Electromagnetic compatibility, come and see | Hacker News Reader

Electromagnetic compatibility, come and see

helentronica.com

98 pointssohkamyung26 comments

In metal shielding cases, slots, not holes, are the big problem. Slots are long openings and leak on wavelengths up to their length. Small round holes leak well up in the gigahertz range, where high-power noise is less likely.

A high-powered switching power supply, like this guy is building, is the worst case. It's inherently a high-power RF emitter. Switching power supplies are spike generators, so they emit lots of high harmonics. Of course this guy has an EMI problem.

I've run into this with a device I'm building that's powered from a USB port and has a switching power supply. The switcher runs around 300KHz, and I can see noise spikes down to 25ns, which is 40MHz. Keeping it from injecting that back into the USB power end is tough. If the device is powered from a computer, that's likely to cause trouble. USB ports usually have current limit detection, and it's instantaneous current that matters. Exceed 500mA for a microsecond and the port shuts down until power cycled. That's good; port shutdown occurs before the computer crashes due to noise on the power bus. Right now, I'm waiting for some 6.8μH ferrite beads to arrive. 1μH wasn't enough.

The gold standard for EMI testing is to test outdoors in a big, open field far from any RF emitters. Even for some IoT devices, an outdoor test range is used for the final check.[1] Here's one for megawatt-range power gear.[2] For outdoor testing, the unit under test sits on a wooden or Fiberglas turntable, all alone, several hundred feet from anything else. Now you can make far-field measurements. RF anechoic chambers are used for convenience, so you don't have to go out in the boonies for test runs.

[1] http://www.echelon.com/assets/blt2781ab44cb62fe72/PC_BOARD.P... [2] http://incompliancemag.com/article/lessons-learned-from-the-...

It is actually impressive how wide-band disturbances are. A hard-switches MOSFET can easily give you disturbances with bandwidths greater than 1 GHz. Closing a switch on a mains circuit can create high-energy wide-bandwidth disturbances in the power line, easily coupling everywhere.

Yeah, the slots were odd; all SMPS I've ever seen had perforated metal for just this reason.

Ah yes, EMC. If you've ever opened something up and seen lots of metal fingers round the edge of the case, or those flat metal cans on the PCB, that's because it wouldn't pass EMC otherwise. The ferrite rings are another common trick and you can sometimes see them on power or USB cables.

Multilayer PCBs also help a lot; it's not mentioned in this article but putting the 100MHz clock on an interior layer with ground planes on the exposed layers would be a good idea.

You do have to design EMC defence in from the start, rather like defensive programming. Sometimes people take the other extreme and cover v1 of the board in extra capacitors, ferrites, TVS diodes etc, then see how many they can remove while still passing EMC. This may explain unpopulated pads you see in consumer equipment.

One of the reasons cheap Chinese Alibaba gear is so cheap is ignoring these requirements. I have a car USB charger that has so much conducted EMI that you can't listen to the radio with it plugged in.

In many high priced low-volume systems you will find another approach: go to the extreme with EMI supression, pass testing and just leave it as is in the final product. Perfect examples are 90's DEC and IBM workstations which are full of completely redundant ferrite beads (I've even seen snap on ferrite bead on cable for power LED).

Obvious problem with that approach is cost, somewhat non-obvious one is that there can be cases when unnecessary EMI supression/shielding/whatnot can be actually detrimental. Part of EMI susceptibility issues on one customer's system were resolved by replacing 20cm run of shielded twisted pair with unshielded one, shielding formed parasitic capacitor with the overall chassis of the product which coupled noise from chassis ground into analog ground.

[Edit: also the fact that your product passes EMC testing says nothing about the EMC issues it will have in field]

Well, yeah: the cable shielding should have been bonded to the chassis at both ends.

I thought you should only ground shielded cables on one end to avoid forming a ground loop.

As said in other comments in depends on the nature of the ground. In this case it should have been bonded to chassis ground and not to the ground plane of analog board, which wasn't feasible due to mechanical constraints.

In general cable shielding should be connected to shielding ground of devices on both ends of the cable as long as the devices have purpose-designed shielding enclosure, which they often don't and in that case it should be connected at the end with lower impedance ground. Another problem are cables where the shielding is also used as return path for normal operation (RF coaxial cables, consumer electronics...) and in that case it obviously needs to be connected at both ends.

My rule of thumb for this is that shielding should be connected only on the lower-impedance ground end except when it is not obvious which side that is or the shield is also return path. Neat solution to the "non-obvious side problem" is used for IEEE1355/SpaceWire and related interfaces: outer shield of S/STP cable is connected at both ends while the per-pair shields are connected at transmitting side for given pair (used cable has isolation between shields).

Very interesting, thanks!

