Creepage distance pri to sec in an offline SMPS

[ocset]

Hi
Can I just check that in an offline SMPS in UK,EU,USA, the primary to secondary conductor creepage distance should be >=8mm?

We have a student with us whose father is a Top Expert Chief Power Electronics Engineer, and he says the rules have just changed and its now 5mm.

Also, this student has layed out a 70W Offline Flyback for us. He has L,N & E connections to the primary mains input connector. The earth is then tracked onto the PCB and picks up the Y capacitors at the primary side. ...Also, this earth copper is then tracked to the primary FET heatsink PTH's to which the plastic TO220FP FET is affixed. However, he has then continued to track this earth on the bottom layer from primary to secondary side. First this earth connects to the secondary diode heatsink’s PTH holes. (TO220 diode). Then this earth continues throughout the secondary side where it connects to a PTH to which the earthed metal “cage” is soldered into. This cage is “holy” and is to reduce radiated emissions as it kind of “box covers” the top PCB side of the entire SMPS.

[EDIT.................
…So, the earthed copper is on primary and secondary sides. However, on both primary and secondary sides this earthed copper tracking/pouring is less than 1mm away from primary and secondary metal conductors.
So, the earthed copper is on primary and secondary sides. However, on pri side, this earth tracking is 2mm away from a pri side trace, and on the secondary side, this earthed trace is 0.5mm away from the secondary side traces..............EDIT]

He says his father told him that all this is right and the regs have just changed.

However, do you agree this is not the case, and his SMPS is very bad?

[Gyro]

The "regs have just changed" is rubbish. However 8mm primary - secondary creepage and clearance is for Class II double insulated products (reinforced insulation requirement).

If you have a mains earth connection to the PSU, which is bonded to the secondary side, then the requirement drops to 4mm (allow a little wiggle room). 1mm creepage and clearance between earth and primary side tracks is clearly not acceptable.

You really need to purchase a copy of the appropriate Low Voltage Directive spec and understand it if you are developing commercial products.

[AndyC_772]

+1

Buy a copy of BS:EN60950, or whatever other document relates to your specific product type.

Read it.

UNDERSTAND it.

As for "his father told him that all this is right", I'll just leave this here: 

[Gyro]

Note that he mentioned the US too. That is UL 62368-1, probably identical requirements but if you're shipping there, you need to check.

I am somebody's father too, and with long experience in the industry. It doesn't make me a 'reliable' source for information relating to meeting safety specs. Second hand word of mouth from your student's dad isn't acceptable.

[ocset]

Thanks,
The following on page 9 gives some creepage distances for PCBs...
http://www.alternatezone.com/electronics/files/PCBDesignTutorialRevA.pdf

Page 117 of EN62368 gives this formally, but needs you to know what material group and  what "pollution degree" you have.
However, to find this out you must purchase further standards it seems.

If you have a mains earth connection to the PSU, which is bonded to the secondary side, then the requirement drops to 4mm

Thanks, Our earth connection is not bonded to the PCB...it just comes in on a earth wire into the mains input connector, and then tracks all over the PCB as described  in the top post. I wondered if this affects things?

Most places where ive worked just get in a "standards consultant" (at big cost) to come in and tell them what they need to do to get through the regs. No attention is payed to the regs.....just the words of the standards consultant are heeded. They tell you how to get through regs with the minimum cost, and if there are any nice loopholes for you to exploit. . Is this common?

[ocset]

Thanks,

1mm creepage and clearance between earth and primary side tracks is clearly not acceptable.

I must admit it was a surprise to me to hear that earth can be tracked from the mains input connector...right round to the secondary side, and be 0.5mm away from tracks in the secondary side and 2mm away from  any tracks on the primary side...giving a total creepage distance of 2.5mm between primary and secondary side conductive tracks...and this is absolutely within regulations (because its earthed copper)
I have been assured that  this is because being earth, it is the safety earth, and so any flash over to it would blow the fuse.
Of course , If a floating piece of copper were tracked between primary to secondary like that...then it would not be acceptable.

[Gyro]

I must admit it was a surprise to me to hear that earth can be tracked from the mains input connector...right round to the secondary side, and be 0.5mm away from tracks in the secondary side and 2mm away from  any tracks on the primary side...giving a toal creepage distance of 2.5mm between primary and secondary side conductive tracks...and this is absolutely within regulations (because its earthed copper)
I have been assured that  this is because being earth, it is the safety earth, and so any flash over to it would blow the fuse.
If a floating piece of copper were tracked between primary to secondary like that...then it would not be acceptable.

