SFC on a Class II ME EQUIPMENT with Class I PS

[eldercosta]

ME EQUIPMENT (MEE) is Class II powered by a Class I External PS, plastic enclosure, with 2xMOPP insulation between mains and output, ouput also galvanically isolated from the PE. The PS is part of the MEE in accordance with 7.9.2.3. Class I PS chosen for EMC reasons. The PS has a IEC 320-C14, it will be connected to a protectively earthed outlet through a IEC 60601-1 compliant power cord in a Professional Healthcare Environment.

PE is considered FE in accordance with 8.6.9. The problem is FE disconnection is not explicitly listed as a SFC; on the contrary, apparently it is considered NC. The PS has Y1 capacitors between L/N and PE for EMC. With PE connected, current coupled from L/N will flow through the Y1 capacitors to earth and the current flowing from primary to secondary/metallic accessible parts will be much lower than the 100uA limit. With FE disconnected, this current will be much higher than the 100uA NC limit.

The only metallic accessible part in the MEE is a USB connector, restricted to connecting USB flash drives and that will have a rubber cover most of the time. I understand 8.4.2 c) would allow us to circumvent the limit, and the USB port will easily pass the test introduced by AMD2:2020 (30Vrms over 10kR corresponds to 3mArms, much higher than the 500uA touch current limit - go figure), but we would rather avoid it if we may argument FE/PE disconnection in this context could be treated as SFC.

Suggestions on how to handle this?

Thanks in advance for your comments.

 

[Peter Selvey]

If a SPS (separate power supply) is Class I then the earth is protective and remains protective even when connected to the MEE. The MEE is Class II in name only, it doesn't do anything because classifications (I, II) are really only intended for the equipment that physically or directly connects to mains supply, and is not something that normally would be applied to secondary equipment. For historical reasons, IEC 60601-1 kept it there even for non-mains equipment (long story), but if you look carefully through the standard, there are no requirements related to Class II for non-mains MEE, apart from the IFU where you have to declare the classification. Otherwise it is meaningless.

In testing, the system is Class I and therefore opening the earth will be SFC, and the SFC limit should apply (500µA). Clause 8.6.9 does not apply because the system is Class I. The clause explicitly references the POWER SUPPLY CORD and also the MAINS PLUG, these are things that are only relevant to equipment that directly connect to the mains supply (100V, 120V, 230V etc.).

 

[eldercosta]

but if you look carefully through the standard, there are no requirements related to Class II for non-mains MEE, apart from the IFU where you have to declare the classification. Otherwise it is meaningless.

Could you elaborate or point directions?

In testing, the system is Class I and therefore opening the earth will be SFC, and the SFC limit should apply (500µA). Clause 8.6.9 does not apply because the system is Class I. The clause explicitly references the POWER SUPPLY CORD and also the MAINS PLUG, these are things that are only relevant to equipment that directly connect to the mains supply (100V, 120V, 230V etc.).

I anticipate the test lab to disagree with us stating it must be tested as Class I when it is Class II. Would you have suggestions on how what approach to take in this context?

In the past we have had NCs (nonconformities) due to diverging (us vs test lab) interpretations on some subclausees but using parts of the standard we demonstrated we are conformant, but in this particular issue I am having some difficulty to find basis to our arguments so any suggestion would be useful.

 

[Peter Selvey]

When you see instances in the standard that mention Class II, you will find in the title or the text of the clause the word "MAINS" or "SUPPLY MAINS" or similar which are defined terms and don't apply to non-mains equipment, but you have to be on your toes so to speak.

For example, Clause 7.2.6 says "ME EQUIPMENT shall be marked with the following:" and then lists things like input voltage, nature of supply, and the Class II symbol (double box). These items are often forced on manufacturers of "main units" that are powered by a separate power supply. But if you look closely, the title of the clause is "Connection to the SUPPLY MAINS", which means: all of items in that clause including the rated voltage are only required for equipment that actually connects to mains. And that makes sense if you think about it. The reason why test labs get this wrong is the IECEE CB scheme which insists on making TRFs that garble the standard and often omit clause headings, and the labs use those without referring to the standard. I'm old enough to have worked with the 1988 edition of IEC 60601-1, which had a Table 2 that clearly showed that non-mains equipment does not need to be marked with rated voltage, current, nature of the supply (e.g. dc symbol) or the Class II symbol. However, even back then the CB scheme TRF did not include any reference to Table 2, leading to many labs erroneously pushing a bunch of inapplicable marking on manufacturers, a practice that remains. As far as I know, all the other references to Class II include something similar where it can only apply to devices powered from the mains. Note that some dc powered devices are genuinely mains powered, e.g. if they get their supply from 12V or 24V bus in an ambulance. In that case, all the Class II stuff is applicable. But for things powered by a separate power supply, not relevant.

