PAT testing?

[myxiplx]

Hi guys,

I realise this isn't particularly DIY related, but I haven't a clue where to find another forum as busy as this populated by qualified electricians, so I hope you don't mind me picking your brains.

I work in IT, and for the last 9 years (across two jobs) we have generally accepted that PAT testing fries computers. Despite our electricians always insisting that the tests are safe, we always experience a good number of failures, and no matter who we ask, we've never been able to find out why this happens. My understanding is that electronic equipment is only subjected to an earth leakage test, which should only affect the power supply. However we've seen blown power supplies, motherboards, processors and hard drives in the past. We've even seen a blown fuse in a mains lead after a PAT test.

We've always wanted to understand what's going on, mainly so we can try to reduce the number of failures we get, and this year I finally have a theory. I've no real experience with electrics however so I'd like to ask you electricians if this sounds reasonable:

1. The first key assumption is that we're going to assume that the kit is already faulty, and is going to fail the earth leakage test. This implies at least one faulty component within the power supply, although it's not bad enough to affect the operation of the computer.

I think this is the key to the problem since we can't think of any way a PAT test would affect a properly working computer, and previously we've always assumed the PAT test was breaking computers, and could not come up with any explanations.

2. Next, am I right in thinking that during the PAT test, the faulty component is going to be stressed much more than usual? Instead of just potentially leaking to earth via that component (which is unlikely to happen), we have deliberately created a circuit using the short, and are running power through it. Am I right in thinking this could cause the faulty component (and indeed other components within the power supply) to fail?

If so, I think this could explain why computers that have been running fine for ages fail after a PAT test. The test itself doesn't cause the fault, but it can highlight existing problems, and actually make them much worse.

I think this is then compounded by the faulty equipment being turned back on (either by the users, or by ourselves not understanding the fault when trying to repair the machine - we're not always told it's a PAT test failure...).

And I think it's this final step that has most likely caused the computer failures over the years - attempting to use a power supply that is potentially quite badly faulty could cause all kinds of problems within the computers.

Our plan to correct this is:
- When any equipment fails the PAT test, as well as placing a notice on the users keyboard, maintenance will physically remove the power lead.
- When repairing failed equipment, IT will replace the power supply without making any attempt to power on the computer.
- IT will then get the computer PAT tested, and only when it passes they will power on the machine to ensure it is working normally.

Do you guys think this sounds reasonable?

 

[comms]

I realise this isn't particularly DIY related, but I haven't a clue where to find another forum as busy as this populated by qualified electricians,

http://www.electricianforum.co.uk
or
http://www.theiet.org/Forums/forum/categories.cfm?catid=205&entercat=y

I think this is the key to the problem since we can't think of any way a PAT test would affect a properly working computer

An insulation resistance test could bugger a pc, they dont like 500v up em.

Our plan to correct this is:
- When any equipment fails the PAT test, as well as placing a notice on the users keyboard, maintenance will physically remove the power lead.
- When repairing failed equipment, IT will replace the power supply without making any attempt to power on the computer.
- IT will then get the computer PAT tested, and only when it passes they will power on the machine to ensure it is working normally.

Sounds ok to me.
Why not get one of the IT bods C&G2377 qualified (done as a 1 or 3 day course), you can then do your own PAT.

 

[ericmark]

My understanding is that electronic equipment is only subjected to an earth leakage test, which should only affect the power supply. However we've seen blown power supplies, motherboards, processors and hard drives in the past. We've even seen a blown fuse in a mains lead after a PAT test.

You are correct but one other test is the “Earth continuity” most machines have a milliamp setting for testing IT equipment as the printed circuit boards will not stand the 25 amp normally used.

The first key assumption is that we're going to assume that the kit is already faulty, and is going to fail the earth leakage test. This implies at least one faulty component within the power supply, although it's not bad enough to affect the operation of the computer.

I think this is the key to the problem since we can't think of any way a PAT test would affect a properly working computer, and previously we've always assumed the PAT test was breaking computers, and could not come up with any explanations.

2. Next, am I right in thinking that during the PAT test, the faulty component is going to be stressed much more than usual? Instead of just potentially leaking to earth via that component (which is unlikely to happen), we have deliberately created a circuit using the short, and are running power through it. Am I right in thinking this could cause the faulty component (and indeed other components within the power supply) to fail?

Other than inappropriate testing like using 500 volt to test insulation when it is know that IT equipment has built in filters with capacitors unable to take 500 volt as earth test already referred to nothing in a PAT test should stress a computer.

If so, I think this could explain why computers that have been running fine for ages fail after a PAT test. The test itself doesn't cause the fault, but it can highlight existing problems, and actually make them much worse.

I think this is then compounded by the faulty equipment being turned back on (either by the users, or by ourselves not understanding the fault when trying to repair the machine - we're not always told it's a PAT test failure...).

And I think it's this final step that has most likely caused the computer failures over the years - attempting to use a power supply that is potentially quite badly faulty could cause all kinds of problems within the computers.

Our plan to correct this is:
- When any equipment fails the PAT test, as well as placing a notice on the users keyboard, maintenance will physically remove the power lead.

The lead is PAT tested independent to the PC and are interchangeable so not much point

- When repairing failed equipment, IT will replace the power supply without making any attempt to power on the computer.
- IT will then get the computer PAT tested, and only when it passes they will power on the machine to ensure it is working normally.

