3
\$\begingroup\$

I am trying to make the CQB50W12-72S12 120VAC input capable by adding a bridge rectifier into the design that is already called out in the application note (app note and data sheet found here: https://www.cincon.com/productdetail/CQB50W12.html).

See picture of what I have drawn up in altium: enter image description here

Because it is AC powered now I figure I need line to earth surge protection in order to get the same or better ratings for 61000-4-5. I looked up some littelfuse application notes on surge protection and saw that you can add impedance between surge clamping components for better coordination. I am going to start building what I have in LTspice but would like to get someone else's eyes on it for a sanity check.

I think that the mov's will help protect the bridge rectifier and remaining surge will be clamped by the TVS diodes that were recommended by the datasheet. I am just trying to achieve the same 2kV line to line rating and 4kV line to earth rating that the application note declares without completely messing up the EMI filtering.

\$\endgroup\$
5
  • 2
    \$\begingroup\$ 120 volts AC produces a peak voltage of 170 volts but, the converter you are using is only specified for continuous operation up to 160 volts DC. \$\endgroup\$
    – Andy aka
    Sep 29 at 13:54
  • \$\begingroup\$ Is MOV/TVS to GND a good idea? It's not allowed for a number of installations / wiring environments / product types. \$\endgroup\$ Sep 29 at 14:19
  • 3
    \$\begingroup\$ @Erv it seems a bit pointless to do any further analysis given the flaw on voltages. \$\endgroup\$
    – Andy aka
    Sep 29 at 15:14
  • \$\begingroup\$ @TimWilliams oh i did not know that, which installations is an mov to gnd not allowed? how do these installations protect against surges and achieve high line to earth surge ratings? \$\endgroup\$
    – user298907
    Sep 29 at 16:25
  • \$\begingroup\$ @Erv Well, maybe that's a solution. Though this drifts off the topic of design here, but it is nonetheless an issue engineers regularly have to deal with. Is there a reason you are required to use this unit, perhaps you were ordered to by a higher-up? If they aren't listening to any push-back on a matter that clearly several other professionals have identified is a critical, functional flaw, then chronic mismanagement may well be the underlying issue here, and there's only one sure way you can fix that: change jobs. \$\endgroup\$ Sep 29 at 18:01

1 Answer 1

1
\$\begingroup\$

Because it is AC powered now I figure I need line to earth surge protection in order to get the same or better ratings for 61000-4-5.

I don't think you can get better immunity as the converter module already complies with 61000-4-5 Level-4 which is the toughest level already (unless a higher one is defined with Level-X by the manufacturer or yourself). Of course, this doesn't mean that you should rely on the converter module's surge protection so it's good to implement a separate protection for the AC-input-side components.

The datasheet on p.3 says the converter module complies with 61000-4-5. But the app note on p.14 (under Safety & EMC) mentions the requirement of external components in order for the converter module to comply with 61000-4-5.

I think that the mov's will help protect the bridge rectifier and remaining surge will be clamped by the TVS diodes that were recommended by the datasheet. I am just trying to achieve the same 2kV line to line rating and 4kV line to earth rating that the application note declares without completely messing up the EMI filtering.

The idea of placing TVS diodes for common- (D1, D2) and differential-mode (D3) surges is correct, however, the selection of the components doesn't seem to be. I don't know if you have a grasp of 61000-4-5's details, but apparently, you missed the fact that the currents during a surge, depending on the class/level of course, can reach hundreds of Amps.

What you need to do first is to check the standard and the required class/level for your application, then pick diodes having correct reverse stand-off and clamping voltages, and most importantly, an IPP rating (for 8/20 us) of higher than the class/level requirement. TVS diodes for surge immunity/protection generally mention what standards they are compatible with. This can be a good starting point. For example, if you have a look at the datasheet or product highlights of SMDJ180A you'll see that it's not compatible with 61000-4-5 as its peak pulse current rating is only 51 Amps.

As for placing MOVs for common-mode surge (across L-PE and N-PE), as Tim pointed out in his comment, you need to be very careful because, since the MOV will always be under mains stress (assuming N and PE have the same potential), a leakage current will flow through it to the earth which can cause self-heating (if not selected properly) and eventually initiating positive feedback and possibly destruction (i.e. heat causes higher leakage current, therefore higher self-heating, so higher self-heating causes even higher leakage current which causes even higher self-heating. This goes on until destruction). This may result in shock or even fire. Also, you may want to place another fuse on the N-side, plus you may want to cover the TVSs with nonflammable material.

\$\endgroup\$
5
  • 2
    \$\begingroup\$ As I answered in a companion question, I strongly suspect they tested to EN50121, where the IEC 61000-4-5 surge is done through quite high resistance, not direct coupled. This seems to perfectly explain the choice of diode. \$\endgroup\$ Sep 29 at 16:27
  • \$\begingroup\$ @TimWilliams referring to your previous comment; littelfuse shows in this slideshow mov's in the configuration i have. Thats kind of where i got it from. psma.com/sites/default/files/uploads/… \$\endgroup\$
    – user298907
    Sep 29 at 16:55
  • \$\begingroup\$ @Erv For luminaires (street lighting), it seems? I don't know about that market specifically; I would guess there is some variation between jurisdiction (some may still be using constant-current loops(!), though with the adoption of LED fixtures I'm betting their days are even more quickly being numbered), but it's definitely not office or medical equipment, for example. \$\endgroup\$ Sep 29 at 17:00
  • \$\begingroup\$ @timwilliams please forgive my ignorance and lack of experience. Would industrial equipment allow for the use of 3 mov's arranged as I have? Or how would I look up this information? It seems to me that the 3 mov's arranged as such would clamp L-L and L-E surges and with coordination with the TVS diodes downstream could increase the surge rating the datasheet specifies. \$\endgroup\$
    – user298907
    Sep 29 at 17:07
  • 1
    \$\begingroup\$ @Erv MOVs clamp the surge pulses to some extent as they behave like back-to-back avalanche diodes, so placing one across L-PE would seem a good idea for common-mode surges. However, placing anything (Y-cap, MOV, TVS, etc) across L-PE will increase leakage current. Your system will probably fall into IEC Class-I where shock protection is achieved with a protective grounding as well as insulation. So the ground leakage may become more of a concern here. Consider GDTs as well, either along with MOVs or as a replacement for them. \$\endgroup\$ Sep 29 at 17:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.