The critical clause for separate power supplies is 8.2.1. This clause gives you two choices: either treat the whole thing as the MEE or handle under Clause 16 as an ME SYSTEM. If you consider the (SPS+main unit) as the "MEE" then it has a single classification which must be Class I if the SPS connects to mains earth.

Clause 8.6.9 is a highly specialized and rare case where a manufacturer of a mains powered MEE (direct to mains) wants to use the earth wire in the power supply for EMC but also wants to retain the Class II classification and treat the earth as FE only. It is frequently misread but what the standard says is that you must have 2MOOP between the FE and all accessible parts (not, as is often misunderstood, between the FE and mains). That can be tough to implement, so it is really only used in rare cases, for example, there may be Y caps or an EMC filter directly at the appliance inlet, but the FE doesn't go anywhere else and is definitely not connected to the secondary in any way, shape or form.

 

[eldercosta]

The critical clause for separate power supplies is 8.2.1. This clause gives you two choices: either treat the whole thing as the MEE or handle under Clause 16 as an ME SYSTEM. If you consider the (SPS+main unit) as the "MEE" then it has a single classification which must be Class I if the SPS connects to mains earth.

Hi, Peter, still struggling with this. Please bear with me.

So, what you are saying is classifying my MEE as Class II makes no sense? Or it may be classified as Class II but it must be tested as Class I because the external PS (considered part of the MEE) is Class I?

Regarding subclause 8.6.9, I think I was misinterpreting it, as I was considering the PS's internal wiring (actually, pcb traces + the IEC 320-C14 PE) as the "internal" screen.

 

[Peter Selvey]

Under 8.2.1 there are two paths you can choose (freely):

(Option 1) the MEE is just the main unit, and the power supply is a separate object not subject to the standard (outside of Clause 16), but the system of the main unit + PSU needs to be evaluated to Clause 16 as an ME SYSTEM. This means, for example, when applying say Clause 7.2 for marking, you don't apply this to the power supply, since it is not in the scope of test. The power supply manufacturer deals with standards compliance under their own responsibility. In this case the Main Unit can be Class II, but as mentioned, this is a toothless classification without any actual impact apart from declaring it as Class II in the operation manual.

(Option 2) the MEE is the Main Unit plus the power supply, and there is no ME SYSTEM. In this case, when (for example) Clause 7.2 is applied, it is also applied to the power supply, so the test lab for example should be checking the marking on the power supply meets the requirements. Also in this case, the MEE is clearly Class I, since the power supply is Class I. There is no separate classification for the main unit.

In reality, Option 2 only fits for custom power supplies where the manufacturer of the medical device takes responsibility for the power supply as well (even if it is manufactured by a third party). The problem is that type testing is more than just a one time test on a single sample, it also involves manufacturing, production controls via the definition of "manufacturer" and also "type test" and the concept of a single test being "representative" which can only exist if there are design and production controls in place, otherwise it's just a random result without any meaning. If you buy a power supply from a supplier with no contracts, agreements in place, you can't plausibly be the "manufacturer" of the power supply so the whole thing falls apart. That said, most test labs don't understand this and will happily test under Option 2.

IEC 60601-1 has messed all this up so it's no surprise that it's confusing. The correct way to handle this is Option 1, but with a twist: you should declare assumptions about the external power supply that are reasonable, and then test your device based on those assumptions, not with a particular (single) sample of the power supply. For example, in a temperature rise test for your device, maybe the design is sensitive to the actual voltage coming in to the unit. You could test with a random sample power supply which is say nominally 12Vdc, but the actual sample is 12.11Vdc. That random test is really meaningless. Rather, you should look at the power supply specification, which could be say 12V±5%, and then test your device with an artificial power supply that covers 11.4V to 12.6V, or even 11V to 13V just to be sure. Alternatively, you design with an internal d/c to d/c that is over specified and can handle 9V to 15V so it's not an issue. The point is that you can't just do a random test on a single sample using randomly selected power supply from another manufacturer and declare everything is OK. The same deal with leakage current tests, fault tests and so on. Each case can be handled by thinking about reasonable assumptions that you can make about the external power supply, and designing and testing to fit those assumptions, rather than relying on a test with a random sample.

IEC 60601-1 does not do this, so really, you just have to do this in the background, and let the test lab do a bit of a fudge. They might, for example, take the simple random sample as being OK, and the regulators too.