Home   |   Technical Articles

Technical Articles

Withdrawal Force to IEC 60884-1: Table 19, Ten Insertions and the 30-Second Hold

2026-11-14

A socket-outlet has two opposite failure modes and one test that has to catch both. If the contacts grip too weakly, the plug heats at the interface and the terminal behind it runs hot. If they grip too strongly, the user pulls the outlet out of the wall box or abandons the appliance. IEC 60884-1 clause 22 answers both with one apparatus and a table, and the table is the part people quote without reading the clause around it.

This article works through the numbers in the current edition, the order of operations, and the six mistakes that turn a compliant outlet into a failed report.

The table has two columns that pull in opposite directions

IEC 60884-1:2022 Table 19 gives a maximum withdrawal force measured with a multi-pin gauge and a minimum withdrawal force measured with a single-pin gauge, plus a maximum for the single-pin gauge as well. The values are grouped by rated current and by number of poles.

RatingPolesMulti-pin gauge, max (N)Single-pin gauge, min (N)Single-pin gauge, max (N)
Up to and including 13 A2401.517
Up to and including 13 A3501.517
Up to and including 13 AMore than 3701.517
Above 13 A to 20 A2502.025
Above 13 A to 20 A3542.025
Above 20 A to 32 A2803.027
Above 20 A to 32 A3803.027

The single-pin figures apply only to the resilient earthing contact assembly. That footnote is easy to lose: the 1.5 N to 3.0 N floor is a statement about one contact, not about the outlet as a whole, and it exists because the earth pin is the one that must still make contact after a hundred insertions of a plug whose pin is at the low end of its tolerance.

The edition matters more than people expect

The 2022 edition re-cut the current bands. Where the earlier edition grouped at 10 A and 16 A, the current table groups at 13 A and 20 A, so a 16 A three-pin outlet moved from the middle band to the first band. Its multi-pin limit stays at 50 N, but the single-pin maximum dropped from the middle-band value to 17 N in one reading and to 25 N in another depending on which edition the laboratory still has on the shelf. The single-pin maximum in the middle band also moved from 18 N to 25 N. Two laboratories citing different editions can disagree about the same outlet without either of them making a measurement error.

Maximum first, and it takes ten insertions

The sequence in clause 22 is not a single pull. The outlet is fixed to the mounting plate of the apparatus so that the axes of the contacts are vertical and the entry holes face downwards, and the multi-pin gauge is inserted and withdrawn ten times. Only then is the gauge inserted again and loaded. The ten cycles are the test: a contact that grips at 60 N on the first insertion and at 30 N after ten may pass a single-pull check and fail the real one, because the spring has taken a set.

Loading is by gravity, with the gauge hanging from a hook that carries a principal weight and a supplementary weight. The total is the gauge, the hook, the tray and both weights together, and the plug should come free. If it does not, the supplementary weight is allowed to fall 50 mm onto the principal weight and the plug is judged to have separated if it then comes out. That drop is not a second chance in the sense of a weaker test; it is a way of overcoming static friction in the guide without adding mass.

Because the load is gravitational, the apparatus geometry is part of the method. The frame on the reference apparatus stands 550 mm tall inside, with 150 mm between the column axes. A short frame forces the weights to run against a guide, and every newton absorbed by that guide is a newton not pulling on the plug.

Minimum, and why it only gets thirty seconds

The single-pin gauge goes into each contact in turn, with the outlet in the same vertical position, and it must not fall out within 30 seconds. Each pole is a separate result: an outlet whose line contact holds and whose neutral contact drops the gauge has failed, and averaging the two is not a thing the standard allows.

Thirty seconds is chosen so that the reading is about grip, not about creep. A contact that lets the gauge go after five minutes has passed the clause as written and will still fail in service, which is why some product standards add an endurance sequence before this test. Where a customer specification calls for a hold of two minutes, that is an addition to the standard, not a reading of it.

The gauge is the measurement, and it wears

Both gauges are hardened steel with a surface roughness not exceeding 0.8 micrometre, and the pins are made to the maximum limit dimensions of a plug pin, not to the nominal size. That choice is deliberate. A plug pin at its largest is the worst case for the socket, so a gauge that represents the largest pin tests the smallest opening the socket is allowed to have.

Wear pushes the gauge the wrong way in both tests. A worn multi-pin gauge is thinner than the maximum pin, so it enters a tight socket more easily and the maximum withdrawal force reads low: over-tight outlets pass. A worn single-pin gauge is also thinner, so it grips less and drops out where a correct gauge would have held: good outlets fail. One worn gauge produces both false passes and false failures in the same afternoon, and neither shows up in the certificate.

A go/no-go check on the gauge pins against their drawing every six months, plus a roughness check on the same schedule, costs less than one retest.

Six ways the test goes wrong

  • A real plug instead of the gauge. A production plug has chamfers, plating and a pin somewhere in the middle of its tolerance band. It cannot represent the maximum pin and it cannot represent the minimum grip.
  • Skipping the ten insertions. The most common omission, and the one that most often converts a failure into a pass.
  • Leaving the hook and tray out of the total. A hook and tray can weigh 200 g, which is nearly 2 N. At the 1.5 N floor of the single-pin test that error is larger than the limit itself.
  • Mounting the outlet at an angle. The load must act along the pin axis. A plate tilted by a few degrees adds a sideways component and the gauge binds.
  • Testing one pole for the minimum. All poles, separately, including the earth contact.
  • No temperature record. The resilient contact is usually brass or phosphor bronze in a thermoplastic body, and grip falls as the body softens. A result taken at 30 degrees C and one taken at 15 degrees C are not the same result.

What the number is really protecting

The lower limit protects the temperature rise test two clauses later. A 16 A outlet that holds its plug at 8 N instead of 20 N may still pass a temperature rise measurement on a new sample, because the contact resistance of a clean, unworn contact is low. The same outlet after fifty insertions has a contact resistance high enough to add tens of kelvin at the terminal, and the temperature rise limit of 45 K is not generous enough to absorb it. Withdrawal force is the early indicator that the temperature rise test only sees once the damage is done.

For the gauges themselves, see our electrical safety test equipment range, or start from the product index.

close
Scan the qr codeClose
the qr code