• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Non-isolated, offline, 230VAC LED driver with LED MCPCB screwed to earthed heatsink

Status
Not open for further replies.

Flyback

Well-Known Member
Hello,

We have an 230VAC offline, non-isolated LED driver with a 300V LED bank.
The LEDs are on MCPCB, which is screwed down to an earthed heatsink.

Can you confirm that this is illegal?, since if the earth wire that connects to the heatsink was (for whatever reason) not secured, then the heatsink would be insufficiently insulated from live mains.

So….we are using the setup of Figure 3 on page 4 of the following App Note on LED MCPCB’s..
http://www.lumileds.com/uploads/539/AB06-pdf
 

Diver300

Well-Known Member
Most Helpful Member
If there is enough insulation to stop the mains, and the heatsink is earthed, then I think it is legal.

A single fault would not cause danger. If the earth wire came disconnected, there could be a bit of leakage, but if it wasn't enough to trip the ELCB with the earth connected, it wouldn't be enough to cause danger. Alternatively, if the insulation broke done, there would be fault to earth and the fuse or ELCB would blow.

The minimum spacing, suitable insulation gap etc. are all for non-earthed appliances.
 

Flyback

Well-Known Member
Thanks, the thing is, on page 9 of the following …

http://www.lumileds.com/uploads/539/AB06-pdf

…it states that there should be 2mm of clearance from an MCPCB copper trace and the board edge.

…The thing is, if the MCPCB is screwed directly to an earthed heatsink, then the MCPCB copper traces may just as well be right at the board edge, because wherever the PCB traces are, they are effectively only a few microns away from the earthed heatsink? (a few microns being the thickness of the dielectric insultion layer on the MCPCB).

Also, i dont think a few microns of MCPCB insulation is classed as being sufficient insulation from the mains?
 

Diver300

Well-Known Member
Most Helpful Member
OK, but it's a few microns that water and other contaminants can't get into. That's why you need the clearance to the board edge.

There a lots of components where inside them live connections get that close to earthed ones. Lots of semiconductors are rated to 1000 V or more, and class Y capacitors are tested to 8 kV. You can get 2-core, 1.5 mm2, 750 V MICC cable that is 7.9 mm diameter. That's only got about 1mm of clearance between cores or between the core and the sheath, and that can be coiled up.

Just use MCPCB rated to the voltage, and earth the heatsink. You would need two things to go wrong for there to be any danger. Add an ELCB and it is three things needing to fail before anyone gets hurt.
 

Flyback

Well-Known Member
Thansk, i assumed all MCPCB insulation was rated to 1.5kV AC?
OK, but it's a few microns that water and other contaminants can't get into. That's why you need the clearance to the board edge.
Thanks, sorry but i dont see it..a copper trace on an MCPCB is electrically separated from any point on or in the MCPCB by just a few microns, (the thickness of the insulation covering the MCPCB.)....so whats the point in the 2mm clearance?....the whole MCPCB is electrically conductive....whether a track is right next to the edge of the MCPCB , or 5mm away from the edge, it is effectively right at the edge.....because the entire MCPCB is made of conductive material (apart from the thin bit of insulation).
 

alec_t

Well-Known Member
Most Helpful Member
whether a track is right next to the edge of the MCPCB , or 5mm away from the edge, it is effectively right at the edge.
Not so. If the trace is spaced away from the edge then the path for arc-over includes the exposed surface (possibly-contaminated) of the insulator plus the insulation thickness, all in parallel with an air-space of unknown conductivity. That path is significantly longer than the insulator thickness alone. It is the air-space which is most likely to break down.
PCB.PNG
 
Last edited:

Flyback

Well-Known Member
Thanks, what if the track is covered in solder resist? Surely solder resist has just as much insulation resistance as the dielectric insulator on top of the aluminium of the IMS board?.....so surely solder resist covered tracks can go right to the edge of the IMS board?...(or at least to within 1mm)
 

alec_t

Well-Known Member
Most Helpful Member
It is the air-space which is the likely breakdown path, so covering the trace with solder resist should help. I don't know whether alumina (or whatever insulator is used) or solder resist has the greater insulation resistance, but my money would be on alumina.
 

Flyback

Well-Known Member
Thanks, i take it "alumina" is the name of the dielectric sitting on top of the aluminium of the IMS?
 

alec_t

Well-Known Member
Most Helpful Member
The dielectric layer may be proprietary. From what I read, alumina is only one possible filler material used in such a layer.
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top