General questions about Medical Device MOPs and MOPPs

[DENich]

Hello to all, I'm a newbie in a field of 60601-1 and 60950-1 standards. So, it's sometimes hard to understand (or "believe") tough requirements of 60601-1. Hope you will clarify some items to me. I'll be very thankful.

I have AC\DC power supply, which is intended to use in ME EQUIPMENT. I have a task to modify this power supply in accordance with requirements of 60601-1. You can see a simplified diagram of it in the figure below*. Input voltage is 230Vac, output voltage is variable and depends on the model and can be up to 1500Vdc.

I intend to assume the metal case and wires (input, output and interface) as ACCESSIBLE PARTS. In accordance with "8.1 Fundamental rule of protection against electric shock" I'm obliged to prevent unacceptable TOUCH CURRENTS and voltages on surfaces of ACCESSIBLE PARTS. My understanding of proper MOPs is shown with red on the diagram.

As I've understood, for MOPP I have to use Table 12 for creepages and clearances and Table 6 for solid insulation. So, for example, for 1900Vdc MOPP (400+1500Vdc MOP between MAINS PART and SECONDARY PART on my diagram) in accordance with Table 12 I have to provide 50mm creepage distance and 28,6mm clearance distance. The transformer between MAINS PART and SECONDARY PART should provide the same clearances and creepages and to have (in accordance with Table 6) 1900*sqrt2+5000=7679Vac dielectric strength. It's hard to imagine such a monstrosity, I mean transformer and PCB which are suits for requirements above. Are my rationales right?

I think, I have to add 1500Vdc 2xMop galvanic insulation between SECONDARY PART and control board (it's marked with "1?" on my diagram). In this case I have not to provide any insulation between parts marked "2?". Am I right? Is there any way to use protective impedance in feedback instead of galvanic insulation?

Thanks in advance.

* Unfortunately it's my first post and I have no permission to insert images in a body of the post. I've opened access to image in my Google disk. For seeing diagramm please search DiagramForElsmar27.09.17.pdf in Goole. It will be only one link. Hope it won't break forum's rules.
.

 

[Ronen E]

Sorry, not my expertise. I attached the diagram to the post as a file, though.

 

[DENich]

Sorry, not my expertise. I attached the diagram to the post as a file, though.

Great thank you for your help with the file, Ronen E. Hope someone will be able to help me with topic questions. They are still actual.

 

[MedMartin]

Hi Denich,

Is the output an applied part? connected to a patient? The MOPP requirement of 60601-1 is only valid for applied parts. A power supply normally has no applied part but this depends of course on your application.

Normally 1 MOP is enough from input or output to PE-connected case. PE connection counts as an additional MOP.
You could use a PE connected shield in the transformer which has only a requirement of 1 MOP of max. 1500V working voltage.

From input to output 2 MOP are necessary from input to output. No galvanic isolation is necessary if the applied part is not classified as "floating", BF or CF. Then "only" the leakage currents have to be limited.
And it is always helpful to talk to the notified body about the actual requuirements.
I hope this helps.

Best regards,
Martin

 

[DENich]

Great thank you MedMartin for your answer.

Our power supply is a capacitor charger. This capacitor charger isn't ME EQUIPMENT by itself but it is intended to use in medical devices. Typical usage is this: the capacitor charger is build in system where to its input is connected a mains socket and its outputt is connected to a capacitor bank. The capacitor bank is connected to a Xe lamp thru a powerful switch (IGBT). The capacitor chager charges the capacitor bank and then the switch opens and energy, which is stored in the capacitor bank, is released to the Xe lamp. Light of the lamp can be used for medical purposes.

You could use a PE connected shield in the transformer which has only a requirement of 1 MOP of max. 1500V working voltage.

Thank you for the idea. I'll think about it.

No galvanic isolation is necessary if the applied part is not classified as "floating", BF or CF. Then "only" the leakage currents have to be limited.

Could you specify a number of paragraph in IEC 60601-1 where this situation is described? I can not recall something similar from the standard.
My logic is this: we have 1500Vdc on the secondary part of the power supply and if we have poor insulation between secondary and primary parts, this 1500Vdc might appear in the primary part. The primary part is connected to the MAINS by a power cord (the cord between the system and "a socket on a wall"). This power cord is intended to use with 220Vac and it can not deal with 1500Vdc. Therefore, we might get unacceptable TOUCH CURRENT on the power cord which should be recognized as an ACCESSIBLE PART for an OPERATOR.

And it is always helpful to talk to the notified body about the actual requuirements.

I've written to TUV, hope they will answer.

Thanks one more time

 

[MedMartin]

Hi Denich,

the applied part of the device is classified as applied part of type B, BF or CF. Only BF and CF need galvanic isolation, they are floating. There is no direct definition of this in the standard, you can read chapter 8.5.2.1 and the associated parts of attachment A.

Maybe there is a particular standard 60601-2-X for your application?
Type BF is used for applied parts which are longer in contact with the patient (e.g. electrodes), type CF for applied parts with contact to open wounds, intracardiac, ...

