Author Topic: EN 60950 and Clearances in Primary Circuits  (Read 1914 times)

0 Members and 1 Guest are viewing this topic.

Offline Darren DeBattistaTopic starter

  • Newbie
  • Posts: 9
  • Country: mt
EN 60950 and Clearances in Primary Circuits
« on: September 18, 2018, 10:22:26 am »
Hi All,

I am designing a new product which has a Live-in Live-out power supply. This product will have no isolation and can be seen as to have a linear regulator from the Live-in Live-out to produce a 3.3V supply for the micro controller. According to the type of product, the standard I need for safety is the EN 60950. Looking at table 2K fro minimum clearances for over voltage category II (2,500V) and pollution degree 2, the minimum clearance for functional insulation is 1.0mm. I'm finding this to not be realistic as not even the components that are used (for example the MCU) has a clearance of 1mm between the pads. I suspect I might be missing something. Should the 3.3V be considered as ELV and use Table 2M for the clearances? Is there anyone who can help me on this?

Thanks
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: EN 60950 and Clearances in Primary Circuits
« Reply #1 on: September 18, 2018, 11:29:35 am »
There's no need for 1mm clearance within a circuit running at 3.3V supply, even if it is on live potential. That whole circuit will require clearance to its surroundings if they are on a different potential. The functional clearance is required e.g. from live potential  to neutral, and for components that have that voltage across them.
Safety devices hinder evolution
 

Offline Darren DeBattistaTopic starter

  • Newbie
  • Posts: 9
  • Country: mt
Re: EN 60950 and Clearances in Primary Circuits
« Reply #2 on: September 18, 2018, 12:39:23 pm »
Hi Capt,

Thank for the reply. I agree with you but issue is that I cannot find anything which I can relate to in the EN 60950. Also, I would anyway expect that the clearances between two traces would be proportional to the difference in voltage between them. Table 2K specifies values for peak working voltage 'up to and including' and start from 71V. Hence I am assuming that the 3.3V area is considered as part of the 71V in table 2K but I maybe be wrong. On the other hand, it doesn't make sense to maintain 1mm for a 3.3V rail. Is there any information regarding non-isolated extra low voltage (ELV) clearances which I maybe not know off?

Thanks
 

Offline capt bullshot

  • Super Contributor
  • ***
  • Posts: 3033
  • Country: de
    • Mostly useless stuff, but nice to have: wunderkis.de
Re: EN 60950 and Clearances in Primary Circuits
« Reply #3 on: September 18, 2018, 01:39:26 pm »
Hi Capt,

Thank for the reply. I agree with you but issue is that I cannot find anything which I can relate to in the EN 60950. Also, I would anyway expect that the clearances between two traces would be proportional to the difference in voltage between them. Table 2K specifies values for peak working voltage 'up to and including' and start from 71V. Hence I am assuming that the 3.3V area is considered as part of the 71V in table 2K but I maybe be wrong. On the other hand, it doesn't make sense to maintain 1mm for a 3.3V rail. Is there any information regarding non-isolated extra low voltage (ELV) clearances which I maybe not know off?

Thanks
I've got the German translation of this standard only, at the very beginning (1.3.1), it says kind of: "This standard applies to parts of the circuit that are relevant for safety". I'd consider 3.3V operated circuit not relevant for safety. Afaik, this standard refers to safety in terms of "danger of fire" and "injury of humans", so as long as this part of the circuit is safe in this sense, the standard most probably doesn't apply. Don't nail me on that, ask your local notified body if you want to know for sure.

Otherwise, 3.3V for sure is ELV, but if it sits on live potential, that whole block can't be ELV to the outside, but within itself. Anyway, don't nail me on that, ask your local notified body if you want to know for sure.
« Last Edit: September 18, 2018, 01:42:46 pm by capt bullshot »
Safety devices hinder evolution
 

Offline Phoenix

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: au
Re: EN 60950 and Clearances in Primary Circuits
« Reply #4 on: September 18, 2018, 10:17:15 pm »
The voltages listed in the tables are in reference to the voltage difference between individual nodes in the circuit (peak, rms), not a blanket requirement for everywhere. Each node pair will give a creepage and clearance requirement. Then you also need to consider if functional (non safety) vs basic vs reinforced.

For your digital circuits at 3.3V the voltage between nodes is 3.3V, plus it is functional isolation so you don't need to provide any compliant clearance.

Your 3.3V might come close to the input traces at 120/240V. Depending on how the 3.3v is referenced to the rest of the circuit will determine the working voltage.

Your input power supply will be at 120/240V with reference to earth (if earthed chassis) and this will require basic isolation.

Between active and neutral will also be 120/240V (and over voltage catagory) but this may be considered functional isolation (if a short couldn't cause a hazard) but should probably design as basic.

You need to determine what is safety isolation and what's functional. There is also a diagram somewhere in there that illustrates requirements between different types of nodes.

« Last Edit: September 19, 2018, 01:44:06 am by Phoenix »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EN 60950 and Clearances in Primary Circuits
« Reply #5 on: September 19, 2018, 03:15:51 am »
Voltage is a difference, not absolute.

3.3V with no transients (we can probably safely assume the circuit would not transmit a surge through the output*), means minimum spacing between them is applicable.  Hazardous mains voltage (common mode) means suitable clearance is required around that pair, however.

*Although, if it's a series dropping circuit, that may be something to consider, as failure modes go.

The standard must be read carefully to determine this, just keep reading.  It's about as much trouble to set up the same design rules in a PCB anyway, so.. eh.

