Fast (But Not Too Fast) Op Amps for Reliable Very Low-Power Systems
February 09, 2016
|By: Jim Harrison
Guest Blogger, Lincoln Technology Communications
Many electronic design engineers find themselves looking for that ‘just right’ op amp for applications like portable test equipment, medical diagnostic equipment, and industrial sensor data gathering. An amp that is fast—but not so fast that it is a pain to design with—and a device that has all the other performance bases very well covered—plus really, really low power.
The low power specification of amplifiers like these not only fits portable, battery-operated designs, but also makes the circuit design all together easier for any design. A smaller power supply is needed, circuit layout is easier, and the circuit generates less noise and interference than higher power versions.
The ideal amplifier for this (and really any) application would also provide DC precision and low noise. If you can find an amp with all these traits that takes only, say, 10µA supply current, why not use it—even if low power is not a supreme design requirement?
Yes, there are lots of op amp choices. The Mouser web site, just as an example, has specs for 7,607 op amps. If you restrict it to those requiring less than 100µA supply current, you still have 1,317 choices. I chose two really good ones from Maxim for a more in-depth look. One has 1.5MHz bandwidth, the other 200KHz—and both feature great performance specs.
The Faster Amp
The 1.5MHz gain-bandwidth product device is the MAX9617. It takes a single 1.8 to 5.5V supply and pulls just 78µA maximum supply current (111µA over the full temperature range). The amp is unity gain stable with a CMMR of at least 116db, and its inputs operate from rail to rail. It has a wide -40˚ to 125˚C operating temperature that I feel will yield higher reliability.
Fig 1. A simple low-noise A/D front end using the MAX9617.
Some other characteristics: Input bias current is just 80pA maximum at room temp, and only rises to 580pA at 125˚C. Offset voltage is no more than 10µV at room temperature and slew rate is 0.7 v/µs. This device particularly shines in offset voltage drift—just 120 nV/˚C maximum.
Input current noise density is 0.100 pA/√Hz at 1KHz and voltage noise density is 42 nV/√Hz. This device’s input voltage noise is especially low for a CMOS input operational amplifier. The great noise specs are particularly helpful in low signal level sensor interfaces.
The kind of specs this amp has would have made it a full instrumentation-grade amp just a few years ago. It comes in a small 6-pin SC70 package and is priced at $1.10 ea/2,500.
Fig 2. An evaluation kit is available for the MAX9617 operational amplifier.
Another Excellent Amp to Consider
For a somewhat slower amp that needs just 5.5µA supply current maximum over its full -40˚ to 85˚C operating range, you can look to the MAX9911. This device has a 200KHz gain-bandwidth and a 0.1 V/µs slew rate. It comes in a 6-pin SC70 or tiny WLP package and is priced at $0.51 ea/2,500.
The CMOS op amp features a really low input-bias current of 1pA typical, making it suitable for high Z sensor interfaces in applications like photodiode and charge sense circuits. Over its full temperature range, the input bias current is just ±30pA maximum. The amp has rail-to-rail inputs and outputs and operates from a single 1.8 to 5.5V supply.
Fig 3. The MAX9911 is available in tiny 6-bump WLP and 6-pin SC70 packages.
For even more power savings, the device has a shutdown mode that reduces supply current to less than 500nA at room temperature, and puts the amplifiers’ outputs in a high-impedance state. Delay time to shutdown is 2µs and re-activate time is 30µs. Over its temperature range, the device’s offset voltage can be ±5mV and CMRR is 60db. Input current noise density is 0.001 pA/√Hz at 1kHz, while the input voltage-noise density is 400 nV/√Hz. A MAX9911EVKIT+ evaluation board is available for $60.
I believe these are two of the best very low power op amps you can get—among the very great many available. They would make a good addition to any EE’s toolbox and they are both, as of this date, in stock in distribution.
As for my background (in case you haven’t read my work), I was a senior technical editor with Electronics Products Magazine for more than 12 years and am now the editor-in-chief of Lincoln Technology Communications in San Rafael, CA. I have held senior design engineering positions with industrial automation and scientific instrumentation companies since 1989, and was a component engineering manager at Nokia Broadband Systems Division in Petaluma, CA, for five years starting in 1996 and with Calix Networks for two years after that. I attended the University of Iowa and was a design engineer at the university’s Physics Research Center for three years.