Clarifying the misconceptions of filter connectors
Filter connectors are a misunderstood product for many components engineers today. The lack of understanding is the result of the fact, that filter connectors are seldom on the ‘what’s new’ headlines and because most engineers don’t think they need them – until they get to the EMC phase of testing. For those who do, they will be glad that they are available. (*See Figure 1.)
Filter connectors are a tool in the world of ElectroMagnetic Compatibility (EMC). Therefore, only those designers who’s responsibility falls in EMC, will have need for this possible solution to their problems. Back when the problem was called EMI, because interference was the big issue, some of you might remember 1983, when the FCC Part 15 rules came out, and most electronics manufacturers (especially the computer folks) scrambled to figure out how to comply, or else risk not selling their product. Now, the broader term EMC is used and applies to every product, in all industries. So, depending on who you work for, in addition to FCC (part 15), you will encounter EMC requirements based on MIL-STD-461, CISPR, FAA (AC20-136), and others from NASA, Boeing, Airbus, a medical device or an automobile manufacturer. That’s just the North American standards. Seek out the European equivalents and you will have a pile of CE directives hidden in all those IEC specifications.
EMC is a two-way street, in that the designers must ensure that their product does not interfere with other electronics, and it must also tolerate some specified level of exterior RF. Thus the concepts of interference and susceptibility were rolled into one and called compatibility….hence the EMC moniker. Controlling EMC really boils down to three methods: shielding, filtering, and most importantly minimizing the energy and frequency spectrum that needs to be contained.
The proliferation of personal electronic devices (tablets, smart phones, E-readers, GPS, etc), has been a two sided coin. The compact handheld devices are seldom connected to another piece of equipment due to wireless capability, thus reducing the input/output connectors that were so popular and the cause of a lot of EMC challenges. But, the abundance of such RF emitting devices, around equipment that IS interconnected, necessitates rigorous attention to EMC. You can quickly get an appreciation for this, when you consider medical electronics in hospitals, test or process control instruments in a factory, and the ever changing rules aboard aircraft. Then there’s all those new electronic devices carried or even worn by military personnel or emergency response workers.
Over the years, an extensive science has evolved, which helps circuit designers minimize their need for filtering and shielding, by careful design and layout of the components in their circuits. But, ultimately, all final designs will utilize all three methods
“Filtering simply means the control of the frequencies,
the energy that they contain and limiting where they go”
This article focuses on the filtering. Filtering simply means the control of the frequencies, the energy that they contain, and limiting where they go. Components that are frequency sensitive are capacitors and inductors. It’s easy to see, that filter connectors, will integrate capacitors, inductors, or both, inside a connector. Most often, the connectors of interest are the input/output connectors, since those are not only the transition from the device to the outside world, but by their nature, they are also the ‘antennae’ for concerning frequencies – both coming and going.
As a result, the most popular and most filtered types of connectors, are D-subminiature and their related families (which actually originated in MIL-C-24308), several circular types like MIL-DTL-38999, MIL-DTL-83723 and aviation industry specific ones like the ARINC 404 and 600 connectors. Typically, such use is for professional equipment, not consumer products.
The capacitor is the primary filter component, typically of a special construction that lends itself to fitting into the tight confines of a multli-pin connector. One lead is connected to a contact, the other to ground. The ground ultimately is connected to the shell and the device enclosure. While some filter connectors use simple ‘chip caps’ mounted on a substrate configured for the contact layout, the higher performance filter connectors use either tubular ceramic capacitors, or planar ceramic arrays. The very high end filter connectors contain multiple capacitors and ferrites to create a Pi type of filter.
Filter performance can range from a few dB (ferrites, chip caps) of attenuation, to 60 or more dB. when used in Pi or higher order filtering, in a properly shielded and grounded connector design. No exception here…you get what you pay for.
The key to a successful design, is to only pay for what you NEED. Understanding and defining your needs is the critical first step. Then choosing the methods available to meet the spec, and effective implementation, will provide a cost effective end product.
So what makes the use of filter connectors desirable? If you have to use just one word, it’s performance. They work. Key reasons why they offer advantages:
1) Filter performance far exceeds comparable filters placed elsewhere in the circuit. There’s a lengthy engineering explanation of why, but a 60dB. Filter inside a connector, might only provide 15 to 30dB if placed somewhere on the printed circuit board behind the connector. While not all designs need 60dB, some modern electronics do, and thus sometimes a filtered connector is the only EMC choice.
2) Higher frequency effectiveness, while related to #1, occurs due to more compact packaging, better shielding around the filters and especially because of lower impedance ground connections. The rapid expansion of ‘wireless’ devices has made the need for high frequency control much more important, than it was even 10 years ago. Useful filtering above 1GHz. is common, and necessary.
3) Space. The drive for compact designs might welcome component removal from the pcb and hide them inside the connector.
4) Component reduction. A 25 position D-sub, when filtered, might remove 25 or more components from the pcb, when replacing an ‘on-board’ filter design.
5) Flexibility of EMC design & performance. A filter connector can easily be changed for one of different filter properties. Some filter connectors are ‘add-on’ adapters, which can be used on location specific installations.
(*See Figue 2)
Some have said that, the use of filter connectors indicates that other aspects of the EMC design process have failed. While true in theory, not all designs, not all markets, have the ability to fully utilize only the alternatives. There always will be many high performance, highly critical applications that simply need the best, need to choose what is known to work, and do not have the development time and cost to avoid use of a filter connector. Price has typically been a perceived downfall of filter connectors, but that must be balanced against the cost of avoidance, which is seldom trivial.
Filter connectors have a broad range of performance and prices. The most popular higher volume, commercial filtered D-subs, start in the $3.50 and up range. Deliveries can range from stock, to only 6-weeks. At the other end of the spectrum, special aviation types, like ARINC 600 can be $1,000, or more, and take 18 weeks. The shock factor of such ‘price & delivery’, will depend on what industry you work in, and what your cost alternatives
are to meet EMC.
There will always be a market for effective use of filter connectors, because designers will match their needs, and apply cost effective solutions that work. There are lots of $1,000 electronic boxes, and $1,000,000 electronic systems, that simply would not meet the EMC needs, without the use of a filter connector.
EMC is still a blend of science and black magic. The science content has moved forward a long way over the years since the first FCC regulations came into effect. Filter connectors continue to meet a specialized need in electronics design. Because each EMC situation is unique, designers must work with filter connector suppliers to optimize their selection. As a result, there are a select group of connector specialists, who do well in this market space, and it tends to be the smaller companies who choose to make filter connectors a focus product. This means they will provide superior technical support in the early consultation stage, use rigorous quality assurance methods in their manufacturing process, and give the designer the right product, at the right price, to meet his need.