inverted input for adc

[himadri117]

0 down vote favorite


I am using ADC122S625 (Datasheet: https://www.ti.com/lit/ds/symlink/adc122s625.pdf).

The ADC has a differential input per channel. My signal is from a RF devide with 3V peak to peak and 1.5V DC offset. I want to use the ADC in Single-Ended Input Operation. So I have choosen a Reference Voltage of 1.5V. Now my question is what should I use for the Inverted ADC input?

First I thought I will pull it to Ground as a normal Single input ADC would use 0 to Va input range. But after reading the Datasheet (page. 15) it seems that I need to feed the Vref (1.5V) to the inverted inputs.

From Datasheet (copied)::

"Single-Ended Input Operation For single-ended operation, the non-inverting inputs of the ADC122S625 can be driven with a signal that has a maximum to minimum value range that is equal to or less than twice the reference voltage. The inverting inputs should be biased at a stable voltage that is halfway between these maximum and minimum values. In order to utilize the entire dynamic range of the ADC122S625, the reference voltage is limited at VA / 2. This allows the non-inverting inputs the maximum swing range of ground to VA. Figure 35 shows the ADC122S625 being driven by a full-scale single-ended source. Even though the design of the ADC122S625 is optimized for a differential input, there is very little performance degradation while operating the ADC122S625 in single-ended fashion."



Please suggest to clear my understanding about the Inverted input.

 

[barry]

It looks like you need to put 1.5 volts on the inverting input.

 

[himadri117]

It looks like you need to put 1.5 volts on the inverting input.

Thanks for the reply. One more thought thats in my mind is the negative input defines the lower boundary of the signal as per the definition. Is it true in this case? If it is true then we will limit the lower boundary to 1.5V.

 

[barry]

No. The negative input defines the middle of the range. It says so right in the documentation YOU posted!

"The inverting inputs should be biased at a stable voltage that is halfway between these maximum and minimum values"

 

[dick_freebird]

If this is real RF, nobody ought to cry about you using a
balun.

 

[D.A.(Tony)Stewart]

It seems to me if you use the Vref=1.25 the optimum Vcm is 2.5V +/- 1.25

but Vref will be 1.5V to give the differential range of 3Vpp and thus the Vcm needs to be between 1.5~2.5V

 

[D.A.(Tony)Stewart]

Hold the Phone.... "RF devide" you mean RC divider.

this ADC does not do RF.....
200 kSPS is max

 

[himadri117]

Hold the Phone.... "RF devide" you mean RC divider.

this ADC does not do RF.....
200 kSPS is max

It is RF Device...sorry for the typo
My max input frequency is 60kHz, so 200 ksps is enough!

 

[D.A.(Tony)Stewart]

Although RF can be radiated at this frequency, if it conducted by some hardware that generates this signal, we do not call it RF, just a conducted signal from some source. If it is radiated such as an induction heater, then say so. RF covers a broad range.

Where is your block diagram?

 

[barry]

Although RF can be radiated at this frequency, if it conducted by some hardware that generates this signal, we do not call it RF, just a conducted signal from some source. If it is radiated such as an induction heater, then say so. RF covers a broad range.

Where is your block diagram?

A mixer is an 'rf device', and can certainly output 60KHz. But this is irrelevant. The issue here is the input reference voltage.