Search the Elsmar Cove!
**Search ALL of** with DuckDuckGo including content not in the forum - Search results with No ads.

SIP/SOP - Mains Power and Medical Device Ethernet Ports



I have some questions concerning the leakage tests that require applying mains voltage to SIP/SOP ports. I have read in other posts that if you specify in the manual that SIP/SOP can only be connected to other medical devices is no longer necessary to apply mains voltage to SIP/SOP.

My question concerns the Ethernet port, if it is intended to be connected to a network is it necessary to apply mains voltage?.

Another question I have is about MOPP in SIP/SOP, my understanding is that if the SIP/SOP are to be connected to other medical devices is not necessary to provide 2 MOPP to the SIP/SOP, that would be only necessary when SIP/SOP are designed to be connected to non-medical devices.

In case that the Ethernet port would need 2 MOPP would it be possible to specify that a external network isolator is needed when the Ethernet is to be connected to a network (instead of providing it internally).

My last question is how, (when it is necessary) mains voltage is applied to the SIP/SOP. (do all the SIP/SOP lines need to be shorted and connected to mains voltage?, and if there are more than one SIP/SOP, do all the SIP/SOP need to be connected to mains voltage at the same time?)

Thanks in advance
Actually the expected warning covers connection to medical and non-medical devices. The warning is a bit of a fudge but accepted by most agencies.

It goes something along the lines of "Any person who creates a system by connecting this equipment to other equipment (both medical and non-medical) should confirm the system as being in compliance with Clause 16 of IEC 60601-1:2005 (ME SYSTEMS)." Some agencies might expect the warning has some more detail, but generally it remains a 1 paragraph warning and realistically it is impractical to include all the detail in Clause 16 in a single paragraph.

With respect to the patient, if the medical equipment has Type F insulation (BF, CF) or if the SIP/SOPs are earthed or there is an earthed intermediate circuit, then there is really no potential harm from SIP/SOPs.

The remaining primary target of Clause 16 is to prevent cumulative leakage currents, mainly earth leakage, for the operator. If each item of equipment is plugged into the wall separately, almost all cases will be OK. Only when power boards are used (e.g. 4 or 6 outlets) does the potential for cumulative leakage become real. Theoretically SIP/SOPs can be involved, but practically this is rare.

So in most cases the SIP/SOP is not important, which is why agencies are happy to accept the warning to avoid the test of mains on SIP/SOP.


What you have said is about connecting different devices, but I don't know if that applies when the ethernet port is to be connected to a clinic network (not to another device), what do you think?

Would in that case be necessary to apply mains voltage to the ethernet port? And to provide 2 MOPP between the ethernet port and applied parts?
Ethernet cable still just connects to another device, typically a hub which is mains powered device complying with IEC 60950-1. If that hub's power supply is plugged into the wall directly then typically there are no serious issues associated with ME systems. There are theoretically issues, but practically not which means again the fudge warning is OK.

By fudge I mean the warning goes in the manual, which eliminates the "mains on SIP/SOP" test, but it is unlikely that serious attention is given to the warning in practice, which is reasonable since there is no serious issue in the first place.

We work with ethernet networks everyday, in offices, schools, homes and so on. It's not dangerous. If the theoretical risks raised by IEC 60601-1 were real there would be issues not just with medical devices but everywhere.


Our product has various industry-standard data interfaces Ethernet, USB, RS232 etc. Our IFU just states 'connection of non 60601 compliant mains powered third party equipment to these ports may create a safety hazard'. Our NB seemed happy with this.
For the Ethernet port, which is intended to connect to the customer's site network, obviously this restriction won't work so we had to find an Ethernet transformer with 4 kV / 2 MOPP insulation. There are a couple of companies that supply a module containing the transformer but they are surprisingly hard to find off-the-shelf as a PCB-level component.
The other area that needs attention is creepage and clearance issues around the RJ45 jack itself if the product enclosure is metal, most off-the-shelf connectors are miniaturised to the point where this is difficult to achieve. The ones we used were from Stewart IIRC.


now I am confused I understood from Peter answer that 2 MOP weren't necessary. any more opinions?

RednBlack do you apply 250V on your Ethernet port when doing touch leakage current or do you skip that part?


It may be that our approach is overkill with respect to Ethernet; the advice we received at the time was that as the USB, RS232 etc ports were intended for a specific use (data collection from 3rd party medical equipment) then it was reasonable to assume that these would be 60601 compliant. We also allow the use of generic USB peripherals - keyboards, barcode scanners etc but these are bus powered so there is no significant risk.
With the customer's site Ethernet network however, we have no influence over the quality of the rest of the customer's equipment, how well the cabling is separated from mains wiring etc. I'm sure we've all seen the pictures of uncontrolled rat's nests of wiring in server racks etc! This was seen as an unacceptable risk so 2 MOP was specified for the transformer.
Our NB test lab wrote the test protocol for our product, we don't do any of the testing in-house but yes, my understanding is that they apply 250V to the Ethernet connector during the enclosure leakage current test. I'd be interested to hear any other opinions on this.
A warning saying only connect to 60601 compliant equipment seems to miss the point. Clause 16 still applies even if two medical devices are connected together, as it is theoretically it is possible to exceed leakage limits even with two medical devices, and there are other issues like interchangeability of connectors that are covered in the clause for ME SYSTEMs.

Clause 16 also does not require 2 MOP (2 MOPP or 2 MOOP). The maximum requirement would be 1 MOP, and this is just one of several alternatives which can be used to achieve compliance.

A 2 MOP requirement would imply the circuit is seriously dangerous, which is illogical since the user can simply pull the RJ45 connector by hand and touch the pins directly.

Ethernet circuits are often isolated to 1 MOP anyway to comply with TNV-1 requirements from IEC 60950-1, and just to provide solid functional isolation. I am not an expert on TNV (it is too many years ago), and but it seems some oversight that IEC 60601-1 that it does not refer to IEC 60950-1 for networks just to be sure, as there are other tests such as impulse to simulate surges that can be found on networks.

If the ethernet interface has 1 MOP then it should pass the mains on SIP/SOP test. It makes sense to utilize that isolation for compliance with Clause 16. This in turn means that the warning is not required.


Thanks, that all makes sense. The 2 MOP decision (some years ago) was influenced by this article, among others:

which links to this product:

Re-reading the article now, it looks increasingly like marketing FUD.

I think the argument is that while IT equipment should in theory be 60950 compliant, in practice there is no way for the user or installer of the ME to know whether this is the case. I've certainly seen enough IT power supplies replaced with cheap-and-nasty non-compliant ones when the orignal failed (admittedly not in a hospital environment). But as you say, 1 MOP should be sufficient to protect against this.
Top Bottom