I'm looking to ensure that some power supply units that are used are adequately earthed. The units vary in load from <1kW to 15kW and have type 1 insulation class. The manufacturer recommends a bonding conuctor rated to carry 80A for the larger units. These PSUs are disconnected and moved from time ot time and I'd like to establish a guidance note for checking earth continuity when re-connected.
HIgh current (e.g. ductor) tests used to be popular when testing steel conduit etc., but that seems to have fallen out of fashion of late - I suspect there's a school of thought that high current tests aren't without their risks and really only proved that something wasn't just hanging on by a fine strand - whereas a low current test plus visual inspection (e.g. all strands go into each terminal) can produce near enough equivalent results with less hazard during the test itself. I don't recall anything demanding high current tests in BS 7671 these days. That's for fixed wiring etc though - when it comes to individual items of equipment or assemblies that come under other standards, you'd have to check the particular standards concerned ... and of course any manufacturer's instructions. 80A seems quite a lot really, and might make quite a pop if something did fail (which presumably is a possibility, otherwise it wouldn't need testing). I suspect others here have better knowledge of other standards than me....
- Andy.
Hi Andy, Many thanks for your reply. Like yourself I am a little old school and that is why I'm checking my knowledge is current. Only the most expensive testers have the higher current tests these days and alot of people only use battery powered units that have the 200mA test. The test equipment complies with BS EN 61557-4. I'm reluctantly inclined to accept that the test is done at this very low current dispite my misgivings as I have experience of issues being highlighted by the higher current rating. You mention steal conduit. I tested some steal conduit years ago and the 25A test revealed that someone hadn't tightened a running coupler up properly.
I expect that the 200mA test has been standardised somewhere, as it is sufficiently high current to pick up an issue, but not high enough to case damage or possible risk of fire.
On the medical equipment I work with, IEC 60601 requires a 25A earth bond test with open circuit voltage not exceeding 6V, or 1.5x the highest rated current of the circuits being tested (whichever is greatest). Our equipment is a fixed piece of equipment (medical Linac) that weighs over 6 tons. This is carried out at installation and maybe once or twice during planned maintenance over the equipment lifetime (rated 10yrs but can be extended after certain tests have been carried out).
A new standard came out a while back for testing medical devices after remedial works/maintenance had been carried out - IEC62353, which gives an earth bond test current of 200mA, to prevent damage to boards, etc. I can imagine 25A through a cable screen due to a poor earth would like damage the cable.
My understanding is that the main earth bonds are not generally removed/replaced during the equipment life, but ones to a power supply, card rack, etc. are likely to be during any planned maintenance or repairs. In these instances, we are only checking the affected bonds, not all of the main ones (unless they were removed).
I can see why bond tests on conduits or trunking could cause an issue, especially where you can't see the entire section being tested because it goes under floors, into voids, etc. and any heating caused by high current through a bad connection could cause a fire.
I remember on PA equipment racks used on the Jubilee line extension (over 20 years ago), that we would get smoke from the cage nuts in the equipment racks during bond testing, which used to worry those observing, as they didn't realise we had unplugged all of the circuit boards.
My personal preference would be the high current bond test as in 60601, but ensure that any signal cables and PCBs are disconnected so that only power earth connections are tested and not grounds/signal earths.
I remember on PA equipment racks used on the Jubilee line extension (over 20 years ago), that we would get smoke from the cage nuts in the equipment racks during bond testing, which used to worry those observing, as they didn't realise we had unplugged all of the circuit boards.
On the same project, network equipment racks had no problems with the 25 A test, but they had been designed specifically with this test method in mind, including consideration of parallel paths. The test was carried out prior to installation of electronic equipment.
I would add that, where resistances to be measured are less than 0.1 Ω, modern Kelvin-method ductor testers can provide higher accuracy and resolution than the old 'conduit ohmmeters'.
However, in the general case of BS EN 50310, the latest version of the standard (BS EN 50310:2016+A1:2020) specifies BS EN 61557-4 or BS EN 61557-5 for test instruments, the tests therefore being carried out at as little as 0.2 A (into 0.1 Ω) and with a resolution of up to 0.1 Ω, so the accuracy of a ductor tester is not needed these days for simple conformity, and "<0.1 Ω" (or relevant steps of 0.1 Ω) appears to be OK. (It may, however, be specified by a designer for a particular purpose in specialist installations.)
My personal preference would be the high current bond test as in 60601, but ensure that any signal cables and PCBs are disconnected so that only power earth connections are tested and not grounds/signal earths.
Yes, testing "as new" racking system prior to the installation of equipment in the rack, and similarly bonding provision in a particular equipment room prior to the installation of equipment, both of which could be carried out with a high current test under controlled conditions.
I remember on PA equipment racks used on the Jubilee line extension (over 20 years ago), that we would get smoke from the cage nuts in the equipment racks during bond testing, which used to worry those observing, as they didn't realise we had unplugged all of the circuit boards.
On the same project, network equipment racks had no problems with the 25 A test, but they had been designed specifically with this test method in mind, including consideration of parallel paths. The test was carried out prior to installation of electronic equipment.
I would add that, where resistances to be measured are less than 0.1 Ω, modern Kelvin-method ductor testers can provide higher accuracy and resolution than the old 'conduit ohmmeters'.
However, in the general case of BS EN 50310, the latest version of the standard (BS EN 50310:2016+A1:2020) specifies BS EN 61557-4 or BS EN 61557-5 for test instruments, the tests therefore being carried out at as little as 0.2 A (into 0.1 Ω) and with a resolution of up to 0.1 Ω, so the accuracy of a ductor tester is not needed these days for simple conformity, and "<0.1 Ω" (or relevant steps of 0.1 Ω) appears to be OK. (It may, however, be specified by a designer for a particular purpose in specialist installations.)
My personal preference would be the high current bond test as in 60601, but ensure that any signal cables and PCBs are disconnected so that only power earth connections are tested and not grounds/signal earths.
Yes, testing "as new" racking system prior to the installation of equipment in the rack, and similarly bonding provision in a particular equipment room prior to the installation of equipment, both of which could be carried out with a high current test under controlled conditions.
We're about to take you to the IET registration website. Don't worry though, you'll be sent straight back to the community after completing the registration.
Continue to the IET registration site
