A field check is not a calibration

Calibration at a service centre is measurement and adjustment against a reference, with a certificate you can show a client. A field check is something else: a quick test that tells you whether the instrument is still the way it was at the last calibration, or whether someone dropped it in the van. The first is done periodically. The second is done often, and ideally after every drive down a bad road and after every knock you heard the sound of.

At Geonextech we see instruments arriving at the service centre with errors that have been building up for months without anyone noticing. Half an hour of checking a month would catch most of them.

Level: the two-peg test

The oldest and most reliable test there is. Drive two pegs about 40 to 60 metres apart. Set the level in the middle and measure the height difference between them. Then move it close to one peg and measure again. If the two differences agree within a few tenths of a millimetre per 100 metres, the line of sight is horizontal. If not, the difference tells you how far the instrument is out, and whether it needs to go for service or can still be used with equal sight distances, which cancels the error. Do the test on flat ground, in still air and with the staffs at the same height above the ground, so you do not argue afterwards whether the error is in the instrument or in the shimmer over hot asphalt.

On an optical level check the compensator as well. The DSZ3 has a button for exactly this: press it and watch whether the line of sight swings and returns to the same place. If it does not move, the compensator is stuck, usually from moisture or a knock. On a digital level do the same two-peg test with the barcode staff, and check that the instrument recognises the staff at the shortest and the longest distance you use.

Total station: two faces and a known point

The bubble first. Level up, turn through 180 degrees, see whether the bubble has stayed in the middle. Second, measure one point in face left and face right. The difference between the two, corrected by 180 degrees, is the collimation error and the trunnion axis error. Every station has a routine in the menu that determines and stores them. Run it on every new site, it takes five minutes.

Third, distance. Measure a known baseline, or simply the distance between two stable marks you have measured before. If you are out by centimetres, the usual culprit is a prism constant entered for a different prism. The classic case is a constant for another brand's prism left in the station months after the swap; everything looks right, only a few centimetres short. Check the atmospheric correction too. On the OTS682 temperature and pressure are read from built-in sensors; on the other models you enter them, and if they are still set from winter, summer will cost you millimetres on every hundred metres. On a robotic station such as the RS10 check the automatic aiming as well: measure one prism manually and automatically and compare.

GNSS receiver: a known point and the IMU

A GNSS receiver has no axes to adjust, but there are three places where error gets in. Antenna height: measure it with a tape and compare it with what is entered in the software. Known point: start every day by measuring a benchmark you know and look at the difference. If it is more than the usual two or three centimetres, the problem is in the base, the coordinate system or the network, not in the point. IMU: on the A90 and the A66 MAX tilt compensation is initialised and periodically checked following the software's routine. After changing the pole or the adapter repeat it, because the geometry has changed. If the receiver has sat in storage for a long time, switch it on outdoors the evening before so it downloads current orbit data and you do not lose the first minutes on site waiting.

Transport, batteries, moisture

Most of the damage we see comes not from work but from the road between sites. This is what we go through before an instrument leaves in the morning:

  • The instrument is in its case with every latch closed, and the case is secured rather than sitting loose in the boot on top of shovels.
  • The total station comes off the tripod for every transport, even a hundred metres by car. The tripod is folded separately.
  • Batteries are charged with the original charger and in cold weather they stay in a pocket, not in the case. Stated running times assume normal temperature.
  • A wet instrument is dried with a soft cloth and left with the case open at room temperature. A closed wet case overnight is the surest way to fog the optics from the inside.
  • Lenses are cleaned only with a brush and an optics cloth, never with a sleeve.
  • The tripod is checked too: loose leg screws and worn tips produce errors you will later blame on the instrument.

IP68 on the receivers and IP55 on the total stations means protection against dust and water in normal use. It does not mean you can leave the instrument out in the rain until morning.

A practical routine that works: two-peg test and two-face check on every new site, a known point with GNSS every morning, a look inside the case every time before you shut the van door. And when a field check shows something you cannot explain, stop and call, instead of "compensating" for it in the calculations. That is what calibration is for.