Home    Bloggers    Messages    Polls    Resources   
Tw  |  Fb  |  In  |  Rss
Ed Rodriguez

Cree LED Bulbs: Good & Bad News

Ed Rodriguez
Page 1 / 4   >   >>
Adrian
Adrian
3/23/2014 3:27:54 AM
User Rank
Photon
keep it simple
You may be right about the US regulations. Yes, I have made interpretations for the sake of this comment as my interest reside in Europe. 

As for fuses rating I preffere to keep it simple and choose them using rule of thumb: 7mm MOV -> 3Amp fuse, 15mm MOV -> 5Amp fuse, etc. To do it state of the art is way too complicated and at the end of the day it counts just the result: pass or fail; it is a cheaper approach to over-rate the fuse and get rid of all headaches. Returns given to failures in real life are another story based on too many other things.

Q: do you have an indea of how the [...] did they measured the total lumen output @25 deg Celsius for a direct AC light engine? Cause I have managed to measure just hot lumens i.e in the real life/luminaire.

Adrian

50%
50%
semiman
semiman
3/22/2014 5:03:07 PM
User Rank
Candle
Re: the nylon screw
It's easy on Metal Core too, just means you have to pick the right metal core or you will never pass.

50%
50%
semiman
semiman
3/22/2014 5:00:54 PM
User Rank
Candle
Re: expectations
Ballast or driver shall comply with ANSI/IEEE C62.41.1-2002 and ANSI/IEEE C62.41.2-2002, Class A operation. The line transient shall consist of seven strikes of a 100 kHz ring wave, 2.5 kV level, for both common mode and differential mode.

There is no specific requirement written into Energstar w.r.t. a combination wave. You are perhaps making that interpretation from C62.41.2, but keep in mind this is a recommeded waveform document, not a testing standard ... and you need to reference table 5.

Of note as well, category A for the ring wave is 30 ohms, not 12 ohms. 12 ohms would apply if a combination wave was used.

Now that said, it is quite possible the manufacture has tested with a combination wave in order to reduce returns, however, that is not called out in EnergyStar.

 

The clamp rating of the MOV does not define the fuse or should I say does not need to define the fuse. The surge rating of the fixture defines the MOV. The 1200A clamp current of the MOV does not define what the fuse does. The I2T of the maximum supported surge waveform defines how much energy will be applied to the fuse and where the fuse does not interrupt ... whether the fuse is fast or slow blow. The 1200A could apply to the interruption rating, but such is not the case. The 1200A surge rating is pretty meaningless if the unit will never see a 1200A surge. 

A fast acting fuse may still be used depending on other design constraints.

Since you referenced the LittelFuse design guide, please go to page 6 where it says pick a fuse with an I2T rating that will pass the EnergyStar ring wave test.

 

For 10KA surge, it all depends on how you rate it, single strike, model code, 61000-4-5 and that will determine how big the MOV is. 6.3-7A fuse may not be big enough for a 10KA surge. 

61000-4-5 defines testing methods and number of strikes but defers to other standards for specific testing regimes, i.e. IEC\EN61547 in jurisdictions and for suppliers that use it.  Many would say the number of strikes of 61000-4-5 is excessive for lighting and that fewer higher energy strikes would provide better protection in many cases.

 

Semiman

50%
50%
Adrian
Adrian
3/22/2014 4:05:32 PM
User Rank
Photon
cheap design
I hereby confirm, it not MCPCB and yes, it have thermal vias.

Given the colour of non solder mask masked parts, it is FR4.

Adrian

50%
50%
Adrian
Adrian
3/22/2014 3:59:32 PM
User Rank
Photon
Re: the nylon screw
Semiman: it may be FR4 + thermal vias.. however it is avery low power thing [9.5W]

I am wasting my time playing with its bigger brothers [80W+, direct AC too] where it is mandatory to be MCPCB.

I know, assumption is the best way to failure, I have just assumed it is MCPCB. On FR4 the EMI/EMC compliance is a joke. Do not ask me how it is on MCPCB!

