Distinguish one-line and simultaneous rejection
Start by identifying which line-frequency interference the measurement must address. Rejection of one mains frequency and simultaneous rejection of both line frequencies are different requirements. The ADS124S0x filter discussion distinguishes those cases, so the general phrase line rejection should not replace the actual interference target.
A notch is a feature of the filter response at a particular frequency. Stating that a mode has one relevant notch does not, by itself, state that it also supplies the other required notch. The document describes combinations that reject one line frequency and combinations that reject both, making the distinction a positive part of selecting the response rather than a wording detail.
Keep the interference target in the comparison before considering how the converter produces its output samples. A requirement for simultaneous rejection should remain identifiable as such instead of being inferred from a nearby single-line statement. This establishes the kind of response being sought without estimating rejection from plots or assigning a system result. It also prevents a change in the mains-frequency assumption from being hidden inside an otherwise unchanged converter description.
Retain mode and data rate together
The named filter mode and output data rate form a pair. The ADS124S0x documentation describes low-latency and sinc3 responses, and the same rate label does not make those responses interchangeable. The filter mode contributes the response shape; the rate setting belongs with the location of the relevant response features.
This is why a data-rate entry alone is not a complete line-rejection specification. Two comparisons can name the same nominal output rate while referring to different filter modes. Conversely, a statement about a mode without the associated rate can omit the condition that determines which line frequencies its response addresses.
Describe the mode/rate combination as one configuration identity, then compare its stated rejection context with the interference target. This provides a concrete answer to what configuration is under discussion without turning prose into a register recipe. The cited revision contains conflicting FILTER wording, so this guide does not resolve that conflict by choosing a bit value. The response names and their documented conditions remain useful even while that implementation detail requires authoritative clarification.
Keep clock accuracy with notch placement
The clock provides the timing basis for both output data rate and notch placement. The cited prose links the response frequencies to the converter clock, so a nominal rate or notch statement should retain the oscillator basis under which it is described. A response is not independent of the timing reference that scales it.
This matters when the clock changes with operating conditions. The document discusses the internal oscillator and the use of an external precision clock where tighter frequency tolerance is required. The distinction is between the nominal response description and the precision with which its frequency features are positioned; these are related but separate concerns.
Keep the clock assumption with the filter-mode and rate statement. That lets a reader distinguish a change in response selection from a movement of the response caused by its timing basis. The discussion does not select a clock, prescribe a circuit or compute a tolerance budget. It explains why a line-frequency requirement cannot be fully compared with a nominal notch description after the clock condition has been omitted. No system rejection guarantee follows merely from naming the intended notch.
Preserve the measured signal bandwidth
A filter that rejects an unwanted line component also has a passband for the wanted signal. The ADS124S0x tables report a minus-three-decibel bandwidth for each relevant mode and rate. This is a separate response coordinate from the presence of a line-frequency notch, and it matters when identifying which signal variation is intended to remain useful.
The low-latency and sinc3 tables show that a shared output data rate does not establish an identical listed bandwidth. Output rate describes how often results are produced; passband describes part of the frequency response applied to the signal. Treating those descriptions as equivalent can conceal a material difference between two otherwise similar configuration summaries.
Compare the useful signal range with the applicable mode/rate bandwidth entry while retaining the clock basis. This adds a fourth concrete question to the rejection review: not only which unwanted component is suppressed, but what wanted variation the selected response describes. It does not estimate a plotted attenuation, certify signal amplitude or replace an application evaluation. The purpose is to preserve both sides of the filter decision instead of treating rejection language as a complete account of measurement behavior.
Sources and further reading
Named ADS124S06/ADS124S08 SBAS660C filter prose and passband tables. FILTER prose inconsistency is unresolved; no bit-setting recipe, curve-derived rejection or system guarantee follows.