EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: PG_LAB on August 10, 2023, 05:32:43 pm
-
Hi,
I'm in process to obtain a CE certificate for a small DC 12V device.
According to the certification lab I have to pass IEC 61000-4-5 test with 500V and 250A surge impulse.
Please note that my device come without any AC power adapter.
I was wondering if any of you can point me how to pass this test. Should I use a varisistor, TVS or other? Any schematic it will be really useful.
I would also like to understand how to properly dimensions the varistoror/TVS.
Any suggestion?
Thank you.
-
12 V input>>CE falls on the mains PSU not your device.
Contact a lab like TUV, ETL for CE requirements for SELV low voltage devices.
Testing requies a very costly test machine and skilled techs.
Jon
-
Just to give more information. My device is for home automation and should be installed
in the house switch board.
According to the lab, my device should be under the following standard IEC 63044-5-2:2017.
Home and Building Electronic Systems (HBES) and Building Automation and Control Systems (BACS) - Part 5-2: EMC requirements for HBES/BACS used in residential, commercial and light-industrial environments
At page 11 of this document is present a table with EMC immunity requirements for DC port.
In particular it's needed:
IEC 61000-4-6 (Radio-frequency common mode),
IEC 61000-4-4(Fast transients-bursts),
IEC 61000-4-5 (Transients surge)
About the last test, there is a remark on the table and it seems only needed when the power cable are longer more than 3m.
It's not my case to use such a long power cable. But the lab suggested that my device should pass this test in order to be
more robust.
If it's just a question to add the right TVS/MOV, I agree with the lab.
I did my research, and it seems that other 12V/24V devices are been certified for this IEC 61000-4-5.
For example Arduino Opta PLC. See the datasheet https://cdn.findernet.com/app/uploads/S8AEN.pdf (https://cdn.findernet.com/app/uploads/S8AEN.pdf) at page 5.
Surges (1.2/50 μs) on Supply terminals common mode EN 61000-4-5 4 kV
If I'm right, Arudino Opta should be able to stand a surge of 4000V and 2000A
In conclusion, I was wondering if some one can point me out how to dimension correctly the TVS/MOV in order to pass the test.
Thank you.
-
Surges (1.2/50 μs) on Supply terminals common mode EN 61000-4-5 4 kV
Hello
if its only "common mode" (both pins equally against protective earth) it is only a matter of isolation distances against environment.
Differential mode would be much more difficult.
But I am unshure wether the 2 Ohms would apply in this case.
with best regards
Andreas
-
My fault, I miss to report the full test of Arduino Opta on the datasheet.
Surges (1.2/50 μs) on Supply terminals common mode EN 61000-4-5 4 kV
Surges (1.2/50 μs) on Supply terminals differential mode EN 61000-4-5 4 kV
Surges (1.2/50 μs) on input terminals common mode EN 61000-4-5 4 kV
Surges (1.2/50 μs) on input terminals differential mode EN 61000-4-5 4 kV
For my device,the test on DC port should be unbalanced transmission (I believe it's differential mode) of 0.5KV
-
Right, if you supply an adapter (and recommend only same or equivalent types for replacement in the manual), you only need to test the supply. If not, you need to test the device itself, which is rough. (Give or take the specifics related to building automation; sounds like that's a bit more intensive in general.)
DC surge immunity isn't a big deal I'd say: ensure adequate connection thickness to a TVS, use a TVS rated for such peak current, and design the following circuit to withstand whatever peak voltage it will develop in the process. 8/20us is pretty brief so the energy isn't too bad, but the peak currents can be quite high as you can see, and even just adding some resistance can help out a lot here. Or if you're drawing enough current that plain resistance is impractical, even a few uH can help out (assuming it doesn't saturate -- air core required).
TVS diodes are fairly often rated for a 8/20us waveform, and at quite impressive currents at that, but mind the voltage drop is also quite high for the same reason.
Or capacitors of sufficient value, and low ESR. The surge generator has about 10uF inside it; if you need to swamp 1kV down to a more modest 10V, you need at least 100 times more or 1000uF at the input. Which is a pretty modest value at low voltages, and a design might already include such for mains ripple filtering or power interruption hold-up time.
Incidentally, 8/20us is the low impedance version of the "combined waveform generator", which is 1.2/50us when high-Z loaded. When you see the latter figure, it's probably at much lower current. This may imply, say, a 22 or 42 ohm test for your case. This is usually the case for long-distance comm wires (e.g. TNV type), where the line resistance is significant, but due to its length, it is subject to lightning induced transients.
You'll have to check the standards in more detail to see which is the case.
Tim
-
You can easily pass these surge test with addition of a TVS diode for example Vishay SMC tranzorb series. For dimensioning you need to calculate peak pulse power with the parameters from your EMC standard. There are application notes for this. But I have done similar test on 24VDC bus systems with long lines and passed all these surge test with a 1500W peak pulse power Transzorb