60601-1 Clause 8.1 a) SIP/SOP voltages

rohan.brash

Registered
Hello all, first time poster here!

This clause has been addressed a lot but I am still confused about a couple of things regarding our application.

We have a protectively earthed medical device that can (but is not intended to) come into contact with the patient. The device has an external AC/DC 24V supply such that the internal working voltage of the device is 24V.

This device has two SIP/SOP ports:
  1. Standard RJ45 ethernet for connection to hospital network or another system that we make. The other system has mains input (internal AC/DC supply).
  2. Another network connection (proprietary connector on our side) intended to only be connected to a specific device that has an external AC/DC 24V supply. This other device unfortunately only has a standard RJ45 input so technically the cable could also be connected to other unintended devices.
The clause 8.1 a) states that Normal Condition includes "the presence on any SIGNAL INPUT / OUTPUT PART of any voltage or current from other electrical equipment that is permitted to be connected according to the ACCOMPANYING DOCUMENTS as specified in 7.9, or, if the ACCOMPANYING DOCUMENTS place no restrictions on such other electrical equipment, the presence of the MAXIMUM MAINS VOLTAGE as specified in 8.5.3;

Our accompanying documents state that additional equipment must comply with relevant IEC/ISO standards and all configurations should comply with the requirements for medical electrical systems.

My questions are regarding the part of the clause "...of any voltage or current from other electrical equipment that is permitted to be connected..." :
  1. Does this imply that application of any working voltage of any permitted device is Normal Condition, or just the signal/dc voltages we expect on the port? For example, for the RJ45 port the connected device could have a mains input so application of mains voltage to the RJ45 is Normal Condition?
  2. Is it reasonable to place an explicit restriction in the documentation for the other port, such that it only be allowed to be connected to the one specific intended device, despite the other end of the cable having a standard connector? If we could, would this lower the MOPP requirements of this port to 24V, as that is the working voltage of the other device?
I hope this is clear enough, please don't hesitate to ask for clarification if needed.

Thanks!
 

Peter Selvey

Leader
Super Moderator
It's a little bit complicated as the open circuit voltage for a external SIP/SOP can in normal and fault condition be relatively high, and the leakage also moderate. This is due to the EMC capacitors that bridge primary and secondary in double insulated devices or connect to the "earth" of an IT device, which isn't always reliably connected through to mains earth. Typically, the EMC capacitors are balanced, so if measured with a high impedance meter you should see around half mains voltage when the external signal circuit is otherwise floating (before being connected to your medical device). Since the EMC capacitors are high impedance at mains frequency (1-10MΩ), as soon as it's connected to anything "vaguely" earthed (relative to ~1MΩ) the voltage drops and the potential of the leakage currents to flow.

Note that all of this is unrelated to the 24V or 5V in the signal circuit itself.

If your system uses an isolation barrier to prevent the leakages from external devices from reaching the patient or operator, then it's important to factor in the open circuit voltage and test the barrier at least for 1MOOP @ mains voltage (1MOOP? Long story but it makes sense).

If there is no SIP/SOP barrier and the medical devices is earthed, then as soon as you make the connection, the voltage drops to SIP/SOP levels (e.g. 5V, 24V) and the leakage current is diverted to earth and ultimately back to mains earth via the medical device's mains cable. No problem!

When the medical device's earth is broken (SFC), these leakage currents could be diverted to the operator or patient. Since these are unknown (we don't know the full installation outside of the medical device, networks etc.) it could be assumed to be >0.5mA, but in reality it's more likely to be somewhere in the 0.05 to 0.5mA range, and rarely exceeding 1mA. Nevertheless, if the external equipment is not fixed, we should assume >0.5mA as a starting point.

Now for the fudge.

Test labs allow manufacturers to include a warning in the operation manual that says connection to other networks and SIP/SOP should comply with Clause 16 of IEC 60601-1, and once the warning is there the test of "MAINS on SIP/SOP" is not performed. However, this is logically wrong. Essentially what this does is place a responsibility on the user to check that the installation does not exceed leakage current limits, which should fail IEC 60601-1-6 (usability) since most users would not do this or know how to do this.

