Analog Angle Blog

Do We Really Need Yet Another Op-Amp?

To the non-analog designer, the vast number of op-amps available from any single top-tier vendor can be overwhelming. Sure, there are selection guides and other tools, but you still have to admire it (and perhaps be just a little scared) when you see literally hundreds of distinct devices — and that's not counting their various temperature and performance grades. Frankly, it can even be a little intimidating to those who know and resonate with all the vagaries and subtle parameters of this basic building block of the analog world.

The interesting thing is that, even with so many pretty good, very good, and quite excellent op-amps already on the market, there seems to always be a need for new and better ones. Of course, “better” is the key word in this situation. What makes one op-amp better than another is not necessarily improved specifications. Instead, it may have a better combination of specs that hits a sweeter spot in the eternal struggle to balance the many tradeoffs and compromises inherent in the analog world.

Several recent introductions substantiate my feelings, as you'll see from these extracts from vendor press releases. (This is an unscientific, somewhat randomly chosen selection.)

(Source: Linear Technology Corp.)

(Source: Linear Technology Corp.)

“Linear Technology introduces the LTC2057HV, a zero-drift amplifier featuring self-calibrating circuitry that provides high DC precision and stability over changes in temperature, time, input range and supply voltage. With 5μV input offset voltage, 0.025μV/°C offset drift and 220nVP-P low-frequency noise with no 1/f noise, the LTC2057HV offers more than 140dB dynamic range while operating on a 60V (±30V) supply.”

(Source: Texas Instruments)

(Source: Texas Instruments)

“Texas Instruments… today introduced a fully differential, zero-drift, 36-V programmable gain amplifier (PGA). At 5μV, the PGA281 offers the lowest offset voltage in its class to improve accuracy and long-term stability… Designers can use the PGA281 in industrial signal acquisition applications, including test and measurement, strain gauges, bridge amplifiers, and medical instrumentation… [It has] an offset voltage of 25μV and an offset voltage drift of 174nV/°C with a maximum gain of 128… CMRR [is] more than 140dB.”

(Source: Linear Technology Corp.)

(Source: Linear Technology Corp.)

“Linear Technology introduces the LT6015 single Over-the-Top operational amplifier with outstanding precision over a 0V to 76V input common-mode voltage range. It incorporates multiple built-in fault tolerant features, resulting in no-compromise performance over wide operating supply and temperature ranges. Over-the-Top inputs provide true operation well beyond the V rail. The LT6015 functions normally with its inputs up to 76V above V-, independent of whether V+ is 3V or 50V. Input offset voltage is 80μV max, input bias current is 5nA and low frequency noise is 0.5μVP-P … Fault protection modes guard against negative transients, reverse battery and other conditions… for applications where the amplifier is at the analog interface to another board, and for high-side and low-side current sensing.”

So there you have it. Op-amps encompass a wide array of performance tradeoffs, ranging from being very, very good in one or two parameters but just so-so in others (for example, speed versus power, or accuracy versus cost) to being pretty good in most areas but not outstanding in any one. Making the right decision is all about making choices among the various tradeoffs — and hey, isn't that what engineering is all about? And many times, spending a little more time (or money) for a better-fitting op-amp can save you effort in avoiding calibration, temperature compensation, and many other nasty real-world issues.

Have you ever been flustered or overwhelmed by too many choices for op-amps or some other components? Or do you like having this many parts and options to evaluate, so you can drill down to find one which is even just a slightly better match for your needs?

Related posts:

7 comments on “Do We Really Need Yet Another Op-Amp?

  1. Bill_Jaffa
    October 22, 2013

    Let me clarify that PGA281 offset spec: it's 5µV (typical) and 25µV (maximum)–both very good numbers that will eliminate the need for trim or calibration in many applications.

  2. hstearnsjr
    October 23, 2013

    Yes, it's time consuming and frustrating, expecially when the application is not critical and most anything would do.

    The hardest search I had was for an op-amp suitable for high end audio,

    DC parameters don't matter, need high open loop gain, low noise,  wide bandwidth etc.  I settled on the 5532 years ago, but i”ve seen better lately.

    Audio seems to be an exceptionally difficult domain because of the >2000:1 bandwidth ratio; huge dynamic range, and the human ear's ability to detect distortion beyond the capability of most any measurement instrument.

  3. eafpres
    October 23, 2013

    @Hoyt–“because of the >2000:1 bandwidth ratio”.  I'm not completely clear what you mean in this case.  Generally good audio is 20 to 20000 Hz, which is 1000:1.  Above 16000 most won't hear it, and the low end becomes more vibration than sonic.  Can you clarify what bandwidth you mean here?

    I'm not disagreeing with you as to the challenge!  Just want to understand more what you are describing.

  4. eafpres
    October 23, 2013

    @Bill–I think the challenge can be as hard or harder with passives.  Let's take something simple like a ferrite chip bead.  The range of parts across vendors is mind boggling.  Frequently as well, the behavior is not ideal across frequencies, so you may have to be careful with signal integrity parts (caps, inductors, ferrites, etc.) to avoid degrading signal, and the tradeoffs there may not be clear from the data sheets.  In some cases you might need to use SPICE or some other simulation to really choose the right parts.

  5. samicksha
    October 24, 2013

    I guess its consideration of Bandwidth gain gain before designing the topology, another parameter is filter cutoff frequency.

  6. yalanand
    October 27, 2013

    Operational amplifiers had their geneses in analog computers, where they were used to do mathematical processes in many linear, non-linear and frequency-reliant circuits. Features of a circuit spending an op-amp are set by outward components with little reliance on temperature variations or manufacturing differences in the op-amp itself, which makes op-amps general building blocks for circuit design.

  7. RedDerek
    November 6, 2013

    I can see adding an opamp due to improved process to increase bandwidth, reduce input bias current, better offset voltage specs, improved operating voltage, packaging, and so forth. And, unfortunately, the old parts have been designed and tested in a circuit that one would not want to necessarily change to a newer part due to the cost of retesting.

    Then compound it with the number of manufacturers. Just looking at the number of opamps that come up when one searches DigiKey ( – today's count is 34102! (not separating tape and reel versus small quantity pricing).

    I guess the real trick is to start designing an opamp to replace several old ones without affecting circuit performance. But some circuits work well with a slow opamp – less noise in the output. At least this has been in a couple of applications I have worked on. When working with repairing electronics from the 1980's and 1990's, the old parts are no longer available and I have to insert a newer product, I try to find a part with the low bandwidth and so forth. However, I do like the tighening up of the input offset voltage specs for many newer opamps do not have the offset adjust pins anymore. Another advantage with newer parts are the FET input type. They do tend to perform better in the old circuits.

Leave a Reply

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