I told them it could accept ±40V input, but...

At one point, a customer consulted me about a design that used an AD converter called the AD4857.

The AD4857 is a product that features a wide input range and a "soft span" function that allows the input range to be changed by setting an internal register. The datasheet states in the features section at the beginning that the input range can be set to ±40V.

The customer then asked if it was possible to input a ± 30V signal, and I explained that it was not a problem.

However, later, when I checked the datasheet, I realized I had made a big mistake.

What was the problem?

The ±40V input range stated in the datasheet is for "differential input" conditions, and your circuit configuration uses single-ended (effectively pseudo-differential) input, which means it cannot meet the input specifications.

In other words, I made the typical mistake of explaining the AD converter based solely on its maximum input range without examining its input structure.

Therefore, from here on, I will introduce the differences in each configuration, including the concept of input range.

Single-ended input method

The most basic configuration for an AD converter is a single-ended input method.

In this configuration, the input signal is converted using the AD converter's GND as the reference. In other words, the range of the input voltage is determined by "how far it can swing relative to GND." For this reason, the input voltage is usually limited to around GND, the reference voltage, or the power supply voltage.

While single-ended inputs have the advantage of a simple configuration, the signal is always evaluated relative to GND, making it directly susceptible to GND noise and potential fluctuations caused by power supply return current. Therefore, performance limitations tend to become apparent in noisy environments or for high-precision measurement applications.

Pseudo-differential input method

Pseudo-differential inputs can alleviate some of the limitations of single-ended connections. In this method, the difference between IN+ and IN-is measured, but the IN-input is typically connected to the signal source's GND​. Therefore, although the measured value is treated as a difference, the voltage at each input terminal is often close to the reference voltage from GND, similar to single-ended connections.

Compared to single-ended inputs, pseudo-differential inputs can cancel out noise and potential fluctuations on the signal source GND (common-mode noise rejection), resulting in a configuration with stronger noise immunity and improved effective accuracy.

Differential input method

In differential input, both IN+ and IN-are signal inputs, and the AD converter converts the difference in voltage between them.

The important point here is that "the input range is defined not by the voltage of each terminal relative to GND, but by the difference between IN+ and IN-." In other words, the ± 40V differential input range that was the source of the failure means that the voltage difference between IN+ and IN-is ± 40V (for example, IN+: 20V, IN-: -20V).

Compared to single-ended or pseudo-differential inputs, differential inputs have the ability to double the input range for a given power supply and reference configuration, thereby expanding the dynamic range.

Furthermore, in a differential signaling system, signals are input to both IN+ and IN-, and the noise from the signal sources cancels each other out, resulting in the most noise-resistant configuration.

Therefore, this configuration is useful for applications that require higher accuracy in conversion.

Summary

This time, based on my own experience of failure, I focused on the concept of input voltage range and introduced the differences in input architectures of AD converters.

When selecting an AD converter, the maximum input range listed in the datasheet is very important information. However, if you don't check which input method that value is based on, you may not actually be able to input the expected voltage. The mistake I made this time, assuming that it could accept ±40V input, was precisely due to a lack of this verification.

Each input method has its own advantages and disadvantages. Therefore, when choosing an AD converter, it is important to check not only the maximum input range, but also the input method, the voltage range relative to GND for each terminal, and even the reference point for the signal you want to measure.

When I was organizing my thoughts on AD converter input methods, I referred to the following article by Analog Devices. If you'd like to learn more about single-ended input, pseudo-differential input, and differential input, I highly recommend reading it.

Examination of various analog input architectures for SAR ADCs

Good luck Tanepens ~Support diary for young analog FAEs~ Article list

・ Part 1: How do you think about ADC resolution?

・ Part 2: Techniques for improving ADC resolution

・Regarding the third session, I stated that it could accept ±40V input, but...

Inquiry

If you have any questions regarding this article, please contact us below.