What Is IP67: Dust-Tight, Thirty Minutes, One Metre
IP67 is a test result, not a promise about water.
The two digits describe two separate tests. The first covers dust. The second covers a short dip in fresh water. Neither digit says the product can be used underwater.
Both come from IEC 60529, the standard that defines the IP code. The standard also sets the exact conditions, and those conditions are where most supply disagreements start.
The Short Answer
6 means dust-tight. The test pulls a vacuum inside the enclosure and holds it in a dust cloud for up to eight hours. No dust may get in.
7 means the enclosure survives a short dip. The test puts it under water for thirty minutes and then checks whether any water got in that could cause harm.
That is the whole rating. One dust test, one dip, both on samples chosen by the factory.
Three limits follow, and a buyer should know all three before a quote is signed. IP67 is not submersible. IP67 is not a pressure rating. And IP67 is not permanent.
Two Digits, Two Very Different Scales
The digits look like a scale from 1 to 7, and they are not one. Each number stands for a different test rig, not a step on a ruler.
| Code | What has to happen | How it is tested |
|---|---|---|
| 5 | Dust may enter, but not enough to stop the product working safely | Dust chamber, partial vacuum, up to 8 h |
| 6 | No dust at all gets in | Dust chamber, vacuum inside the case, up to 8 h |
| 7 | No harmful amount of water after a short dip | Immersion, 30 minutes, fresh water |
| 8 | Continuous immersion, depth and time set by the maker | Agreed test, longer or deeper than 7 |
An X in either place is not a zero. It means no level was assigned, usually because the test was not run.
So IPX7 says nothing about dust either way. A seller writing it for a dusty job has answered half the question.
The jump from 5 to 6 is the one that costs money. A 5-rated housing may let a little dust in. A 6-rated housing may let none in, and the test is built to prove it, with a vacuum pulling dust toward every joint.
That is why a “dust resistant” claim and a 6 are not the same thing to buy. One is a description. The other is a chamber test with a number attached.
The Conditions Behind the Seven
The immersion test is more specific than the marketing line suggests, and the details decide whether your product actually passed it.
Standard says the water is fresh. It says the dip lasts thirty minutes. And it defines depth in a way that surprises people.
The depth is judged in one of two ways. Either the lowest point of the enclosure sits 1 000 mm below the surface. Or the highest point sits 150 mm below the surface. Whichever gives the deeper submerge is the one that counts.
So “one metre” is not measured to the middle of the product. A tall enclosure is therefore treated differently from a flat one. A factory quoting “one metre” without the second clause may be testing the easy case.
The standard tests samples, not a production run. A report covers the units the lab received, and says nothing about unit ten thousand.
The dust side has its own detail. The chamber draws air out of the enclosure to create a partial vacuum, then lets dust-laden air back toward every seam. The run can last up to eight hours.
Why 67 Is Not a Licence to Dive
The most useful sentence in the whole standard is a warning about how the numbers stack. Water ratings are not cumulative past the jet tests.
A product that passes the dip test does not automatically pass the jet test. The standard says so directly, and it shows how to write a claim that covers both: list them together, as IPX5/IPX7.
That single rule answers a common support ticket. A unit rated IP67 fails after a hose or a pressure washer, and the owner calls it a defect. The unit was never tested against a jet. It was tested against still water.
Two more limits sit in the same paragraph of the standard. The water is fresh, and fluids such as salt water, oil and solvents are outside the rating. A seawater dip is a different test, and it usually needs a different seal material.
Temperature is the other quiet failure. A sealed box carried from a cold truck into a warm room pulls moist air inward as the inside pressure drops. A vent that lets air pass also lets vapour pass.
Condensation inside a sealed unit is not covered by the rating. It is still the most common reason a rated device fogs its lens.
Thirty minutes is also a test duration, not a service life. A gasket that passes on day one can harden in sunlight, flatten under a screw, or take a compression set after a few hundred closures.
IP68 Is Not One Better
Buyers often treat 68 as the next step above 67. The standard does not define it that way.
For 8, the manufacturer states the depth and the time, and the test is run to that statement. The standard expects the depth or the time to exceed the 7 test, and it allows extra conditions such as temperature cycling before the dip.
So an IP68 claim without numbers is incomplete. Two suppliers can both say IP68, one testing at 1.5 m for thirty minutes and the other at 3 m for an hour.
The same caution applies across systems. A NEMA 6 enclosure is often listed against IP67, but the two standards test different things, and the mapping is not an equivalence.
Where the Rating Actually Dies
The rating belongs to the enclosure as tested. Change one opening, and the claim needs a new test.
In production the failures cluster in the same few places. Each one is a design decision that a buyer can see on a drawing before a tool is cut.
| Opening | How it fails | What to settle early |
|---|---|---|
| Port cover | Lint on the seal, or a cover left loose by the user | State that the rating applies only when closed |
| Gasket | Too little squeeze, or a gasket that takes a set | Groove depth and squeeze ratio on the drawing |
| Ultrasonic weld | A void along the seam that only shows under vacuum | Weld bead width, and a leak test on every unit |
| Pressure vent | Added for sound, then assumed to be sealed | Name the vent grade, and keep water out of its path |
| Cable gland | Cable pulled sideways opens the seal | Strain relief, and a gland rated for the cable size |
The gasket is a spring, and it needs a defined squeeze. Too little and the seal leaks at the first pressure change. Too much and the rubber takes a set and never springs back.
