+ Post New Thread
Results 1 to 9 of 9
  1. #1
    Advanced Member level 5
    Points: 20,536, Level: 34
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,370
    Helped
    1201 / 1201
    Points
    20,536
    Level
    34

    Transformer creepage with faraday shields

    I'm starting a medical PSU project (~1kW), low leakage and low EMI are critical. I planned to address this by using faraday screens in the DC-DC transformer, in order to reduce or eliminate the Y capacitor between primary and secondary.

    The one thing I'm stuck on is how the use of the shields affects the my insulation category and creepage distance. Normally, without any faraday shields, you can just use appropriately rated triple insulated wire and be done. But if the screen(s) are put between the windings, then I expect that it would eliminate the protection gain from the TIW, assuming the shield connects directly back to the primary or secondary circuit ground. So does that mean there's no point in using TIW with a shield?

    What if the shield is connected to the circuit ground via a ~10nF Y capacitor (rated for the working insulation voltage)? Could the shield then not be considered an issue for insulation failure, since it's effectively an open circuit at 50/60Hz? I've seen this trick mentioned in a couple texts, but can't find any reference to it in IEC 60950 or IEC 60601-1.

  2. #2
    Advanced Member level 5
    Points: 9,475, Level: 23

    Join Date
    Aug 2015
    Posts
    1,538
    Helped
    557 / 557
    Points
    9,475
    Level
    23

    Re: Transformer creepage with faraday shields

    if the shields touch then you will have lots of C from mains to o/p. Shields are subject to the same insulation requirements as winding wire - you might consider 3 parallel wires wound as a shield for the pri then the TIW sec.


    1 members found this post helpful.

    •   AltAdvertisement

        
       

  3. #3
    Advanced Member level 5
    Points: 20,536, Level: 34
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,370
    Helped
    1201 / 1201
    Points
    20,536
    Level
    34

    Re: Transformer creepage with faraday shields

    Quote Originally Posted by Easy peasy View Post
    if the shields touch then you will have lots of C from mains to o/p.
    I don't see why the capacitance between a primary shield and a secondary winding would be more than between the primary and secondary. Assuming the insulation is the same for both cases (though in that case the shield will increase leakage a bit).
    Shields are subject to the same insulation requirements as winding wire
    This is where I'm getting lots of mixed messages from sources. For example, I've also seen suggestions that using a safety screen (rated for 25A) to protective earth may actually relax requirements on creepage and insulation.
    you might consider 3 parallel wires wound as a shield for the pri then the TIW sec.
    In such a case, would you make the shield out of TIW, and the primary with standard insulation?



    •   AltAdvertisement

        
       

  4. #4
    Advanced Member level 5
    Points: 9,475, Level: 23

    Join Date
    Aug 2015
    Posts
    1,538
    Helped
    557 / 557
    Points
    9,475
    Level
    23

    Re: Transformer creepage with faraday shields

    3 wires of PEI or PUR functional insulation to the primary - then safety insulation to the sec

    as to a single earth shield - yes you are right - as long as the shield can easily carry any fault current - usually a physically large choke is needed to block the RF pickup by the earth shield which then gets into the LISN ( along the earth wire )

    a pri shield and then an earth shield overcomes this

    if the shields touch then you will have lots of C from mains to o/p.
    I mean't if you had large caps in series with these shields and they then touched ( metal to metal ) inside the TX ...


    1 members found this post helpful.

    •   AltAdvertisement

        
       

  5. #5
    Advanced Member level 5
    Points: 20,536, Level: 34
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,370
    Helped
    1201 / 1201
    Points
    20,536
    Level
    34

    Re: Transformer creepage with faraday shields

    Quote Originally Posted by Easy peasy View Post
    3 wires of PEI or PUR functional insulation to the primary - then safety insulation to the sec
    Could you elaborate? "3 wires of functional insulation"... you mean three layers of insulation between the primary and shield?

    as to a single earth shield - yes you are right - as long as the shield can easily carry any fault current - usually a physically large choke is needed to block the RF pickup by the earth shield which then gets into the LISN ( along the earth wire )

    a pri shield and then an earth shield overcomes this
    Yes, I definitely wouldn't consider using a safety shield without also using EMI shields on both sides. I don't plan at this point to include a safety shield.

