Advertisement

Analog Angle Blog

Giving Connector Contacts Adequate Consideration

Any engineer with even a little practical experience knows that interconnection cables and connectors are a necessary and important element in the reliability picture. They are an all-passive, totally analog component of the analog and digital system. Interconnects which are improperly installed, poorly tied down, have marginal strain relief, are flexed too often, or run in a too-tight bend radius can lead to outright system failure.

Even more frustrating, they can be the source of hard-to-catch intermittent operation. Worse, in many cases you may not be able to say with certainty that you have actually identified the problem; instead, it may have gone away for now, perhaps because of something you did, or perhaps just due randomness.

For these reasons among others, designers often spend a large part of their project time selecting the best cable and connector to use. The selected pairing must handle the frequencies and power levels, and also the mechanical demands such as lockdowns, jackscrews, or retaining clips. (Of course, in some cases, an industry standard or common practice defines these factors, but in many cases, the designer has a wide range of possibilities and is not constrained.)

Still, it’s easy to study the cable and connector body, but what about the critical electrical contacts within the multipin connector? They often don’t get much attention, other than to make sure they can handle the current and frequency. Yet that’s a very limited view of the situation.

How so? Check out the clear and well-written article “Contacts for Hi-Rel Connectors: Comparing Technologies” in the November 2017 issue of Tech Briefs. Written by a specialist at high-reliability connector vendor Harwin PLC, the article clearly identified the different types of contacts offered for various connectors and the key attributes of each, both pro and con.

For example, the twin-beam contact, Figure 1 , offers two points of contact to the round or square mating pin for redundancy, and is a good choice for shock and vibration settings. However, if used with an oversized or misaligned mating pin, the twin beam contacts (usually made of phosphor bronze) can become deformed and get a permanent “set” that can actually impede continuity.

Figure 1

The often-used twin-beam contact offers a combination of reliability, ruggedness, and overall performance, but is somewhat more costly than simpler configurations. (Image source: Harwin PLC via Tech Briefs)

The often-used twin-beam contact offers a combination of reliability, ruggedness, and overall performance, but is somewhat more costly than simpler configurations. (Image source: Harwin PLC via Tech Briefs)

Another commonly used contact is the tuning-fork configuration, Figure 2 , which is low cost since it can be fabricated using a pressed-metal process. It is used with square mating pins on connectors such as those compatible with the PC-104 standard. The downside is their relatively low spring tension makes it a lesser choice for vibration environments.

Figure 2

The tuning-fork contact design offers performance which is almost as good as the twin beam in many cases, but at a lower cost; here it is shown as part of an insulation-displacement connector assembly. (Image source: Meritec)

The tuning-fork contact design offers performance which is almost as good as the twin beam in many cases, but at a lower cost; here it is shown as part of an insulation-displacement connector assembly. (Image source: Meritec)

These are just two of the contact configurations which the article discussed. Also covered were the single-beam, circular-stamped clip, circular-turned multi-fingered, spring, spring-loaded, and hyperboloid contacts. This article had a very high ratio of “useful information/time spent reading” than many other technical articles I have read on various topics.

It also cleared up an area where I was a little fuzzy in my internal understanding. Several months ago, I worked on a project related to connectors, and found that nearly all are made to order rather than stocked as standard items. Distributors of connectors carry relatively few finished ones in stock, except for the most-widely used, very common varieties. The user selects the connector body type, number of contacts, contact type, contact finish, and other parameters, and the vendor then provides these with very short lead time for modest quantities.

This makes sense, as the number of connector combinations with all the available options, though not infinite, feels close to it. No one can carry all these as stock items, of course. But it also means that there’s a subtle risk: if the exact connector/contact combination is not available with short-enough lead time, someone may be tempted to “temporarily” use a combination that is.

As a consequence, the product may fail in-house tests or even field tests, due to that Bill of Materials (BOM) substitution. The bigger danger is that the units will work fine for a while, and procurement will forget about the need to get the right contacts. The field failures accrue and designers will need to spend many hours trying to figure what happened and why.

The lesson is simple: spend the needed time selecting the right connector and contact type, and be sure it is actually used. If you can’t, at least fully document that the contact being used is not the one that is preferred, and why.

Have you ever had factory or field problems due to choosing the right connector body but using wrong contact type with it, or due to a subtle switch to what you called out on the BOM?

<>Related Articles

  1. High-Speed connectors increase speeds, keep their cool, and contain EMI
  2. Give Me Back My External Wi-Fi Antenna Connector, Please
  3. Battery and connector diversity: Sensible design, or casual engineering?
  4. Obscure AC line connectors challenge designers, consumers
  5. Don’t neglect the humble battery connector

2 comments on “Giving Connector Contacts Adequate Consideration

  1. AubreyKagan
    December 23, 2017

    Bill

    Thanks for the pointer to the connector review. One of the best kept secrets in the connector world is the number of make and break cycles that can be made before the connection starts to degrade. The article keeps the secret. Even when I worked for a connector company it was impossible to get official confirmation of rumoured numbers. I might add that those rumours a probably far less than one imagines.

    I have written two blogs on connectorswhich I hope are of interest

    “Which PCB Connectors Are Best?” 

    (h t t p s ://w w w .eetimes.com/author.asp?section_id=36&doc_id=1320698)

    and “Unusual connectors”

    (h t t p s ://w w w .embedded.com/electronics-blogs/without-a-paddle/4439649/Unusual-connectors)

    remove the spaces to get to the URLs

  2. davebirdieee
    December 27, 2017

    Years ago, when I worked for GE Computer Division, we investigated  the requirement to have a hot plug capacity. This allowed insertion/removal of system boards while the power was on. 

    We found, after some in house tests, that our existing board connectors pitted when inserted/removed with power on – no surprise actually. But, what was more problematical was that after a relatively small number of cycles, the connection failed completely. After feedback with our connector vendor and some design changes, the problem was remedied.

    Many years later, in a situation where the hot plug functionality was again required on a system under design (different company), this historical fact was replayed in a meeting involving various contributors. The system design manager “looked” into it by asking the connector vendor what they thought about the issue. They replied that they had never heard anything like that before. Our system design manager decided that it was not an issue, and proceeded to utilize the existing connector without futher investigation. He just didn't want to deal with it.

    Who knows whether the issue became a reliability problem or not. Chances are, that it was OK, or at least would never have been discovered. But, it does go to reinforce the unpleasant truth that way too often, potential design issues are skipped over, rather than actually delt with.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.