Rather, the correct approach would assess the probability of (a) leakages from external devices exceeding 0.5mA combining with (b) broken earth in the medical device, together with (c) a realistic assessment of the potential harm of such leakages up to 1mA. For me, (a), (b) and (c) are all low except in special cases such as medical devices in cardiac locations, critical care, where it's important to avoid high leakages floating around. Anyway, for a typical medical device, the risk should be acceptable and the warning itself is just for liability :).

If the risk is deemed unacceptable, then the designer should consider isolated I/O ports, which may not be that expensive. Other options include a EP terminal (potential equalization) and instructions for the user to connect the EP bar if available, which adds a second earth in case the main one fails. This could make sense in areas where EP earthing is standard practice and users can be expected to follow the instructions.
 

rohan.brash

Registered
Thanks for your thorough response, Peter!

So, after some more deliberation based on your feedback, we have the approach below.
  1. IFU contains the magic words that any SIP/SOP configuration should comply with the requirements of clause 16 (for liability)
  2. All accessible conductive parts are protectively earthed.
  3. There will be 1MOOP/MOPP (@ max mains voltage) between the SIP/SOPs and secondary circuits, but only functional separation between these SIP/SOPs and PE due to the connectors we are using.
Note: This approach assumes that the patient is likely (but not intended) to touch the PE chassis of the device, and has negligible probability of touching secondary circuits directly. The contact to the PE chassis (if any) will likely be over a large area but through intact skin. Additionally, there are no patient connections. We do not have an additional potential equalization terminal at this time.

Scenario 1: SIP/SOP device is in Normal Condition with leakage current less than 0.5mA. This current breaches the (functional) separation to the chassis (still NC), and we have a break in our device PE connection (SFC). -> Safe as touch current still below 0.5mA.

Scenario 2: SIP/SOP device is in Single Fault Condition (e.g broken earth connection) with leakage greater than 0.5mA. This current also breaches the (functional) separation to the chassis but is redirected through our intact PE to earth. -> Safe as touch current remains low

Scenario 3: SIP/SOP device is in Single Fault Condition (e.g broken earth connection) with leakage greater than 0.5mA. This current breaches the 1MOPP to secondary circuits (e.g GND) which is also a Single Fault Condition. Operator then touches the secondary circuit (somewhat likely) or operator touches secondary circuit and patient at the same time (very unlikely but possible) -> Safe as requires at least two single fault conditions.

My questions are:
  1. Does this make sense or am I missing something?
  2. Given that the device is intended for use in an OR where other devices can be assumed to comply with relevant standards, would a test house accept the assumption that leakage is less than 0.5mA in NC from other devices? Would they even care as long as the IFU contains the clause 16 stipulation?
Thanks!
 

Peter Selvey

Leader
Super Moderator
The analysis looks good, especially for OR situation where it is more likely that the earthing of devices and leakage in the room is periodically checked and possibly using equipotential earthing (back up earthing), isolation transformers and not using MPSOs (multiple portable socket outlets, which can accumulate leakage). These are all practices that reduce the risk of wayward leakage currents in a critical environment. For Q2), in practice you could expect <0.1mA for an OR, and using an assumption of 0.5mA add an extra safety factor.

One extra point is that while low currents in the 0.2 ~ 0.5mA range are unable to practically cause serious direct harm, as in stopping the heart or breathing, they can be felt and it can be painful. The effect arises from the high open circuit voltage which if a person "brushes" the energized part means that an arc can form just prior to firm contact. This arc concentrates the current allowing a person to "feel" leakage currents as low as 0.05mA, well below the typical threshold of 2mA that is often found in literature. Once there is firm contact the current spreads and the arc effect disappears.

The reason to raise this point is that if a person brushes a metal frame that is floating at ~120Vrms (half mains) and has a potential leakage of 0.5mA (or more), the pain from the arc effect could induce an involuntary action, which in turn could produce a medical error (indirect harm?), especially in a OR. In terms of IEC 60601-1, the limit of 0.5mA in SFC is acceptable, but it's just a note that in an OR, the issue of cumulative or SIP/SOP leakage in systems should be considered a little more seriously. That said, your solution of 1MOPP to the secondary and functional insulation to PE seems fine, noting than an OR should also be controlled better than other areas in health care.
 
Top Bottom