    I mean't if you had large caps in series with these shields and they then touched ( metal to metal ) inside the TX ...
    Well certainly that would cause a large increase in leakage currents. But it would be much worse without the series capacitors on the shields...



  6. #6
    Advanced Member level 5
    Points: 9,475, Level: 23

    Join Date
    Aug 2015
    Posts
    1,538
    Helped
    557 / 557
    Points
    9,475
    Level
    23

    Re: Transformer creepage with faraday shields

    3 wires of PEI or PUR(II) for the shield. Functional isolation is that for the function of the power circuit, i.e. not safety, if the peak volts from any point on the pri to shield is say 650V ( e.g. flyback) then functional insulation has to withstand this over time.



  7. #7
    Advanced Member level 5
    Points: 20,536, Level: 34
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,370
    Helped
    1201 / 1201
    Points
    20,536
    Level
    34

    Re: Transformer creepage with faraday shields

    Quote Originally Posted by Easy peasy View Post
    3 wires of PEI or PUR(II) for the shield.
    I'm assuming by PEI you mean Ultem or something similar?

    So here's the most basic method of construction I can think of. Margin wound, no TIW, just using tape insulation.
    Click image for larger version. 

Name:	zzzzz.png 
Views:	4 
Size:	41.9 KB 
ID:	152658

    Basically I'm just want to fact-check these assumptions of mine:

    1. Using TIW for the primary winding alone is pointless, due to the presence of the primary shield.

    2. Using TIW for both the primary shield and the secondary winding would allow me to eliminate all the tape insulation.

    3. In the case of 2, the creepage from primary to secondary would be <8mm, so I would still need the margins (since the primary is not TIW).

    4. Using TIW for all windings and the shield would allow me to get rid of all the insulation tape, and the margins.



    •   AltAdvertisement

        
       

  8. #8
    Advanced Member level 5
    Points: 9,475, Level: 23

    Join Date
    Aug 2015
    Posts
    1,538
    Helped
    557 / 557
    Points
    9,475
    Level
    23

    Re: Transformer creepage with faraday shields

    You need to read thoroughly & understand the related standards if you are to take responsibility for the safety of the design.

    PEI = poly ethyl iamide - a common winding wire insulation for > 180degC it cannot be soldered

    PUR = poly urethane - commonly red & solderable

    if using approved TIW no margins are needed - if using margins and PUR or PEI then the lead outs need to be sleeved to meet standards - and the tape used for isolation needs to meet standards - the insulation on the wire does not count - except TIW.

    If you use TIW as a shield and simply cut the end of the shield wdg, then the Cu is exposed and you need another insulation system - so may as well use 3 x PUR in //


    1 members found this post helpful.

  9. #9
    Advanced Member level 5
    Points: 20,536, Level: 34
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,370
    Helped
    1201 / 1201
    Points
    20,536
    Level
    34

    Re: Transformer creepage with faraday shields

    Quote Originally Posted by Easy peasy View Post
    PEI = poly ethyl iamide - a common winding wire insulation for > 180degC it cannot be soldered

    PUR = poly urethane - commonly red & solderable
    I'm familiar with PU (never heard it called PUR), polyimide, and polyamide. Never heard of poly ethyl iamide as a wire insulation, and google doesn't turn up anything. Are you sure you don't mean polyester imide?

    if using approved TIW no margins are needed - if using margins and PUR or PEI then the lead outs need to be sleeved to meet standards - and the tape used for isolation needs to meet standards - the insulation on the wire does not count - except TIW.
    Yes, AFAIK IEC 60950-1 and 60601-1 are consistent here.

    If you use TIW as a shield and simply cut the end of the shield wdg, then the Cu is exposed and you need another insulation system - so may as well use 3 x PUR in //
    Good point. I suppose one option would be to terminate both ends of the shield strands to bobbin pins, but not connect it to the PCB. That would occupy quite a few bobbin pins though.



--[[ ]]--