Measuring and Tracking Backlash for Reliable Focusing

Why measure backlash?

In deep‑sky imaging, a stable, repeatable focus is essential to take full advantage of your telescope’s resolution and to keep FWHM/HFR low. Backlash is the mechanical slack that appears when you reverse the focuser’s direction: for a short distance, commanded steps do not translate into motion. If you ignore it:

• Autofocus can converge to a biased focus position (higher HFR than necessary);

• Repeatability suffers (the best focus you reach depends on the direction you approached from);

• Over time, the evolution of backlash becomes a mechanical health indicator: a sudden increase can hint at a tensioning/pressure issue, contamination on the drive surface, or a part that needs adjustment.

Measurement method

We measure HFR (or FWHM) while scanning the focuser position across the focus point:

1. Outward sweep: increase the position (tube moving out) in regular steps and record HFR(position).

2. Inward sweep: start on one side of the range and retract (decrease the position) in regular steps.

Each HFR=f(position) series is approximately U‑shaped. We fit a separate quadratic regression to the outward and inward points:

HFR ≈ a·x² + b·x + c

The minimum of each parabola (its vertex) gives the best‑focus position for that direction. The backlash is the horizontal difference between the two minima:

Backlash ≈ x_min(outward) − x_min(inward)

As a cross‑check, we also search for the horizontal shift that best aligns the two fitted curves (by minimizing the mean squared error). The resulting shift closely matches the “minima difference” estimate.

Practical notes

• Always finish focusing from the same direction (or enable backlash compensation) for maximum repeatability.

• Repeat the measurement under stable conditions (seeing, target altitude, temperature) and scan symmetrically around the expected focus point.

Results

2022 dataset

• Minimum (outward): 8382.22 steps

• Minimum (inward): 8369.91 steps

• Backlash from minima: ≈ 12.3 steps

• Best aligning horizontal shift (MSE): 13 steps

Reading: backlash around 12–13 steps.

2024 dataset

• Minimum (outward): 9058.24 steps

• Minimum (inward): 9043.07 steps

• Backlash from minima: ≈ 15.2 steps

• Best aligning horizontal shift (MSE): 15 steps

Reading: backlash around 15 steps, consistent across both estimators.

Evolution 2022 → 2024

A slight increase in backlash is observed (≈ +2–3 steps). This change is modest and consistent with normal wear and with the measurement scatter in HFR.

Plots and data

The figures “HFR vs. position” for 2024 and 2022 are provided alongside

Conclusion

The Lacerta MFOC + OCTO‑PLUS 2″ combo delivers an excellent precision and very little drift over time. Within two years, the backlash changed only mildly (from ~12–13 steps to ~15 steps), and the autofocus is crisp and repeatable provided we always finish from the same direction or set compensation around 14–16 steps.

On algorithms: Ekos now offers autofocus routines that are largely insensitive to backlash (e.g., linear/one‑pass approaches). They are robust but typically take longer to run. We still perform this backlash measurement regularly—both to monitor the mechanical health of the focuser and because we sometimes prefer faster focus strategies for certain imaging sessions when conditions and targets allow (comets, exoplanets, etc.)