# The Diamond Plot

Q : My signal is within the instrumentation amplifier’s specified input voltage range, but the output seems saturated. What’s happening? Is the part broken?

A : On February 18, 2013, a spectacular and successful diamond heist was carried out in about five minutes, while loading the cargo onto an aircraft headed from Brussels to Zurich. An estimated value of \$350 million makes it one of the largest diamond heists of all time. With thieves disguised as police officers (down to the armbands), an elaborate plot was executed quickly and with high precision, similar to what would be seen in a Hollywood movie. The passengers didn’t notice until everyone had to disembark the airplane. The robbery started an international manhunt to bring the burglars to justice.

Those of us working with electronics need to be concerned with a different diamond plot. When using in-amps, designers often observe strange phenomena. The problem is sometimes accurately explained as “the output is saturated,” but other times the description can be a little more cryptic: “the gain error is very high,” or “the amplifier is very nonlinear,” or simply “it doesn’t work when it should.” We don’t have a flow chart to solve customer problems, but if we had one, “check the diamond plot” would come right after “make sure the part is powered up.”

In this context, the diamond plot, often found in the data sheet, shows an in-amp’s input common-mode voltage vs. output voltage range. If the operating conditions fall within the contour, the device should operate properly; otherwise, the output will be invalid due to saturation of internal nodes.

For readers unfamiliar with in-amps, these linear devices amplify the voltage difference between their inputs, independent of the input voltage relative to the supply voltage. The input common-mode voltage, which is the average of the two input voltages, is rejected by the amplifier.

Naturally, operation is restricted to a limited voltage range, which most people would expect to be less than the supplies, so it’s generally not a problem. However, the common-mode voltage does not simply disappear as it enters the circuit. Instead, it’s internally subtracted from the desired signal. This means that the amplified signal and the common-mode voltage must fit within the voltage rails. The mechanism by which the common-mode voltage is subtracted depends on the particular circuit topology, giving the contour a specific shape, which can be an octagon, hexagon, or parallelogram. “Diamond plot” can be a bit of a misnomer, but these contour plots provide the circuit designer with useful information regarding the proper operating range given the input voltages, desired output swing, reference voltage, and supply rails.

This problem becomes more challenging when working with low supply voltages and single-supply applications, because the diamond plot becomes much smaller and the operation range becomes even more restricted. Modern in-amps such as the AD8226, AD8227, AD8420, and AD8422 aim to expand the diamond plot as much as possible. To simplify low-voltage designs, the AD8237’s diamond plot exceeds the supply, as shown in Figure 1.

So, next time you’re designing with an in-amp, remember to consider the diamond plot. At least this plot won’t get Interpol knocking on your door to recover the stolen rocks!

Figure 1

Input common-mode voltage vs. output voltage.

— Gustavo Castro (gustavo.castro@analog.com) is an applications engineer in the Linear Products Group in Wilmington, Mass. His main interests are analog and mixed-signal design for precision signal conditioning and electronic instrumentation. Prior to joining Analog Devices in 2011, he worked for 10 years designing high-performance digital multimeters and precision DC sources at National Instruments. Gustavo received a Bachelor’s degree in electronics engineering in 2000 from Monterrey Institute of Technology, Mexico.

## 7 comments on “The Diamond Plot”

1. etnapowers
July 1, 2014

“The problem is sometimes accurately explained as “the output is saturated,” but other times the description can be a little more cryptic: “the gain error is very high,” or “the amplifier is very nonlinear,” or simply “it doesn't work when it should.” We don't have a flow chart to solve customer problems, but if we had one, “check the diamond plot” would come right after “make sure the part is powered up.”

Among the possible reasons for the saturation of the output there is also the load variation: a load that requires a different power amount due to , for example, a degradation of its leakage current, may saturate the output when the load driving capability is overpassed.

2. etnapowers
July 1, 2014

“these linear devices amplify the voltage difference between their inputs, independent of the input voltage relative to the supply voltage. The input common-mode voltage, which is the average of the two input voltages, is rejected by the amplifier.”

The CMRR (Common Mode Rejection Ratio) is the parameter utilized to describe the quality of an amplifier, being the ratio between the amplification factor of the differential mode voltage and the amplification factor of the common mode voltage.

Ideally the CMRR should be  as big as possible.

3. etnapowers
July 1, 2014

The input impedance of these amplifiers should be as high as possible, ideally infinite, to decouple the amplifier from the input source, and for the same reason the output impedance should be as low as possible, ideally zero.

4. vasanjk
July 2, 2014

Steve

I never understood the importance of the “plot” before reading this blog. If the signal input amplitudes are sufficiently small and as long as the used gain factors do not throw the output to the rail, for example in case of patient monitoring like ECG and EEG, the plot goes unnoticed.

5. Steve Taranovich
July 2, 2014

@vasanjk—-I was “bitten” by this kind of design oversight in my early design years, but Burr-Brown engineers stressed the Common mode plot to me when I worked for them. There was even a program on TI which is a rollover from Burr-Brown that can tell you if your input signals are going to be a problem—you insert your design criteria and the INA part number and it will let you know if there will be a problem.

I can't seem to find it, but you can ask TI apps on the E2E community

6. Davidled
July 7, 2014

Saturation point (SP) limits the circuit, and defines partially product electric characteristics. Saturation could be defined by each component as well as the integrated system. SP could be slightly different among each component. To investigate SP, designer checks these parameters: temperature; current; and voltage.

7. etnapowers
July 8, 2014

That's correct, the saturation has to be accurately simulated by designers that have not only to simulate the effect of the  variation of parameters like temperature, voltage and current but also the worst case scenarios in terms of drift of the parameters (montecarlo simulations) and load-line variation

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