What’s wrong with the Electrical Safety Testing section of the new AS/NZS3551

Posted by

What’s wrong with Appendix B of AS/NZS3551:2012

In a word, LOTS!


Firstly all of the performance specifications have been left out from the 2004 version. The accuracy tolerance for the Ground Wire Resistance Test is missing along with the voltage requirements for Insulation Resistance Testing and the regulation required for Mains Contact Current Testing.


Contact Current Testing has been added but it is flawed. Applying Ohm’s Law one can see that you have a 1Kohm Measuring Device in parallel with 0.3Ohm medical device return earth and to get 100uA fail, 100mV would have to be on the protectively earthed case but this would generate 333mA current flow through the medical device being tested. This then increases to 1.5A when the neutral is opened.

To really add insult to injury, the last Contact Current Test opens the earth line so you now have exactly the same test as the Earth Leakage Test but under Contact Current the fail level is 500uA, (same as earth leakage fail level in 2004), and later under Earth Leakage Tests the fail level is 5000uA. How you can do an identical test and have two different fail levels, a factor of 10 times different in the same standard has everyone mystified.


Applied Part Current Testing & Mains Contact Current Testing made perfect sense as API was current sourced from the medical device & MCI was current sinking into the medical device.

To open the neutral and earth with API testing also made sense because you are presenting the applied parts with the only return path via the measuring device to ground, what is commonly referred to as the worst case condition.

Similarly, the worst case condition for MCI testing is having all of the mains conductors present as this gives the applied parts the maximum pathways for the current to sink into the medical device.

So adding for MCI testing a Neutral Open Test and an Earth Open Test is a useless waste of testing time.
Then of course we have to keep up with fashion and change the names of the tests. It took me ages to change my thinking in the 1980s to calling what were referred to as Patient Circuits to Applied Parts.

Now they want to call API testing, ‘Patient Leakage Current’ and are you ready for this, MCI testing they want to call ‘Patient Leakage Current – Mains on Applied Parts’.

I cannot believe that our smart engineers and scientists have disappeared and we have to follow IEC like sheep.
For a more detailed review of what I see as problems with AS/NZS3551:2012 along with my recommendations to fix them, email me and request a copy.

The paper I will send you was the one I delivered at the SMBE (NSW) Training Workshop at Mudgee NSW, March 2013.

Kevin E. O’Brien


Interesting comments at one of the Melbourne Workshops

Posted by

Interesting Comments from Mike Flood at a Melbourne 3551 Workshop


Mike Flood is a very experienced Bio-Medical Engineer & is a committee member for the new AS/NZS3551:2012. He has been travelling around Australia conducting workshops to explain the new standard. Mike’s presentation was mainly focused on giving an overview of the standard & my impression was that Mike was heavily involved in the formulation of the procedures & Risk Management sections of the standard whilst other committee members were responsible for Appendix B, the electrical safety testing section of the new standard.
My area of interest as a member of the Bio-Medical industry for over 45 years and a designer of electrical safety test equipment for medical devices for 20 years is Appendix B. Below are some interesting comments and a lot of other problems not discussed at this time were raised during the SMBE (NSW) Conference at Mudgee NSW in March 2013.

Questions for Mike Flood prior to the Workshop:

A: Mike’s response, “It is only common sense to do it that way”.
My Comment: I agreed with Mike & pointed out that even though it was left out of 3551:2012 in Appendix B, it is still in Appendix K of 3551:2012 & is allowed under IEC 62353.

Q: The Contact Current test when applied as drawn touching protectively earthed parts doesn’t work as can be seen by just applying Ohm’s law?
A: Mike didn’t respond directly but left me with the impression that I had raised a valid point and that work had to be done to clarify this test.

Q: Why are we changing the terminology for Applied Part Current and Mains Contact Current to Patient Leakage Current and Patient Leakage Current – Mains on Applied Part.
A: Again Mike didn’t respond directly but I believe that whatever IEC does we tend to follow them like sheep even when in this case Mains Contact Current is technically not leakage current but is current sinking into the medical device from an external power source.

Three year changeover:
Mike commented that when a new standard is released, they consider the old and new standard running together for a three year period. At the end of that period, the old standard is phased out and the new standard applies.

During the workshop:
Q: How can we have pass/fail thresholds set by IEC Standards when they run at 220V mains and we run at 240V mains.
A: That is not correct, Australia as of 2000 is 230V +10%/-6% and Europe is the same.
My comment: Mike was right after I did some research but that doesn’t take into account that this is a field test standard and testing may be done with 240V mains or higher so Leakage Currents for example will be higher.

After the workshop:
Q: On the basis of Risk management as you have outlined, the BME can decide what tests apply to a medical device & leave out some of the tests.
A: Mike’s response, “That is correct”.


Blowing the LID OFF AS/NZS3551:2012 + Amendment 2

Read More

MEMIE calls for the HE-003 Committee behind AS/NZS3551:2012 to be replaced!

Read More