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.

  • Resolved

EMI XTR116

Prodigy 30 points

Replies: 5

Views: 3496

I have a design using the XTR116 With a isolated Power supply and external loop Power. It Works well, but we fail on emc immunity testing (both conducted and radiated suceptibility) in freq range 10-80MHz. By setting an output current to say 18mA, this current will change down to 5mA at the worst during the suceptibility test. Removing the noise, and the current get back to normal. Any ideas on how to get rid of this immunity problem?

NOTE: The noise is picked up through the external current loop.

Brgds
Magne Farstad - Powex AS

  • Hello Magne,

    Please take a look at one of our "CerTIfied" precision reference design, TIPD126, that provides radiated immunity testing data using the XTR117 current loop transmitter.  The XTR117 is very similar to the XTR116 but without the internal precision reference.  http://www.ti.com/tool/tipd126

    You can take a look at the schematic on page 27 of the reference design PDF and the full design files are available on the web page as well.  The two big differences between your design and the reference design are the lack of ferrite bead filtering for EMI/RFI rejection, as well as some type of transient voltage suppressor device which is helpful in passing lightning surge (IEC61000-4-5) and other high voltage transient immunity testing.

    If you'd like, please send me a private message with your PCB layout and we can review it to see if there are any improvements we can suggest to help with the EMC performance.

    Regards,
    Collin Wells
    Precision ADC Applications

  • Hello Magne,

    I'm not with TI; I'm just another engineer who's used the XTR116, done EMC, etc.

    A couple of questions and considerations:

    Your ACPL-C87A seems to have a differential outout voltage.  Can it be connected single-ended?

    If not, I wonder how a single-ended ouput device (GND2 connected to IRET) would fare under test.

    Also, on the conducted, I assume you are coupling fast transients and surges into the AC of your isolated supply (and not the DC) per IEC 61000-6-2, Table 3, NOTE 3 ... ?  Because the onus is on the supply, not your device and the supplies are almost always covered in this regard.

    On the radiated, are you talking capacitively coupled?  to which side/cable?  Voltage input or the current loop?  

    Also, IEC 61000-6-2 has performance criterion B (4.2) and C (4.3) to consider ... wherein performance can be affected under test but returns to normal operation after the test.

  • In reply to Kris Paxton:

    Hello Magne,

    Kris pointed something out that I missed the first time which is that the ACPL-C87A output is differential.  I'm a little concerned that you have a fundamental grounding issue present on the input side of the XTR116 because I don't see how the opto-isolator is referenced to the IRET pin of the XTR116 which is required to create a proper common-mode voltage for the input of the XTR116.  

    As Kris suggested I would recommend implementing the circuit in the ACPL-C87A datasheet to convert the differential output of the opto to an appropriate single-ended input to the XTR.  I've copied the circuit below.  

    In a configuration similar to this the GND2 net of the opto needs to be connected to the IRET pin of the XTR so they share a common GND potential and will therefore interface properly.

    Also, where do +5V2 and GND2 come from?  It looks like you have a separate supply voltage/GND for these devices which may also cause issues if not properly biased to the XTR116 IRET pin.

    Regards,
    Collin Wells
    Precision ADC Applications

  • In reply to Collin Wells:

    Hello Collin,

    Yes, it is a separate isolated dcdc converter from XP Power providing the +5V2 and GND2.

    Since the ciruit Works fine, except during some frequencies during the immunity testing, I will first do a test adding ferrite beads on the mA-lines to verify if that is sufficient to pass the test.

    Thanks !

    Best regards

    Magne Farstad - Powex AS

  • In reply to Magne Farstad:

    Hi Collin,

    To introduce minimum additional Circuits, we did find a solution only adding 3 capacitors.
    The solution that passed the EMI test was to add 10nF capacitor between pin 5 (GND2) and 6 of the ACPL, and the same between pin 2 and 3 of the XTR. In addition we put a 100n between pin 4 of the XTR and pin 5 of the ACPL (GND2)

    kind regards

    Magne

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.