AD4020 in single-ended operation

Thanks for your query.

The AD4020 has an input common-mode voltage requirement which requires that the average voltage of the IN+ and IN- pins be nominally VREF/2 (+/- 100mV).

The proposed single-ended circuit would unfortunately violate this requirement, as the common mode would end up spanning from 0 to VREF/2.

This is why the EVAL-AD4020FMCZ board is configured by default with VREF/2 driven to the inputs of the ADA4807-1s driving the AD4020 inputs.

A single-ended to differential conversion can be implemented either with two op amps or with one fully-differential amplifier (FDA).

The op amp option is pictured in Figure 76 of the ADA4807 data sheet. The single-ended input signal is applied to the input of a non-inverting op amp, and the antiphase signal (the other "half" of the differential output signal) is generated by applying the output of the non-inverting amplifier to the input of an inverting amplifier with gain of -1. This configuration is also supported by the EVAL-AD4020FMCZ board by modifying the on-board passive components (most notably, R87 acts to feed the output of U12 to the inverting input of U14).

The FDA option is a simpler circuit with fewer components but unfortunately the recommended ADA4940-1 (U2) is not populated on the EVAL-AD4020FMCZ board by default. Personally I would recommend prototyping with the AMC-ADA4945-1EBZ board, which comes populated with the ADA4945-1 with feedback networks and appropriate VOCM voltage already generated by default. This board slots into J8 and J9 headers to bypass the amplifiers populated on the EVAL-AD4020FMCZ board. Refer to Table 4 in the EVAL-AD4020FMCZ User Guide for instructions on configuring the EVAL-AD4020FMCZ board to work with the AMC-ADA4945-1EBZ board.

The op amp approach can also be implemented on the op-amp compatible AMCs, which are described in UG-1224.

The main differences between the two architectures are that the FDA option is simpler with fewer active components, but it does not provide a high-impedance input, since the signal source will be driving its feedback network instead of a high-impedance op amp input.

Let me know if you have further queries.

-Tyler

Hi, Tyler! 

I followed your advise by modifying EVAL-AD4020FMCZ circuit. I choose to work with two ADA4807 op amps. I made some changes to achieve Fig.76 circuit. 

But problem is that i got an almost +6V at coax input VIN+ somehow. VIN- is doing fine, it's about +2.5V. Power part of evaluation board at it default values. 

Here is changes i made: R17, R22, R37, -> jumper. R32, R87 -> 1kOhm. R28, R27 -> removed.

Thank you for help!

Do you mean you're seeing +6V at the SMA input that's connected to the U12 non-inverting input?

Do you see this voltage while you're driving it with an external signal source?

It would make sense if it were "floating" to +6V while you're not driving the pin, because R27 and R28 have been removed and are no longer driving the input of U12 to 2.5V (VREF/2).

-Tyler

Do you mean you're seeing +6V at the SMA input that's connected to the U12 non-inverting input?

Do you see this voltage while you're driving it with an external signal source?

It would make sense if it were "floating" to +6V while you're not driving the pin, because R27 and R28 have been removed and are no longer driving the input of U12 to 2.5V (VREF/2).

-Tyler

Hello!

That was a bad idea to left floating opamp's input and trying measure something:) After connecting external source i got what i exppected, however decoding spi output to real values was kinda tricky for me.

Everything works fine for now. Thanks for helping me out!