top of page

Chromatography Theory Chapter 2 – Data Acquisition

  • Writer: Chromperfect
    Chromperfect
  • Sep 24
  • 2 min read
Chromatography Software for data acquisition

This chapter explains the math and practice of data acquisition for chromatography: sampling rate, digitization limits, noise vs. drift, converter resolution (bits → µV), and how to maximize dynamic range so both tiny and huge peaks are measured accurately.


What is chromatography data acquisition?


A detector outputs a continuous signal. The chromatography data acquisition system samples it at fixed intervals and stores a digitized raw file:


  • Discrete in time: each point is the average over the preceding interval.

  • Discrete in amplitude: each point is rounded to a converter step (quantization).Some information is lost, but with proper settings you capture everything that matters for height, area, and retention time.


Sampling rate


Rule of thumb: ≥10 points across the full baseline width of each peak.


  • LC: often 1–5 Hz (peaks up to ~1 min).

  • Capillary GC: 20–200 Hz (sub-second peaks).

  • Conventional GC: 5–25 Hz.Too slow → flattened apices, shifted RTs, area/height errors. Too fast → larger files (okay, just heavier).


Range (converter limits)


AD/DA converters have fixed upper/lower limits. If the signal hits them:


  • Upper limit: peak tops clip → height/area lost.

  • Lower limit: baseline segments clip → small peaks vanish.Fix by adjusting detector zero/attenuation so the signal stays comfortably within range during the run.


Noise and drift (quick recap)


  • Noise (fast): detector flicker, pump pulsations, electrical hum. Sets the practical detection limit.

  • Drift (slow): temperature/bleed/etc. Usually benign; handle column bleed via baseline subtraction (see Ch. 3).


Resolution (bits → step size)


Converter resolution = step size of the y-axis. Example: 20-bit over 1 V → ~1 µV/step (~10 µV needed for a 10-step minimum).


  • Peaks ≥10 steps high quantify reliably.

  • Quantization adds tiny “round-off noise.” If detector noise ≫ step size, raising bit depth mainly measures noise more precisely (not magical sensitivity).


Maximizing dynamic range (practical)


  • Use most of the converter span without touching limits.

  • Avoid feeding a tiny chart-recorder output (e.g., 10 mV max) into a 1–2 V ADC—this wastes ~99% of range and hurts low-level linearity.

  • If detector outputs 5–10 V, use a simple resistive attenuator to match ADC range, then recalibrate.


Quick setup checklist


  •  Choose sampling rate for ≥10 points across your narrowest expected peak.

  •  Set detector zero/attenuation so baseline and tallest expected peak sit well within ADC limits (allow headroom for drift).

  •  Confirm bit depth / range → smallest peaks are ≥10 steps.

  •  Verify noise & drift on a blank; if needed, address pump/electrical noise and plan baseline subtraction (see Ch. 3).

  •  Archive size okay? If not, reduce rate where peaks are wide.


Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page