables.

    Hi Dileep,

    One possible reason for not meeting emissions requirement even after using ferrite beads on power supply as well as signal lines is generally due to large PCB size. If ferrite beads are not enough then using stitching capacitor formed through PCB layers (as described in SLLA368A) would be the only option to improve radiated emissions further.

    If you could share us your schematic, PCB layout, emissions test setup image and emissions test results then we could check and comment if there is room for improvement to meet emissions requirement without implementing stitching capacitor. Do share us the data, thanks.

    Regards,
    Koteshwar Rao
  • Hi Koteshwar,

    Thank you for your interest in this.

    I have to admit, I didn't think stitching capacitance would be a factor at 65MHz. In SLLA386, Figure 10, there seems to be no difference between the emissions at 60MHz between the two cases, with and without stitching capacitance.

    In our design, there is a 220pF THT capacitor between the ground planes. This part has an SRF >100MHz so I expect it to still be quite effective at 65MHz. Figure 10 only shows the effect of the stitching capacitance above 200MHz.

    I understand that there is no foil on top of the box in Figure 14 but the lining of the box must be just one or two mm from the DUT and will still be a pretty effective screen for one side of the DUT.

    Yours,

    Alex.

  • Hi Alex, Dileep,

    Were you able to resolve your ISOW78xx emissions issues? I see that no one has responded to this thread in a few weeks and I wanted to ensure you are getting the support you need. If you still require more support please respond to this thread or open up a new thread.

    Best regards,
    Dan
  • Hi Dan,

    Thanks for following up.

    I have a modified board going into pre-compliance testing on Monday.

    I have added ferrite beads to all the signal lines to the ISOW, in addition to those already present on all the cable connections off the board and on the ISOW power supply.

    I also have a case with screening to try out if that doesn't work out.

    I'll report back when I get the results....

    Alex.

  • Hi Dan,

    I have results from our pre-compliance tests to Class B.

    Adding ferrite beads to all the signal lines on both sides of ISOW7842 made no significant difference and we still fail Class B by about 3dB at 65MHz.

    Adding two areas of screen above and below the ISOW7842 does make a difference and brings us to a pass but with zero margin.

    These screens are about 50x50mm and are not grounded or connected together.

    I will need a further 3dB improvement at 65MHz to achieve a secure pass, any suggestions will be very welcome.

    Alex.

  • Results:
    Adding beads to the ISOW signal lines (in addition to the power lines) made no measurable difference.

    Screening did though. So, if we need to pass Class B, we will need to implement a screen.