My logic is this: we have 1500Vdc on the secondary part of the power supply and if we have poor insulation between secondary and primary parts, this 1500Vdc might appear in the primary part. The primary part is connected to the MAINS by a power cord (the cord between the system and "a socket on a wall"). This power cord is intended to use with 220Vac and it can not deal with 1500Vdc. Therefore, we might get unacceptable TOUCH CURRENT on the power cord which should be recognized as an ACCESSIBLE PART for an OPERATOR.

The logic sounds reasonable, the operator should be safe

Best regards,
Martin

 

[Peter Selvey]

In the case of high voltage it needs a more detailed analysis than a simple block diagram.

It's normal in safety to make worst case assumptions partly just to keep it simple. For example, in "mains circuit" we assume that all insulation is stressed at 230V, and we assume the stress will be there 24 hours a day for service life of the device. Of course in reality an individual part might only be 50% of mains, and may be stressed for an average of say 30min every day. But it's easier to assume the worst case than do a detailed analysis and have different limits and tests for every part. That approach works well largely because of the wide availability of parts designed around 230V.

That might not work for high voltage systems. In that case it might be more efficient to assess the true working voltage between parts that require safety related insulation. That's likely to find there are only a few points that are actually stressed by the high voltage, and you might find that even for these points the actual peak/rms values are lower than just adding 400Vdc + 1500Vdc = 1900Vdc.

Next it needs to be checked if it really needs to be "patient" (MOPP) or if operator limits are OK (MOOP), and whether 1 MOP or 2 MOP is needed, which might depend on earthing and overcurrent protection. That I can't say for sure without seeing the product but it is possible to design the system in such a way that only 1 MOOP is required (and I would do it that way if you can). There is a huge difference between 1 MOOP and 2 MOPP.

 

[DENich]

Hi MedMartin,
Thank you for the reply. I think, I understand your point. I believe, I should recognize our module has no APPLIED PARTs. It has only ACCESSIBLE PARTs.

Maybe there is a particular standard 60601-2-X for your application?

Thank you for the suggestion. Looks like IEC 60601-2-22 "Medical electrical equipment. Part 2-22. Particular requirements for the basic safety and essential performance of surgical, cosmetic, therapeutic and laser equipment" suits for system in which our module might be installed.

 

[DENich]

Hello Peter Selvey,
Thank you for the reply.

Actually, it was a key point of my question. I wanted to know whether I should understand 60601 strictly or there are some "tricks" which can make my life easier.

That's likely to find there are only a few points that are actually stressed by the high voltage, and you might find that even for these points the actual peak/rms values are lower than just adding 400Vdc + 1500Vdc = 1900Vdc.

As far as I'm concerned, I should consider a fail of any component which is not a MOP as a NORMAL CONDITION. Therefore, I believe, voltages on any point of a scheme which are not separated from it by MOP should be considered as maximal possible voltage within this scheme. Am I right?

I understand that an assumption that superposition of 400Vdc and 1500Vdc equal their sum is the worst case. For sure, the voltages can be coherent and in this case their superposition will equal 1500Vdc.

Next it needs to be checked if it really needs to be "patient" (MOPP) or if operator limits are OK (MOOP), and whether 1 MOP or 2 MOP is needed, which might depend on earthing and overcurrent protection. That I can't say for sure without seeing the product but it is possible to design the system in such a way that only 1 MOOP is required (and I would do it that way if you can). There is a huge difference between 1 MOOP and 2 MOPP

I intend to recognize our module has no APPLIED PARTs (only ACCESSIBLE PARTs) and I think we need to use requirements for MOOP.

Could you give few examples where 1 MOP is needed? Just to get the main idea.

 

[Peter Selvey]

In switching power supplies (which I assume this is), generally faults will cause the supply to cease operation which means the higher voltages stopped being generated.

For that reason, most test labs assessing switching power supplies will use a reference voltage of the mains supply voltage in general (for the whole isolation barrier), and the actual working voltage at a particular point but only if that part exceeds mains voltage (rms or peak). In practice most of the circuit is below mains voltage (e.g. optocouplers, feedback), and there are only a few key traces and parts in the primary circuit that are above mains voltage (400V supply, switching transistor, transformer). It would would be common in assessing a switching power supply to print out the circuit diagram and use a highlighter to show the different areas.

For 1 or 2 MOOP: generally the secondary HV circuit will be current limited (electronic over-current protection). If you surround this part with low impedance to the 0V pole (in this case not earth, but circuit's 0V, but this could also be connected to earth), then a fault should trigger the overcurrent and drop the volts to <60Vdc. It's possible that no MOP is required for such areas, or 1 MOP if the 0V grounding is reliable but could fail over the life, or 2 MOP if the grounding is considered unreliable. This "grounding" does not need to meet the 25A test, only low enough impedance to reliably carry the current from the HV supply in a way to trigger the over-current protection.

However if for example there is a plastic case and you route the HV wiring near that without any 0V parts in the way, that would definitely need 2 MOP.

And if the HV output is floating, it's a different story again because we may not be able to rely on over-current protection as much.

The point is to know where the source of harm is and have a plan. Don't worry about 601 at the start. You have 1500Vdc, that is dangerous. What's your plan to stop people getting a shock? Play it by ear and hope it passes 601? Usually not, the designer will usually have a reasonable strategy because they know 1500Vdc is dangerous. Figure out what that plan or strategy is and then see how it fits in with 601. Sometimes that requires a little lateral thinking, but it can often fit.