Adrian

50%
50%
Adrian
Adrian
3/22/2014 3:49:08 PM
User Rank
Photon
expectations
Semiman: I have expectations [from you]!  So:

- the transient test is part of EMI/EMC tests and is done according to IEC61000-4-5

- Energy Star Program requirements for integral LED lamps [Based on IEEE C.62.41.2] Class A operation, require testing on ring wave 2.5kV, 6kV 1.2/50us, 0.5kA 8/20us with a source having internal resistance of 12ohm.

- the MOV is selected to withstand the max surge current, and for the above class A it's diameter have to be 7-10mm: the fuse is selected to support the amount of energy clamped by the MOV, the so called i2t. The 7-10mm MOV require a fuse [delay type!] rated greater than 3A. Or the fuse will blow at the first spike clamped by the MOV!
The [1.2kA] rating ot the MOV is EXACTLY selecting the fuse rating. The fuse have NOT to blow while the MOV clampe a spike. This is the reason why delay type fuses are used and NOT fast acting fuses.

- e.g. on outdoor Class C High locations, the MOV have to withstand 10kA and therefore its disk diameter is 25-30mm and the pair fuse is rated above 6.3-7A

Yours

 

50%
50%
semiman
semiman
3/22/2014 3:34:07 PM
User Rank
Candle
Re: the nylon screw
Now, the nylon: the thing, as a 2 wires light engine have to withstand some 4kV pulses between its metalic parts and any of the grid vires [L or N]. As you know, the air dielectric rigidity is some 1kV/mm so the design[er] taked this into acct and moved all traces not closer than 4mm to the edges or conductive things connected to the heatsink; 

 

Given it's residential, I would pretty much guarantee that the surge voltage rating is only 2KV. Hipot test will be 1200VAC per UL1598.

Now that all said, I had assumed this was a metal core board, but zooming in, it looks like there are thermal vias around the driver section so perhaps not metal core. Thoughts?

Semiman

 

50%
50%
semiman
semiman
3/22/2014 3:23:12 PM
User Rank
Candle
Re: Shifted or just different
Excellent call on the Exclara. I got on the computer and zoomed in. Exclara is the only one using the QFN18 package that I am aware of so i can't think it would be anyone else.

 

50%
50%
semiman
semiman
3/22/2014 3:12:45 PM
User Rank
Candle
Re: Shifted or just different
You will notice use of "wiring" ... and actual wires outside the fixture would be fused.

Not specifically EMI, but surges, and potentially EFT, at least from a classification rating.

The 1200A surge rating of the MOV is not the specific spec that pertains to the fuse. It is more complex than that. Those small fuses have fairly low interrupting current ratings (well below 1200A).

What comes into play is how the fixture manufacturer rated the fixture for surges (if they have). North America has no specific standard so they could have just picked one arbitrarily or they could have used an international spec, 61000-4-5 just for the surges and/or 61547 at a fixture level. That would determine the test waveform and interval from which the I2T rating of the fuse could be evaluated to ensure it does not blow from surges.

Can't make it too big either of course.

Semiman

50%
50%
Adrian
Adrian
3/22/2014 3:03:25 PM
User Rank
Photon
the nylon screw
Ron: given the size of the board and the 3 lateral screw fixing, there is needed a central screw too. This is to prevent the board to bend in the middle.

Now, the nylon: the thing, as a 2 wires light engine have to withstand some 4kV pulses between its metalic parts and any of the grid vires [L or N]. As you know, the air dielectric rigidity is some 1kV/mm so the design[er] taked this into acct and moved all traces not closer than 4mm to the edges or conductive things connected to the heatsink; HV resistant solder mask allowed them to decrease the distance between components/traces and edges to less than 4mm; the 3 side crews are wery well positioned from this perspective.

If the central screw would have been metallic, than given the very small distance between his head and the surrounding components, the light engine would have not resisted to 4kV tests. So, they correctly used a non conductive screw.

Adrian

50%
50%
Page 1 / 4   >   >>
More Blogs from Ed Rodriguez
Gentlemen, start your engines.
When does it make sense to reduce emitter size to get more "punch"?
A guide to the appropriateness of Seoul Semiconductor's AC LEDs for high-bay applications.
flash poll
follow us on twitter
like us on facebook
All LED Lighting
About Us     Contact Us     Help     Register     Twitter     Facebook     RSS