2.5mm separation might meet Creepage due to the interposed earth track, but it certainly does not meet Clearance. If you are being assured of these things by your "student" the you are not supervising him properly. The idea is that you teach him - he doesn't assure you that 'It'll be fine'!

At some point there becomes a transition from the standard to common sense and product reliability. Such tight clearance will lead to breakdown under transient conditions, leakage if no-clean flux is used etc. Do you really want the fuse blowing on a significant proportion of your products. Also, with such a lazy earth copper pour, are you certain that the fuse will blow before a narrow point in the trace (potentially breaking your secondary side earth connection)?

Our earth connection is not bonded to the PCB...it just comes in on a earth wire into the mains input connector, and then tracks all over the PCB as described  in the top post. I wondered if this affects things?

It would be irresponsible for me to comment positively, given your description of the routing in your original post - The earth trace sounds to wander all over the place. Conservative design says that you need to have a direct and substantial routing between the (what sounds like) PCB mounted mains connector and the output ground signal.

Most places where ive worked just get in a "standards consultant" (at big cost) to come in and tell them what they need to do to get through the regs. No attention is payed to the regs.....just the words of the standards consultant are heeded. They tell you how to get through regs with the minimum cost, and if there are any nice loopholes for you to exploit. . Is this common?

Absolutely not. Standards consultant sounds like an expensive and unreliable way of doing things and leads to the sort of ignorance of the standards that seems to be evident. You need to build up in-house knowledge that can be applied to every stage of the design and implementation (you certainly shouldn't be looking for "nice loopholes" with safety related standards). That way you will have confidence in the product when you take it to the test house... Please tell me that you are using an accredited test house rather than self-certifying.

[ocset]

2.5mm separation might meet Creepage due to the interposed earth track, but it certainly does not meet Clearance.

Thanks, though the creepage and clearance allowance between live and neutral traces just downstream of the mains input connector is 2.5mm.
If they arent suspected of flashing over between them, then why should it flash over from  a pri side trace to earth trace that are 2mm apart?

Also, with such a lazy earth copper pour, are you certain that the fuse will blow before a narrow point in the trace (potentially breaking your secondary side earth connection)?

Its a failrly thick kind of "Poured track", at least 4mm wide everywhere, apart from one 6mm section, which is well on the secondary side, which is a 2mm wide track. Even if this pour did blow somewhere along it, how would that then make the clearances/creepages worse?

Such tight clearance will lead to breakdown under transient conditions

In our case we have a MOV after the input connector and a 150uF electrolytic aftert the mains diode bridge, and as you know, very few if any mains transient would ever be able to lift that cap's voltage much above 390V say.

[Gyro]

2.5mm separation might meet Creepage due to the interposed earth track, but it certainly does not meet Clearance.

Thanks, though the creepage and clearance allowance between live and neutral traces just downstream of the mains input connector is 2.5mm.
If they arent suspected of flashing over between them, then why should it flash over from  a pri side trace to earth trace that are 2mm apart?

You have edited your OP - you said 0.5mm between Primary side and Earth get your facts straight. Regardless, you are required to meet at least 4mm Clearance between primary and secondary - you seem to be only thinking of creepage! How can I see what those are?

Also, with such a lazy earth copper pour, are you certain that the fuse will blow before a narrow point in the trace (potentially breaking your secondary side earth connection)?

Its a failrly thick kind of "Poured track", at least 4mm wide everywhere, apart from one 6mm section, which is well on the secondary side, which is a 2mm wide track. Even if this pour did blow somewhere along it, how would that then make the clearances/creepages worse?

So you are relying on a 2mm wide track 'fuse point' to ensure the earthed ground to your secondary circuit. Either it needs to be earthed or it doesn't. I have no idea what it is feeding, whether secondary side fed items are user accessible. What your enclosure is. what your use case is.

Such tight clearance will lead to breakdown under transient conditions

In our case we have a MOV after the input connector and a 150uF electrolytic aftert the mains diode bridge, and as you know, very few if any mains transient would ever be able to lift that cap's voltage much above 390V say.

Again, you were saying 0.5mm Primary - Earth when I made the statement.


Look, this has turned into yet another typical 'Treez' thread.   You haven't provided any concrete data to base this thread on - Where is the schematic and PCB layout that you are talking about? - It would be expected of any normal poster. You expect people to second guess you on what you do provide, change it half way through, and expect your questions to still be relevant. You rely on second hand information from a student's dad rather than reading and understanding the relevant specifications, and basically expect people here to  do your work for you!

I'm done!