Field guide
Measuring focuser backlash
Backlash is the slack your motor takes up before the drawtube starts to move. It only costs you anything when the focuser changes direction — which is exactly what an autofocus sweep does, twice.
01 What you are measuring
Turn the motor one way, then reverse it. For the first fraction of a turn the gear teeth are simply crossing the gap between them. The shaft rotates. The drawtube does not. Drive it yourself:
Press Out a few times, then In. Watch the needle stall.
Reverse direction and the first 60 steps do nothing at all. That is your backlash. Every autofocus V-curve reverses at least once.
02 Before you measure anything
Backlash only bites on reversal. If your imaging software always finishes its approach from the same direction, the number barely matters. Most packages call this final inward move or overshoot compensation. Turn it on first.
Measure only if you still see asymmetric V-curves, or if your software asks for a figure it can compensate with. Then pick a method below.
03 Method A — dial indicator
The accurate one. Twenty minutes, and it double-checks the microns-per-step figure the calculator needs anyway. Do it with the camera attached and the tube at a typical imaging altitude.
-
Set up against the drawtube face
Magnetic base on the focuser body or OTA. Indicator tip square to the flat face of the drawtube, pre-loaded about half a millimetre.
-
Drive out 500 steps, then zero the dial
This takes up every scrap of slack in one direction. Now the gear teeth are hard against each other. Zeroing here is what makes the next step meaningful.
-
Drive out 200 more and check the dial
The needle should move by roughly
200 × your microns-per-step. If it doesn't, your calibration is wrong and nothing downstream will be right. Fix that before continuing. -
Reverse. Drive in, five steps at a time, counting
The needle will sit dead still. Keep counting. The moment it first twitches, stop.
-
Steps counted before the needle moved = your backlash
Repeat twice more. Use the largest of the three, not the average — you want the compensation to clear the slack every time, not most of the time.
04 Method B — on a star, no tools
Less precise, and seeing will scatter your answer by roughly a fifth. But it needs nothing you don't already own, and it measures the system as it actually images.
-
Loop short exposures on a moderately bright star
Bright enough for a stable HFD reading, faint enough not to saturate. Two to three second subs are plenty.
-
Drive out until HFD is climbing steadily
You want to be well outside the focus zone, on a clean slope. Stop.
-
Reverse. Drive in ten steps at a time, counting
HFD will hold flat while the slack takes up. When it begins to fall, stop. Your count is the backlash.
-
Run it three times
Seeing moves HFD around on its own. Three runs tells you whether you measured backlash or measured the atmosphere.
05 What the guides leave out
- Measure loaded, and pointed high. A bare drawtube on the bench gives a number that does not survive four kilograms of camera at zenith.
- It is direction-dependent. Moving against gravity is not the same as moving with it. Measure in the direction your software will actually approach from.
- Cold grease is stiff grease. A July figure is not a January figure. Re-check when the season turns.
- A huge reading is usually a loose coupler, not a bad gearbox. If an EAF or FocusCube reports much north of 200 steps, check the grub screws on the shaft coupler before you believe the number. Pegasus calls this out in their own manual. Slipping hardware presents exactly like backlash.
- Zero is a legitimate answer. Direct-drive and worm-drive units — NightCrawler, Esatto, the high-precision Q-Focuser on a good focuser — genuinely add none.