Interesting video on testing type "A"'s RCBO and issues with MFT 's

[LukeD]

 

[loz2754]

Watched it yesterday.
The main thing they were plugging was that BS7671 tables for testing RCDs only apply to type AC RCDs. Therefore, the ECA chap was recommending that we test all RCDs on the type AC setting on our test equipment, regardless of the type of RCD under test.

Then the Lewden technical chap added that, for a type A, we should test first on the AC setting, then test again on the A setting. The results we record would be the results from the type AC, as this is what BS7671 is asking for on the schedule of year results.

Interesting discussion indeed.

They also touched on where in the circuit we should be testing RCDs....

 

[Mikegh]

I'll watch when I get a chance

Interested to go through it all in detail

 

[GBDamo]

I've watched this twice now and it is a bit of a WTF moment.

It looks like everyone is looking at each other to figure out how the hell we ended up in this mess.

Manufacturer is getting massive returns as the devices are "failing" their answer is they're not failing by the specs that we use but don't publish.

MFT maker just looking blank.

The ECA chap made most sense with his "just test AC type as its the only test we have published targets to test against"

More comedy from the electrical industry.

 

[buzzlightyear]

It was watching what came first the dinosaur or the chicken, WTF every body was pluxed.

 

[Mikegh]

I've watched this twice now and it is a bit of a WTF moment.

It looks like everyone is looking at each other to figure out how the hell we ended up in this mess.

Manufacturer is getting massive returns as the devices are "failing" their answer is they're not failing by the specs that we use but don't publish.

MFT maker just looking blank.

The ECA chap made most sense with his "just test AC type as its the only test we have published targets to test against"

More comedy from the electrical industry.

Well I'm not surprised to hear this

I reckon there's definitely a knowledge gap to be filled in regard to testing of A and B type

Thankfully we don't series the RCDs over here which complicates matters a little further


Now maybe everyone is fully up to speed but I could see where I was mistaken a few minutes into the video

 

[SparkySy]

What gets me is that megger are basically selling a lie and their tester can't properly test the type A B or F properly as it doesn't meet the manufacturer standard for testing them but does meet their standard for testing them wtf!
And that the manufacturer doesn't want to give up their standard information as they have probably paid a fortune to acquire it so they can make things to those standards.
That has really made my brain hurt!

 

[buzzlightyear]

What gets me is that megger are basically selling a lie and their tester can't properly test the type A B or F properly as it doesn't meet the manufacturer standard for testing them but does meet their standard for testing them wtf!

Megger are supposed to be top brand, me if I worked there I would start throwing the testers out the window and starting again.

 

[LukeD]

Megger are supposed to be top brand, me if I worked there I would start throwing the testers out the window and starting again.

No one is talking to each other in the industry to ensure things are rolled out properly

 

[GBDamo]

What gets me is that megger are basically selling a lie and their tester can't properly test the type A B or F properly as it doesn't meet the manufacturer standard for testing them but does meet their standard for testing them wtf!
And that the manufacturer doesn't want to give up their standard information as they have probably paid a fortune to acquire it so they can make things to those standards.
That has really made my brain hurt!

I think the tester works it's just that there is nothing to relate the results to as that I the manufacturer's proprietary information and, for all we know they may well al be different.

 

[SparkySy]

The megger guy states that they don't test to the same standard as the manufacturer and have a different standard that it complies to, he also says that their margin for a range is huge and they try and shrink the range to make testing more accurate.
As stated above its about time we all got on the same team and started communicating properly after all we are all supposed to be after the same outcome!
We as sparks don't want to be returning goods, the manufacturer doesn't want loads of returns, and our tester should be able to test to the same standard as the manufacturer and all the test data should be widely available surely?!



I'll wait for the optimistic GIF!

 

[ElectroChem]

I quite liked the ECA guy's point - we're not out here popping fuses or even tripping breakers 'just to check they work' - we trust that reputable manufactures have made a product that will work as specified when it needs to.
If we can't be sure an RCD will work properly unless it's tested every 3/6/12/whatever months, are we really sure it will ever work?

 

[WallaceP]

I never understood why there is a need to time the rcd performance. We are happy to measure the earth fault loop impeance and from that say the mcb will operate correctly. I wonder if any other coutries feel this level of detail is needed.

is this a hangover from back when RCDs were seen the way AFDD’s are thought of now with not a lot of confidence in there ability.
Even when recording the results of an RCD time test where you do 7 tests but have one value to record and a tick seems funny to me.

 

[WallaceP]

Just watched the full video. Actually pretty good.
One concern I have after watching this is disconnection times. They spoke of capactive loads slowing the disconnection time done.
Now what happens in a real world fault where someone is relying on an RCD for fault protection but there is a large Cap load. Will they be able to get the disconnection time the need for safety?

 

[SparkySy]

That is a very valid point, after my experience on Saturday with Some Schneider RCBO'S I would say possibly not. I was given some remedial work by a company that had an Eicr done and several RCBO'S were reported as not meeting additional protection requirements. I like to test everything thoroughly so went through the sequence at the output terminals and no they didn't trip within 40ms. So removed outgoing circuit conductors and tested again, now tripping @ 32ms and 12ms. These were type AC RCBO'S on office RFC's with several PC's plugged in. All but 1 circuit were very similar once the circuit was disconnected. I asked what they wanted to do and their representative asked me to change them anyway. So I have now replaced them with type A RCBO'S, all of them now meet the required times.
This got me thinking was this either capacitors or some DC leakage blinding the older AC type rcd?
Unfortunately ran out of shut down time to experiment further!
Sy

 

[Pretty Mouth]

Now what happens in a real world fault where someone is relying on an RCD for fault protection but there is a large Cap load. Will they be able to get the disconnection time the need for safety?

I was told that when this happens, the RCD does disconnect in its usual time, but the capacitor fools the tester into thinking the circuit is still live until it has discharged.

Can someone more knowledgeable than me please confirm or deny that this is what happens?

 

[WallaceP]

I was told that when this happens, the RCD does disconnect in its usual time, but the capacitor fools the tester into thinking the circuit is still live until it has discharged.

Can someone more knowledgeable than me please confirm or deny that this is what happens?

Thanks! Thats good to know.
I feel testing is way behind with current tech. The guy from meggar touched on it too, when saying they aim for 5% for earth loop testing but the requirement is to be within 30%. We are worrying about 0.1 ohm and the machines we are getting the values from can way out on this. Don't get me started on some 2 lead No trip tests. completely useless....

 

[timhoward]

Finally watched it, interesting!
I was wondering at first how the guy from Megger could possibly come out of this looking good, without losing the "You need one of our newest and shiniest machines to test type A RCD's" selling point in spite of the both test and acceptable results not matching the manufacturing standards.
To give him his due, I didn't see the "Test them twice" line coming. Pure genius. Once for BS7671. And once to justify the extra cash we've just spent for a test that isn't actually required.

 

[Risteard]

Finally watched it, interesting!
I was wondering at first how the guy from Megger could possibly come out of this looking good, without losing the "You need one of our newest and shiniest machines to test type A RCD's" selling point in spite of the both test and acceptable results not matching the manufacturing standards.
To give him his due, I didn't see the "Test them twice" line coming. Pure genius. Once for BS7671. And once to justify the extra cash we've just spent for a test that isn't actually required.

To be fair that advice has been around for a while, e.g. with a Type B RCD to test as AC, A and B.

Although the ECA view on this was interesting to note, and suggests that all of these tests really aren't all needed for BS7671 compliance.

 

[goasis]

An interesting video, with intelligent input. Does make you think about how much you do "because that's what we do" without thinking too much about it.