Advertisement

Blog

Extreme Analog Design: Don’t Forget Those Passives

Those poor passives — being what they are, they often don't get the consideration and respect they deserve. It's easy to worry about the performance and reliability of semiconductors for rugged designs at temperature and vibration extremes, but passives can easily be overlooked. After all, what can go wrong with a resistor or inductor, compared to an IC?

(Source: Vishay)

(Source: Vishay)

Lots of things can go wrong. First, consider the number of passives associated with analog circuitry. My experience and anecdotal evidence show that a digital part of a circuit typically has two passives per IC (usually bypass capacitors). The analog circuit spanning DC through RF (plus power supply) usually has at least five and often up to 10 passives per IC. Further, these are usually a mix of capacitors, resistors, and inductors, plus specialty devices such as circuit-protection components.

Of course, what's extreme depends on the situation. We most frequently think about operation at high temperatures and sometimes low temperatures as the problem, but there are other things to worry about. Vibration is second to temperature, because it can fracture connections and internal structural elements.

(Source: Coilcraft)

(Source: Coilcraft)

But, wait, there's more, as those low-budget infomercials say. Some designs are challenged by truly unique conditions. Consider Coilcraft's recently announced AE425PJB inductor series. Not only do these passives have a suspended core design to pass 80g vibration testing/1,000g shock testing and operate at up to 155°C, but they also are compliant with NASA’s low-outgassing specifications, which are critical for operating in the vacuum environment of space.

Some applications are at the opposite end of the atmospheric spectrum. Vishay makes thick-film resistors, such as its RCWP series, that are highly resistant to sulfur-laden environments. You might expect such conditions in harsh applications such as mining, but they also exist near automotive exhaust systems.

Let's give those passives — resistors, inductors, capacitors, connectors, wire/cabling, and others — their due respect, considering the vital role they play in all circuits, especially analog ones. Perhaps they finally are getting some of the recognition they deserve: the European Space Agency recently held the first International Symposium on the subject, Space Passive Component Days.

What sort of extreme design challenges have you faced? Did you use a component-by-component approach, or did you just put the whole thing in a box and try to protect the complete assembly?

Related posts:

14 comments on “Extreme Analog Design: Don’t Forget Those Passives

  1. RedDerek
    November 22, 2013

    Over my years, I have not had too much extreme design challenges. But to summarize…

    • Military – surface vehicles and aircraft and ship board
    • Space
    • Down-hole applications
    • Commercial use
    • Helicopters

    Most of the time, it was a team effort and others look over the design and question about temp, altitude, humidity and vibration.

    As a one-man show with many helicopter applications, it is now down

    • design
    • Critique
    • re-design
    • re-critique
    • repeat until satisfied

    Hardest part is the critique step where all things need to be concidered – especially vibration now.

  2. D Feucht
    November 22, 2013

    My laboratory is not in an air-conditioned environment but is subjected to tropical jungle air which can frequently be at 80 degF and 80 % RH. What fails the most (and failures are not infrequent) are integrated circuits, not passive components. I suspect it is caused by water vapor penetrating the glass-passivated circuits at the seals around the pins.

    Humidity is another environmental parameter for which better protection could be sought in component design. The future population of humanity is largely found in the 40degN-40degS zone around the equator.

    Thanks for drawing attention to passives, Bill.

  3. amrutah
    November 23, 2013

    Bill, thanks for the post on passives.

       I have been designing crystal oscillator for the RTC on chip and as you mentioned the passives on the chip (every passive device) is to be treated with utter importance as it will define the life of the clock/timer.  The passives define the frequency variation (jitter characteristics), power drawn.

       Choosing an appropriate crystal with its parasitic passives is very important

  4. Davidled
    November 24, 2013

    I wonder what type application require for vibration in the passive component. Did any passive component define the vibration requirement? For example, vehicle requires a lot of Electronic module. In driving condition, there is a lot of vibration and rattle. Hence, the bracket is holding the electronic module to prevent the vibration and rattle. This is one of the limited experiences.

