Hygrometer
Why salt works
A saturated salt solution — one with undissolved crystals still sitting in it — has a fixed water activity. Whatever the air above it does, the solution pulls it towards one particular relative humidity and holds it there.
For sodium chloride that value is 75.3% at 20°C, and it is remarkably stable across temperature: about 75.5% at 15°C and 75.1% at 25°C. That insensitivity is what makes it a practical reference. You are not creating 75% humidity; you are letting physics create it.
The values come from the NIST work on saturated salt solutions, which is why this is a genuine calibration procedure rather than a folk remedy.
What you need
- Table salt. Any kind; iodised is fine.
- A bottle cap, egg cup or similar small container.
- An airtight bag or sealed container the hygrometer fits inside.
- Water.
- Eight to twelve hours.
The procedure
- Make the slurry. Half-fill the cap with salt, then add water a few drops at a time and stir, until the salt is wet through but not submerged. It should look like wet sand. Undissolved crystals must remain visible — that is what "saturated" means, and a clear solution will not work.
- Seal them together. Put the cap and the hygrometer in the bag or container, keeping the sensor above the salt and not touching it. Seal it as airtight as you can manage.
- Leave it alone. Eight to twelve hours, somewhere with a steady temperature — not a windowsill, not near a radiator. Do not open it to check.
- Read it through the bag if you can. If you must open it, read within a few seconds: a capacitive sensor starts responding to room air almost immediately.
- Compare with 75%. The difference is your instrument's error at that point.
What the result tells you
| Reads | Error | What to do |
|---|---|---|
| 73–77% | Within ±2% | Good. Use as is. |
| 70–80% | Up to ±5% | Typical consumer performance. Adjust, or note the offset. |
| 65–70% or 80–85% | 5–10% out | Adjust if possible; otherwise label it and subtract. |
| Outside that | Badly out | Suspect a failed sensor or a botched test. Re-run before condemning it. |
Some digital hygrometers have a calibration button or a recessed adjustment. Many do not, and hair hygrometers usually have a small screw on the back that rotates the needle.
If yours cannot be adjusted, that is fine. Write "reads 6% low" on a label and stick it to the back. A known offset is as good as a correct instrument, and much better than false confidence.
The three things that ruin it
- Too much water. If the salt dissolves completely, the solution is no longer saturated and no longer holds 75%. Wet sand, not soup — this is the most common failure by a distance.
- Not long enough. The air needs hours to equilibrate and so does the sensor. A reading after two hours is a reading of a system still on its way somewhere.
- Temperature drift. A container that warms in afternoon sun and cools overnight never settles. Somewhere dark and stable, at ordinary room temperature.
A fourth, less common: a leaky seal. If the bag is exchanging air with the room, you are measuring the room.
Checking a second point
One reference point tells you the offset there and nothing about the rest of the range. Sensors can be accurate at 75% and several percent out at 30%.
If you need to know, run the test again with magnesium chloride (about 33% RH) or potassium chloride (about 84%). Two points give you an offset and some idea of the slope, which for anything where a few percent genuinely matters is worth the second evening.
When to bother
Annually for anything you depend on — a humidor, an instrument case, a curing space. Immediately if a reading looks wrong, or after the device has been through condensation or a spell of extreme humidity, both of which can permanently shift a polymer film.
And when you first buy it. Consumer hygrometers are frequently several percent out of the box, and finding that out on day one is considerably better than discovering it through a ruined box of cigars.