Do you guys think this sounds reasonable?

I would think there is something wrong with the way they are being PAT tested. PC’s being only required to in the main be tested every four years can of course have been faulty for a long time but one big problem is the way they boot. The PAT test machine will power up the PC for a very short time with can mess up PC and is tested with monitor disconnected so you can’t see what is going on. In old days I had floppy which I would stick into drive and it would stop them booting but today often they have no floppy drive. To stick in a CD like for example the Norton anti-virus which on old ones at least stopped boot with question as to how to continue would stop PC’s from having power removed while in the middle of boot but to put these into PC requires them to be running to open door of CD drive. Many people who PAT test are given a very limited time sometimes they are expected to test over 100 items per day so including time to go between each item they have 4.8 minutes to complete test with around 2 minutes to complete an automated test this gives no time to mess around putting CD’s in drives.
I have PAT tested many PC’s and I can’t recall one failing to work after the test so I would think the tester is doing something wrong. I was the manager for the PAT testing and using the Robin top machine I was able to program in exactly what was tested when the ID number was entered into machine.
But many companies do not even bother keeping to same ID each time tested and rely solely on skill of tester to work out appropriate tests. Often the testers are only semi-skilled or electricians who have never been sent for PAT testing training so mistakes are easy to make.
In house testing is normally very good but where outside firms are employed the pressures they put there testers under often results in mistakes I have heard of electricians testing 250 items per day and anyone with a little sense knows this must in the main be just sticking on labels. With the Robin I used because the testing was auto recorded one had to plug in the machine even when you knew the item would pass. As when a patch lead is used to power PC and PC passes you know the lead is OK but I still had to test lead independent or there would be no record of test.
With manual testers this is not the case and testers and double enter results so two items are tested with one test.
The lead should have 25 amp continuity test but PC would be 200ma test so to do both together would mean using 25 amp test on PC. Also some PAT testers do not give the user the option to use the lower current test and the tester would need to use his installation test meters to do the test. It’s bad enough carting a PAT tester around with all the paperwork without adding other meters one would need a trolley to carry it all.
Not saying I can excuse bad practice while testing but I can understand it.

 

[ericmark]

Why not get one of the IT bods C&G2377 qualified (done as a 1 or 3 day course), you can then do your own PAT.

That also sounds good idea to me.

 

[OilLecky]

I agree with comms. Get your own people qualified to do the PAT.
An insulation resistance test on a 220 / 240 volt piece of equipment is carried out at 500 volts. Putting 500 volts into a box full of electronics is not a good idea. Even if this level of voltage does not destroy it, the effects of over voltage downstream of the Power Supply will cause a short life expectancy of many components. The Code Of Practice for In-Service Inspection and Testing of Electrical Equipment (16th edition IEE Regs) states, in Para 15.5 Insulation Resistance Testing, "This test should not be applied routinely to information technology equipment, unles it complies with BS EN 6090. Equipment not constructed to this standard may be damaged by this test."
Using my PAT tester, an AVO PAT4DV3, it tests IT (and all associated) equipment as follows:
1. Visual Inspection (I do that bit!)
2. Earth Continuity (0.1 Ohm or less.)
3. Operation at 230v (i.e. it switches on.)
4. Earth leakage. ( 0.5mA or less. The tester measures this leakage from line to earth while the computer is switched on at 230 volts.)
There is no way that I would put 500volts into an electronic box.

 

[Spark123]

I think the PAT4 measures the leakage as part of the operation test, this test replaces the insulation resistance test. The earth continuity test is carried out at 200mA.
The 16th edn regs are more associated with the fixed wiring whereas the code of practice for in service inspection and testing of electrical equipment is more to do with current using equipment.

 

[tim west]

Agree with Spark123 100ma or 200ma test for IT equipment The term IT can cover a multitude of equipment not just PC's such as telecoms equipment for instance and the leakage test can be substituted for the Insulation Resistance test in the case of IT.

 

[plugwash]

Why not get one of the IT bods C&G2377 qualified (done as a 1 or 3 day course), you can then do your own PAT.

This seems like a good idea to me, they will know things like what keys can be pressed on thier particular machines to stop the boot process and if thier test load consists of mostly IT equipment they should remember what tests are appropriate to use.

 

[OilLecky]

Spark & Tim. Yes the earth continuity test for IT equipment, using the PAT4, is 200mV when there should be a maximum resistance, of the earth line, of 0.1ohm. The test voltage is applied across the earth pin of the connector and the metal chassis of the equipment. It is not applied across the equipment internal components. As you know this is not the same as an insulation resistance check though, where a high voltage is normally applied across the equipment line and neutral to earth, so being applied across equipment internal components. With IT equipment the normal operating voltage is applied and any leakage to earth, in mA, is measured.
i.e. Earth Continuity & Earth Leakage are different and separate tests and must be carried out in the correct sequence.

Either way, lets not get bogged down with detail, I think that the IT men in question should get qualified to PAT test their own kit but they need to understand a certain amount of electrics / electronics to do a good job of protecting the equipment during testing. I have come across many IT people who neither understand how the electronics of a computer work nor how the software works! I don't really understand what they know..... or do!!!