That is why groove depth belongs on the drawing next to the material name. A supplier who receives only the letters “IP67” will choose both.
A vent is the one that catches people. A small membrane lets air pass so a speaker sounds right, and it keeps liquid water out. It does not make the unit a sealed box, and it has its own pressure limit.
A leak check on the line is a process control, and it sits beside the workmanship rules in the workmanship standard. One inspects a joint, the other inspects a seam.
One more habit is worth building. Ask for the test report, not the badge. The report names the sample, the conditions and the lab. It is also the only document that answers the question a customer will ask in a year.
The Leak Test You Can Run Every Day
The immersion test takes thirty minutes and a tank. Most lines cannot do that to every unit, so they test with air instead.
The unit is pressurised, sealed, and the pressure is watched for a set time. A drop means a leak. The test takes seconds and leaves no water behind.
Two numbers decide the result: the pressure applied and the decay allowed. Both are set by the factory. Neither comes from the standard.
That is the trap, and it is worth naming. An air test matches a water dip only if someone proved the match once, on units known to pass and units known to fail.
Without that proof the line is measuring its own convenience. Ask for the correlation record when you ask for the test report. It is one page: the leak rate that corresponds to a failed dip.
Salt, Sun and the Seal
The rating says nothing about corrosion. The dust test uses talc, and the water test uses fresh water. Neither one ages the seal.
Two forces do that in the field. Sunlight breaks down many rubbers. Salt attacks the metal around the seal until the gasket has no flat surface left to press against.
So an outdoor product needs three answers, not one. The IP code, a stated gasket material, and a salt fog result on the assembled unit.
A page that lists only the IP code answers the smallest of the three questions.
Sixty-Five, Sixty-Six, Sixty-Seven
Three codes get mixed up in quotes, and the difference is the shape of the water, not the amount.
65 adds a jet from a 6.3 mm nozzle at low pressure. 66 uses a 12.5 mm nozzle at a higher flow. Both are aimed at a product being hosed down.
67 swaps the hose for a tank. The product is not sprayed at all. It is lowered in and left there.
That is why the three do not replace each other. A unit built for a dip may have no defence against a directed jet. A jet arrives with force, and it finds a seam that still water never moved.
Outdoor gear that gets rained on and washed sits in the 65 or 66 world. Gear that may be dropped in a puddle needs the 7 as well, and a claim that covers both is written with a slash.
Ask which test the product actually passed. A quote that lists all three codes without saying which one was run is a wish list, not a result.
Six Lines for the Order
For a product sold in the EU, the same result also sits behind a declaration of conformity. The CE file guide shows how that file is built.
An IP claim that is not written down is a marketing line. Six lines turn it into something a lab can check and a claim can lean on.
One. Write the exact string, not a paraphrase. IP67, or IPX5/IPX7 if the product meets jets as well as dips.
Two. Name the standard and the edition. IEC 60529 owns the definitions, and the IEC’s own summary points back to the full text for the test conditions.
Three. Say whether this is a type test on samples or a leak check on every unit. They answer different questions, and only the second catches a bad gasket on a Thursday.
Four. Put the depth rule in the report. The published test conditions measure from the lowest point at one metre, or the highest point at 150 mm, whichever is deeper.
Five. State the condition of the unit during the test: covers closed, fresh water, and how long it dries before the check.
Six. Require a retest after any change to the seal path. A new port, a different gasket rubber or a moved screw is a new enclosure.
The electrical side is a separate file. A sealed enclosure can still fail a hipot test, because water and voltage are different hazards.
These lines fit the same brief you already send used for every other requirement, so nothing new has to be invented.
Questions Buyers Ask
Is IP67 waterproof?
It is water resistant under one defined test: thirty minutes, fresh water, at the stated depth. That is narrower than the word waterproof suggests, and the standard avoids the word for that reason.
Can I use an IP67 product in the sea?
The standard’s water is fresh. Salt water, oils and solvents sit outside the rating, so a seawater use needs its own test and often a different seal material.
Does IP67 mean it survives a pressure washer?
No. Water ratings stop being cumulative above the jet tests, so passing the dip test says nothing about a jet. A product that meets both is written as IPX5/IPX7.
Is IP68 better than IP67?
It is different. For 8, the maker states the depth and the time, and the test follows that statement. A claim of IP68 without those numbers cannot be checked.
Does the rating cover a phone dropped in a puddle?
Only if the water is fresh, the time stays short and the seals are intact. A drop can crack a housing or lift a gasket, and the rating assumes an undamaged unit.
Does IP67 cover condensation inside the case?
No. The rating covers water entering from outside under test conditions. Moisture that arrives as vapour and condenses on a cold lens is a different problem, and it needs a vent or a desiccant.
Send me the enclosure drawing or the IP claim from your quote. I will tell you which openings still need a rule, whether the depth condition is written down, and what to add before a tool is cut. No cost, no obligation.
Check My Enclosure
Questions & Comments
We read every one and reply within 24 hoursIf the drawing says IP67 and nothing else, send me the line. I will say which openings still need a rule before a tool is cut.