It's funny, part of what drew me to electronics and circuitry as a hobby was how predictable it is, compared to how sloppy chemistry and biology are. But when you get into high frequencies and EMI and wireless signals...woof.

It's really amazing how much work goes into making that sort of thing work like, at all. I have enough trouble getting a simple bit-banged serial or parallel interface to work without having to worry about occult ley lines.

It actually depends on the nature of your ground at both ends of the cable. The book by Ott, pictured but not referenced in the article, had a whole chapter on the subject if I remember correctly.

Also copper tape is wonderful but it will also cut the hell out of you so wear gloves.

Summarizing a number of application notes into one sentence: Generally speaking yes, except when you don't.

I'm pretty sure you're right...

Running an engineering house and dealing with emission/immunity issues all the time. Also have a small precompliance capability in house. This is a very good summary to the topic! I really like the pointers to the pitfails like "notice the vertical ground plane next to the LISN".

One thing to remember: if you see external EMC measures (shielding fingers, extra ferrites, copper tape, line filters) on the final product you can be sure the development has been too agile and regulatory compliance an aferthought. Happens way too often.

One thing with precompliance in the office environment to keep in mind. The "magic environmental noise substraction" on EMI receivers does not work that well if noise is stronger than the signal. Just pure statistics at play.

No mention of using a spread spectrum clock source. Nearly all modern clock generator chips have such an option. There is debate as to whether or not it actually improves EMI, but it definitely helps you beat the regulations.

Also, as a non-hardware person, a ferrite choke on the cable is the very first thing I think about for EMI just because it's the single most visible EMI reducing device in anyone's daily life (e.g. that annoying round lump on the end of your VGA cable)

Unbelievable amount of depth, well written, and I learned a lot. The bottom line is important too: If the EMC is not taken into consideration from the beginning of the design, it will be a devil’s job to make it work in the end. Chances of being able to fix EMI when your product is finished will drop closely to zero, while the costs of doing so will skyrocket.

First step is a doozy.

Yes, you do need to take it into account, but even so, there are no guarantees you _are_ going to pass the first time around. EMC is as much dark art as it is engineering, and some things you simply cannot predict up front. We always tell our customers we do the best we can up front, but to expect EMC issues nonetheless.

It may be worth adding that EMC compatibility is only one step in getting your product ready for market introduction. Depending on the category of the product (consumer, medical, automotive, industrial, high-power, wireless etc etc) there may be many more tests you need to pass and many more standards you need to adhere to before you are allowed to sell your product.

Henry Ott's /Electromagnetic Compatibility Engineering/ [1] is a really good book about the topic. Also mentioned in the article.

[1] http://www.hottconsultants.com/EMCE_book_files/emce_book.htm...

Great article! Going to forward it to my manager (non EE) who I keep denying requests for MILSTD-461 testing.

The switching supplies ICs (with integrated FETs; like from Linear Tech, TI, etc) are noisy where you might not think. Extremely fast rise times on the FET generates white noise through 2 GHz, even if it's only running at a few 100 kHz. Putting wrong bypassing/filtering parts on can only make it worse as they will act like an antenna. It may not even show up in a EMC test but enough to degrade RF receiver sensitivity a few dB.

Lots of interesting info in here! I have a DSL modem that smells like burning plastic at all times, so I put it outside in a metal box. I had wondered about the extent to which it diminishes the signal in the house. This passage helpfully explains:

"[...] any opening, or slot, on the shield reduces its effectiveness. The slot behaves as antenna, with the same radiation pattern as a wire of the same length."

More like wire of the half length (monopole) with good reference. Slot is pretty much equal to a dipole and resonates best at double the wavelenth of its longest axis.

This assumes single narrow and straight slots. Quite fancy radiation patterns, polarization and resonance points can be implemented with fractal shapes of slots.

It's amazing how many products cut tiny corners which end up resulting in a lot more EMI than would otherwise be necessary.

One other phenomenon I've learned about recently is that some switching power supplies can create all sorts of weird harmonics when in the presence of strong RF fields. Ideally you can test your device to eliminate such behavior even if its own generated emissions are to spec already.

I'm a bit confused. Can someone point out my error?

- device gets tested, fails

- temporary fixes and "hacks" on top of the product were improvised

- device passes

How is the CE conformance of the real devices endured? The device under testing was a one off with lots of tape, right? Do you need to come back at some point with your final, straight off the manufacturer's line device? If not, what is certified?

CE isn't certification but declaration by the manufacturer that the device complies with relevant regulations. Testing by independent laboratory only serves as source of paperwork that limits manufacturers liability should somebody want to contest that declaration.

In the not so distant past I was working getting rf pumped light emitter engines through a variety of EMC tests. All of this rings very familiar.

SMPS I deal with now are much less challenging to certify, by comparison.

"gravy silence" :0 fun to read, thanks!

Reply on news.ycombinator.com