     

  5. eafpres
    November 25, 2013

    I have had many, many problems with coaxial cable & connectors.  In my work with antennas for various in-building applications, as well as some consumer electronics, there were severable classes of issues.  Many times the cable connected to the antenna was soldered to a PCB or metal element or both.  It was easy to damage small RG-174 etc. coax when soldering.  The next class of issues was around the connector.  Many times it was hard to control the crimp of the ferrule, which mechanically holds the connector by crimping the braided shield against the connector body.  This was such a problem that in many cases we had to do continuous sampling and pull testing.  

    Another class of issues was that in many cases, the cable entry to an antenna housing had to be sealed.  The cable jackets are frequently not conducive to sealing, either they aren't bondable with common adhesives and things like O-rings often could not create a good enough seal..  

    Temperature extremes are also a common challenge for cable.  In one project we had all these problems together.  It was automotive, and the antenna, used to communicate to tire pressure sensors.  The customer mounted the antenna very near the exhaust pipe on the undercarriage  of a car.  The specs were -40C to +125C, IP67 water resistance.  We had to use a PTFE jacket on the cable to meet the temperature spcs.  The PTFE would not seal, even using an RTV silicone.  Eventually we had to treat the jacket with an acid to prime it so the silicone would bond.

     

  6. jkvasan
    November 25, 2013

    Bill,

    Nice post.

    We had the problem of high frequency signal creeping in in a Electro-surgical Cautery. The activation switch is in the handpiece and the device is supposed to only sense the switch closure and not the high frequency itself (one arm of switch is high frequency signal). We used appropriate Ferrite beads from Wurthe to suppress this interference.

  7. amrutah
    November 25, 2013

    @DaeJ:“wonder what type application require for vibration in the passive component. Did any passive component define the vibration requirement?”

    I think the MEMS capacitors are one good example where a mechanical stress (or vibrations) is used to generate equivalent electrical signal which can be further processe.

  8. samicksha
    November 26, 2013

    Thanx Jay, you keep good comment here, yes it can be added to an inductor to improve, in two ways, its ability to block unwanted high frequencies which are considered to be noise.

  9. PCR
    November 26, 2013

    Thanks Bill for the paradigm shifting article. 

  10. Victor Lorenzo
    November 27, 2013

    Hello All,

    Thanks Bill for the article.

    Yes, some times we overlook the real importance of some parameters from those 'tiny' passive components and just concentrate on the 'big' diigital components.

    A couple of years ago we designed one portable device that included one ISO15693/ISO14443A RFID reader for securing accesses. As of course housing size was a constrain, we had some difficulties in the design for achieving the intended reading distance (>7cm), mostly due to antenna size. The device was designed so it could stand for more than 12 hours inside a freezer at 4ºC and included conformal coating so it could also withstand over 80% condensing RH when taken out to ambient temperature.

    Everything was fine with the first production batch (at that time the designers were in charge of component stockage before phase-0 production)… but the devices from the second production batch (real production) had a 'small' problem, they were not able to read the RFID tags just after being taken out from the freezer.

    The root cause for it was, as lead time for some capacitors was too long, in the production department they decided to replace them with some others of same package, tolerance and value (and of course with a very lower price), but they did not pay attention to that small parameter called 'temperature coeficient'. At the end we needed to replace the affected capacitors in all devices from the second batch (fortunately not too many, around 150 devices in total).

    –Victor

  11. Victor Lorenzo
    November 27, 2013

    Good point, and ferrite beads not only contribute to reducing HF noise which affect other parts of the circuit, they also contribute to reduce or minimize the conducted high frequency signals (conducted emissions) that could lead to serious EMI compatibility problems.

    –Victor

  12. samicksha
    November 27, 2013

    You are right Victor, and to add at high frequencies, ferrite bead inductors work like resistors instead of inductors.

  13. jkvasan
    November 28, 2013

    @Samicksha,

    Selecting the right ferrite bead for the application does matter as there are plenty of ferrite bead variants.

  14. yalanand
    November 30, 2013

    If we compare circuit with passive elements like capacitors , resistors, inductor etc then there are many many parameters to consider which may go wrong with different conditions. That is the reason analog circuit design with these elements is complex than digital design.

Leave a Reply