I'm not sure how much leeway there is in the rules.  One example I've designed with: a hot-swap controller rated for 100V but packaged in a 0.5mm pitch MSOP.  No way that meets UL.  Haven't heard any problems from the customer, so I guess it's probably gone through.  Those kinds of things are a bother...

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Darren DeBattistaTopic starter

  • Newbie
  • Posts: 9
  • Country: mt
Re: EN 60950 and Clearances in Primary Circuits
« Reply #6 on: September 19, 2018, 07:51:25 am »
Hi All,

Thanks for the replies. The 3.3V rails are surely ELV and hence need functional isolation in between them but what I'm concerned about is that the clearances are calculated considering the overvoltage category. The circuit is a series circuit with the mains supply and although there a the 3.3V shunt regulation, I think I anyway have to consider the 3.3V rail as being on an overvoltage category II section (Example, in the EN 61010 standard, when a secondary voltage is concerned, taking the output from an isolating transform is seen differently from when taking the output from an (example) buck regulator since that the possible transient seen on the output due to mains overvoltage is different and hence, clearances are listed differently in the standard (Table 6 vs Table K.17)). If this is so and the 3.3V are to be considered on an over voltage category II, Table 2K indicates a clearance of 1.0mm for a function isolation of <71V. Having said that, I did see what two other competitors have done and I've seen that they actually placed a polygon pour throughout the whole PCB which is connected to mains Live and only had a clearance of 0.5mm to everything. But the PCBs are not really from reputable companies and I don't really like the idea off having 0.5mm only between Live in and Live out lines. As I see it, although they are both Live lines, when the load is OFF, the whole 230VAC mains will be dropped across the line switch and hence I must keep the appropriate functional isolation (from table 2K results in 1.5mm at 420V peak). Am I right?

As what T3sl4co1l pointed out, some components don't even have the mentioned isolation but here is how it become confusing. The commonly used BTA16 triac is rated at 600Vpeak with non-repetitive peak voltage of 800V. Considering the 600Vpeak, the clearance between the pins should be 3mm for functional insulation but the leaded version has a pitch of 2.4mm which makes it impossible to maintain the required clearance on the PCB (even if operated at mains voltage, its still not possible to maintain the 1.5mm clearance between the pads on the PCB). Yet this component passes UL1557.

Now basically I have two question remaining:

1. Figure 2H shows the different isolations to maintain between different circuits (I think this is the diagram Phoenix refereed to). The first line shows what to maintain between Primary circuit and similar circuit or conductive part. Does the 'similar circuit or conductive part' means between primary circuits irrespective of the voltage difference in the sense that, between my Live-in line and my Live-out line I can keep minimum of 1.00mm functional isolation and not the 1.5mm as listed for a voltage difference of 420V? (this would agree with the competitor's PCBs I've seen although they maintained 0.5mm and not 1mm).

2. Should I consider the 3.3V rail as also being connected to an Over voltage category II section? If so, the consequences would be that table 2K  requires a clearance of 1.0mm while if not (due to the input MOV for example), the minimum of 0.2mm can be maintained. (Note , table 2M gives clearance for secondary circuits and places the burden on the reader as to which over voltage section to apply considering the measured secondary transient over voltages. But a secondary circuit is defined as one derived from an isolating transformer and hence does not apply here).


Edit: On thinking regarding the clearances, I came to realize that being a serially connected product having Live-in and Live-out, if an over voltage is to happen and an arc/short is produced, this would not happen directly between Live and Neutral as in standard mains connected products but will happen between the same Live wire, resulting in the voltage being developed across the load (which would have its own safety isolation clearance). Could this be and accepted argument and hence being the reason why competitors left minimal clearance between the Live-in and Live-out?

Thanks


« Last Edit: September 19, 2018, 08:32:24 am by Darren DeBattista »
 

Offline Phoenix

  • Frequent Contributor
  • **
  • Posts: 420
  • Country: au
Re: EN 60950 and Clearances in Primary Circuits
« Reply #7 on: September 19, 2018, 09:31:16 am »
While I don't have the standard in front of me - my memory is that functional isolation is purely for circuit operation, not safety, by definition. I don't believe you need to be compliant for functional isolation - the failure of which would not cause a hazard anyway.

Last product I designed with HV-SELV barriers the test lab didn't look into the circuit details much passed the input and output terminals anyway.

EDIT:
1.2.9.1 FUNCTIONAL INSULATION: insulation that is necessary only for the correct functioning of the equipment
NOTE FUNCTIONAL INSULATION by definition does not protect against electric shock. It may, however, reduce the likelihood of ignition and fire.

Also see 5.3.4 - If function insulation doesn't meet the required amount it is assumed as a short for testing against the listed risks.
« Last Edit: September 21, 2018, 01:29:33 am by Phoenix »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EN 60950 and Clearances in Primary Circuits
« Reply #8 on: September 19, 2018, 04:59:33 pm »
You're misusing/misunderstanding the overvoltage category -- once it's applied to a semiconductor (FWB, regulator, thyristor, etc.), the overvoltage across that component/module is limited to its breakdown voltage. :)

It would be quite remarkable, after all, if you had a 1500V surge between +3.3V/0V, even after, say, placing a TVS across it!

The common mode voltage (and so, clearance from traces in the group, to unrelated traces) remains as it is, of course.

Even so, there can be some really sketch things... like plain TO-220s rated for 1500V.  You can stagger the pins on the PCB (if your assembler doesn't mind a leadform operation), but what the heck is up with the pins entering the body of the device itself?  Who knows.  (Some devices have a modified package with extended molding around the middle pin, or slots, which neatly solves that issue, when you can find them.)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf