Reliable Reliability

July 20, 2010

  With an IQ of 181, you might think it’s hard for me to amuse myself, but it’s not true. I like a little light reading in the evening…things like Kreck and Lück’s The Novikov Conjecture (A group G is a-T-menable, or, equivalently, has the Haagerup property if G admits a metrically proper isometric action on some affine Hilbert space, page 205) or Carver Mead’s Collective Electrodynamics (Because the system spatial attributes of the system in an eigenstate are stationary, the system in an eigenstate cannot radiate energy, page 106.)

 Fascinating, riveting stuff, I’m sure you agree.

 This led me to think a little bit about semiconductor reliability, though I confess my thinking might have been influenced by the consumption of a bottle of Puligny Montrachet Grand Cru Domaine des Comtes Lafon. In case you’re unfamiliar, that’s a wine very similar in appearance to the Franzia Crisp White wine you can buy in a box at the local market.

 I mentioned Carver Mead above. Among his many accomplishments, he was the guy who contributed to Moore’s Law (which says transistor density on an integrated circuit will approximately double every two years) when Gordon Moore worked at Fairchild Semiconductor in the 1960s.

 In 1965, Gordon had just started making his plots, where he’d plot the logarithm of the number of transistors on a chip as a function of the year. They’re little hand-drawn plots. I still have some around. One day we were talking about his plots. He said “You’re working on electron tunneling that happens when things get really small, right?” Yeah. “Well, wouldn’t that limit how small you can make a transistor?” Yes. “Well, how small is that?” Gordon has a way of asking these very simple questions that you really think you should know the answer to, and I didn’t. I said, well, I have to go and think about it. I’ve been thinking about it ever since. Carver Mead, from a speech at Telecosm 2006

 The driver for Moore’s Law is the fact that as we make transistors smaller, they get cheaper to manufacture AND they work better. Isn’t that something? They work better. Think about that. And, what do we mean by working better? They consume less power and they can switch more quickly. This is the small miracle that fuels the marvelous advances of the digital revolution.

 Today Fairchild is a leader in power management and mobile technology and in our factories, over 50 years after Fairchild Semiconductor was originally formed, we continue to take advantage of the state-of-the-art in transistor lithography.

 There are not many businesses like ours where customers expect continuous improvement in products-with lower prices every year. Generally, if you buy the bargain brand at the big-box discount store, you expect lower quality…fewer features and less reliability. If you buy an inexpensive car, you expect it to be less comfortable and less reliable than a premium brand. But, that’s not how the semiconductor business works. Regardless of what you pay for our product, you expect high reliability…robust and rugged components.

 Year-by-year, we practice continuous improvement and do our part to produce parts with very low failure rates. As a semi-random example, we publish a 3.65 FIT (Failure in Time) rating for the n-channel FET NDT3055. This works out to one failure for 3,127 years of operation.

 Let’s take a closer look at what this means. We can’t build up a significantly significant number of parts and test them for 3,127 years. We’d love to, but we can’t. The FIT rating is based on accelerated testing of sample parts with numbers plugged into a formula.

 The basic idea is that by aging sample devices with high humidity and overvoltage stress, we can estimate the failure rate without waiting for tens of thousands of years.

 As I mentioned above, we do our part to create robust devices for our customers. However, is there also a role for the customer? Of course there is.

 In building up our reliability estimate, we use a temperature stress factor adapted from the Arrhenius equation-which includes this term:

eEa/k(1/Tu - 1/Ts)

Where:

Ea = Semiconductor activation energy

k = Boltzmann’s Constant

Tu = use temperature (K), or the die temperature in the design.

Ts = stress temperature (K) used in the accelerated life test.

 We control the stress temperature and it is based on the semiconductor process, generally either 150 degrees C (423K) or 175 degrees C (448K). You control the use temperature. Lower operating temperatures result in higher reliability. That’s your part of the job.

 So, let’s say you want to increase the reliability of a system. Using the free MTBF tool referenced at the end of this blog, we can see the predicted effect of reducing the operating temperature from 100 degrees C to 90 degrees C.

 At 100C, the calculated FIT is 1009.

At 90C, the calculated FIT is 860.

Is that enough of an improvment? That depends on your needs.

Please note I am not saying anything about the fundamental reality of these numbers. They are numbers, but there is a fair amount of speculation and unconsidered factors in a real life design.

 It’s getting late and I notice there is about a half a glass left in the bottle. I suppose I could put the cork back in and save that last bit for later.

 Or not.

 The unpublished papers referenced below are available by request.

 The author would like to thank Thomas Welch and Raymond Oakley who collectively contributed three unhelpful suggestions and two rude comments during the writing of this article.

 References

On Acceleration Factors used in Failure Rate Prediction - Unpublished paper by Thomas Welch, Director, Quality and Reliability, Fairchild Semiconductor

Secrets of Mean Time Between Failure - Unpublished paper by Raymond Oakley, Staff Customer Quality Engineer, Fairchild Semiconductor

Free MTBF tool from Advanced Logistics Developments, Free MTBF Tool

Failure Mechanisms and Models for Semiconductor Devices, JEDEC Publication, JEP122E

Reliability by Design, A. C. Brombacher, John Wiley and Sons

Reliability, Maintainability and Availability Assessment, Mitchell O. Locks, Hayden Book Company

Collective Electrodynamics, Carver A. Mead, The MIT Press

The Novikov Conjecture, Matthias Kreck and Wolfgang Lück, Birkhauser Verlag

Tags: Dr. FAE, Fairchild Semiconductor, reliability, semiconductors

Would you care for an IC?

May 7, 2010

Instead of answering one of the many technical questions that cross my desk, I hope you’ll indulge me while I digress to tell you about the party I attended in Manhattan last week. Perhaps you saw pictures from this gathering…it was hosted by the widow of a moderately infamous oil mogul and many of the single-name-only crowd were there: Angie and Brad, Jewel, Cher, Sting, Jen, Björk, Moby and others, as were a random collection of B-list partiers…rappers, lobbyists, literary agents, fashion models….and the ever-present John Q. Fields-you know, the ubiquitous actor with the very high IQ (but not quite so high as mine, thank you very much).

The first thing you’ll want to know is how an engineer gets an invitation to a party like this, but I can’t help you. That’s a path you’ll have to blaze on your own.

After a stare-down with the bartender who tried to insult me with an exceptionally average Cognac when, from a prior visit to this expansive apartment, I knew there was a bottle of Cognac Dudognon Héritage HENRI IV hidden in a secret storage nook behind the hallway Matisse print (Woman with a Hat, or rather, an inept counterfeit). I was minding my own business by the fireplace when a pretty TV actress planted the heel of one of her Jimmy Choo slingbacks directly on my instep. After accepting her profuse apology, we struck up a conversation and she asked what I do for a living.

“I work for a semiconductor company…a famous one with a long history of accomplishment and invention.” But it had no meaning or relevance to her. She looked at me as if I was speaking in tongues-until I explained.

“Unless you’re an electrical engineer, you might not be familiar with Fairchild Semiconductor, but it’s highly likely that you own many of our products. For example, do you own a brand name flatscreen TV? Then, most likely, you own numerous Fairchild components. Do you have a game console? The most popular ones are built with Fairchild parts. Do you drive a car with a navigation system? Do you have a cell phone with a digital still camera? Do you have a DVR? A computer? An Energy Star® air conditioner or washing machine? A printer? A music player? We’re everywhere and you are our customer! Your life is probably overflowing with hundreds of our tiny, energy-saving, feature-enhancing components. You might not know our name, but we reduce the energy consumption of your computer, provide the backlighting for your LCD TV, increase the talk time on your cell phone, and generally power your world.”

As you might imagine, she swooned-and peppered me with questions about our support for soon-to-be-released cell phone features and enhancements, which I will share later.

Then, eavesdroppers quickly spread the word. Many stopped by to shake my hand and kiss my cheek…and thank me for our contributions to their way of life…for helping to create and power all the beautiful things that enrich and enable our fabulous modern lifestyle.

Snifters of the secret stash of cognac finally came my way, but I could hardly find a quiet moment to enjoy a sip.

Oh, my friends, celebrity can be such a heavy burden.

But, have no worry.

I can handle it.

I’ll be fine.

Tags: Dr. FAE, energy-saving device, Fairchild Semiconductor, feature-enhancing components

Zero ESR Filter Capacitors — Effect on Feedback Loop Compensation

March 16, 2010

Dear Dr. FAE
I did a good deed and replaced the output filter capacitors in my voltage mode control DC:DC power supply with a zero-ESR (Equivalent Series Resistance) versions. Now my power supply is unstable when the load rapidly changes and my boss is mad. Is it true that no good deed goes unpunished? — Frustrated in Fortuna

Dear Frustrated,
It’s such a simple problem… I’m constantly surprised at how people overcomplicate things. If you’re designing a feedback loop, then negative feedback is your friend-it contributes to stability. Positive feedback contributes to turning your amplifier into an oscillator. The feedback loop uses an inverting amplifier. That’s 180 degrees of phase shift. Unfortunately, there are other factors that add to this 180 degrees and if they contribute too much, it eats into your margin and moves closer to changing your sweet negative feedback into dangerous positive feedback.

Beyond mentioning voltage mode control, you haven’t given me any circuit details, but that does not trouble me. I see nearly all and know almost everything-it’s one of the key advantages of having a very high IQ (181, thank you for asking). You have a buck converter and it has an LC network on the output that looks like this:

Figure 1 - SPICE Simulation

If we pretend the capacitor ESR is zero…it’s a low pass filter and the -3db corner frequency is:

It’s a 2-pole circuit, so the slope of the response above the corner frequency is -40dB/decade. This is illustrated by the SPICE simulation result-note the characteristic resonant “peaking” of the output around the corner frequency (the solid trace) and the very rapid phase change (dotted trace) from zero to 180 degrees. That rapid phase change makes the loop difficult to stabilize. Any signal anywhere near that frequency quickly becomes positive feedback. Oh no!

As a rule-of-thumb, we’d like the slope of the response to be -20dB/decade when it crosses the zero-gain line with a maximum phase lag of 90 degrees (represented by a 1-pole circuit). If only there was a way to change the -40dB/decade response to -20dB/decade. For free. For once, we can use a parasitic element to assist with the desired response…the frequency response zero created by the capacitor ESR.

The corner frequency for the zero created by the RESR and CAP network is located at:

Beyond this corner frequency, the slope of the response changes from -40dB/decade to -20db/decade…that’s the slope we want when the gain crosses zero. Let’s see what adding a reasonable value of ESR (100 mOhm) in series looks like in the frequency range we’re interested in.

Figure 2 - Add Capacitor ESR

Notice: now the phase change is headed toward 90 degrees and not 180 degrees.

It’s great to reduce ESR. After all, the ESR losses represent wasted energy. So, all hail zero ESR capacitors (including ceramic capacitors). However, the power supply loop design must take this into account. Typically, this means adding a capacitor to the Type-2 compensation network to create a Type-3 network. Later, I will be happy to discuss the Type-3 network in infinite detail. Another way to simplify the feedback network is to use a current mode controller (which allows us to visualize the output as a current source and sort of removes the filter inductor from the feedback loop, but that’s another topic for yet another day).

I want to make sure to address all of your questions, even the off-topic philosophical ones.

So, is it true that no good deed goes unpunished? Yes, indeed. Sad, but true.

References:
Abe Pressman’s Switching Power Supply Design (Second Edition) McGraw-Hill
Optimum Feedback Amplifier Design for Control Systems, Timothy E. Biesecker (http://www.venable.biz/tp-03.pdf)

Tags: buck converter, DC:DC power supply, feedback loop, Zero ESR Filter Capacitors

What’s Up with Feedback Loops?

October 23, 2009

Dear Mister Baffled,

If we’re talking about making people’s lives easier, you could connect our parts to well-behaved loads with well-behaved sources and we could sell you expensive resistors in fancy packages and make a lot of money.

However, most parts that provide a regulated output have a feedback loop for control and enhanced stability. Sometimes this feedback loop is buried inside the part…in that case we design a control loop we think will work well for a variety of circuits a customer might hook it to. Even a simple device like a Low-Dropout (LDO) regulator has a feedback loop from the output to control the conduction of the pass transistor. This is why, in rare situations, when a load is ill-behaved or a board layout is poor or bypass capacitors are improperly selected-the output will oscillate. In a feedback loop, we’re always trying to balance the aggressiveness of the response (the speed that the regulator will respond to load changes) with stability over temperature, component variation, worst case circuit and device influences. This brings up subjects like Bode plots and phase/gain margin that are uncomfortable for some people.

So, when you look at designing with a part like the FAN21SV04 and see the external feedback loop components, we’re trying to do you a service. You can adjust the feedback compensation to meet the transient requirements of your design.  So I hope you agree - including a feedback loop will make your life easier.

Now if you’ll excuse me, I have to visit my therapist.

Tags: Dr. FAE, external feedback loop, FAN21SV04, feedback loop, LDO

Dr. FAE: Voltage Feed-Forward Feature for Power Supply Controllers

September 21, 2009

Dear Dr. Fred A. Engleberry. Could you please explain the advantage of the voltage feed-forward feature of many of your power supply controllers? - -      Puzzled in Peoria

Greetings to you in Peoria.

My IQ is 181 so, certainly, I could explain the advantage of voltage feed-forward. In anticipation of your next question, I will elaborate…

This morning, over a warm, half-caf, double-short, non-fat latté at the local coffee shop, I was enjoying a recreational review of Kreck and Lück’s Novikov Conjecture (Geometry and Algebra) which says the following:

Finally, we indicate the proof of Theorem 16.2 for arbitrary n. The idea is to work inductively. If f is a diffeomorphism on Tⁿ X P with P a 1-connected manifold, one can isotope it so that it preserves N^n-1 X P.

My thinking might be illustrated more clearly with a transfer function from Erickson and Maksimovich, Fundamentals of Power Electronics:

This formula clearly shows that input voltage is not a variable. It contains a built-in assumption that the input voltage is invariant. Adding input voltage greatly complicates the transfer function.

Dear Doctor FAE, pardon me, but I do not recognize that answer as plain English.

-         PiP

Very well, I shall explain without the crystal clarity of the simple equation. The control loop of a DC-DC converter operates by sampling the output voltage and adjusting the pulse width modulation of the power train. The control loop acts as follows: if the output voltage changes, then we adjust power supply to counteract the change and keep the output stable.

However, if the input voltage changes, the power supply must respond to this change too. We could wait for the effect of the input change to appear at the power supply output, but wouldn’t it be glorious if we could monitor the input voltage and adjust the PWM immediately without waiting for the output voltage to change…if we provided some direct control method that did not complicate the feedback loop?

That’s the advantage of voltage feed-forward.

We do this by allowing the input voltage to directly modulate the slope of the PWM ramp. With an increased input voltage, the slope of the ramp increases and crosses the feedback signal sooner, giving a shorter output control pulse. Get it? Thus, increasing Vin reduces the PWM control signal outside of the output voltage control loop.

Now, if you’ll excuse me… I’ll be seeking a refreshing nap.

Tags: DC-DC converter, power supply controllers, voltage feed-forward

Why did my FET fail?

August 25, 2009

We’re introducing a new blogger to the site: Dr. Fred A. Engleberry. Dr. F.A.E. holds a PhD from MIT (Muckton Institute of Talknology) and has several months of valuable experience with applied technology. We are pleased to have Dr. F.A.E. available to answer questions collected from customers around the world.

Dr. F.A.E, “Smoky” from General Specifics Inc. sent a question…”Why did my FET fail?” Without further ado, we’ll turn the session over to Dr. Fred.

First of all, Smoky, you’re probably expecting a lot of annoying questions about your design. Such as, what frequency you’re running at, what the gate drive circuit looks like, what the load is and what supply voltage is present. Some design engineers might try to determine whether there is an avalanche condition beyond what the device might be reasonably expected to tolerate, whether the gate drive is insufficient or oscillating, whether the load is inductive, whether voltage spikes creep too close to breakdown voltages on the gate or drain, or whether the total package dissipation is being exceeded.

However, let’s say in this instance I could see your schematic and Bill of Materials (BOM). Your gate resistor, R42, as noted on the schematic, should be one ohm, but the BOM shows 1,000 ohms. Replace this resistor with the proper value and you will find that your FET turn-on and turn-off rise and fall times will become reasonable and you will avoid gate oscillation – and your FET design will become robust.

Want more information on FETs? Check out our website for MOSFETs at http://fairchildsemi.com/products/mosfets/index.html

Tags: Dr. FAE, FET designs, FET failed, gate oscillation, MOSFETs