USB Port (Signal I/O) Isolation as per IEC60601-1 - Medical Device

hi everybody,
I am new to this forum.
We are developing a medical instrument which has USB port for software updations and data logging.
The USB controller is inbuilt to the processor and only D+ and D- lines are available to the outside.
Since USB is a signal I/O part, as per standard, we need to provide isolation for this. I couldn't find any bidirectional isolators that can be used for USB port.
If anybody has experience with this type of problem, please give me some suggestion.
First of all, please clarify if the USB isolation is required or not because there will be no patient connection while the USB port will be operated.

Any suggestions or references will be greatly appreciated.
thanks in advance

Just bumping this thread. Is anybody able to help with regards to this subject?

...still no takers on this one...(?) I´ll give it a try though I´m sure there are people here that can tell for sure...

As far as I understand, there is no explicit requirement that signal I/O shall be isolated. What you need to do however, is to make sure that any Applied Part as well as Accessible Part will not be connected to the mains voltage under normal or single fault condition.

So, if your connect your USB device to e.g a motherboard, and the applied part is also connected to the same motherboard, the isolation might just as well be between the applied part and the motherboard.

t.

Thanks Temujin,
I understood your point.But, my doubt is that, don't operator needs protection from the internal secondary voltage. From what i understood, the USB port should also have 2MOOP(means of operator protection). So, the port should have isolation from internal 24V DC.This requires creepage and clearence distances.Is there any way to isolate bidirectional D+ and D- lines of USB.
Thanks again

royal77 said:

From what i understood, the USB port should also have 2MOOP(means of operator protection).

Click to expand...

Where did you find the requirement that a USB port shall have 2MOOP?
Is ther any way the patient can access the port of can the operator use it when the patient is connected to your device?
By the way what kind of instrument are we talking about?

According to 60601-1 (3rd edition) if the means of protection is to be classified as MOOP (Means of Operator Protection) you have 3 options:
1: Treat it as a Means of Patient Protection
2 / 3: Apply IEC 60950-1 - Information technology equipment - Safety - Part 1: General requirements.

THough you will have to assess in your risk assesment that the the USP Port will not be subject to the same requirements as for Applied Parts.



t.

As per clause 8.1 a of 3rd edition, application of voltage on SIO parts is a normal condition. Doesn't this mean that, we should have 2 means of protection?
The patient cannot touch the port and the operator will not use this port while connected to the patient. But, as per hazard analysis, there is a chance of somebody using this port while being connected to the patient.
FYI, this is a biopsy surgical instrument.

Are there any components which can isolate bidirectional high speed signals?

Regards

Hi again...

To your original question, I must say I have no idea...

royal77 said:

Are there any components which can isolate bidirectional high speed signals?
Regards

Click to expand...

Note that Clause 8.1 of the 3rd edition says Normal condition is e.g. "(...) 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.(...)

So if you specify in you user documentation that connection to usb while patient is connected is not allowed, this situation is not a normal condition.

But as I said, I am not an expert on this, and I might be wrong.

t.

Hello

I also do not know if there are components which can isolate bidirectional high speed signal.

Regarding the interpretation of the requirements, SIP/SOP isolation is always a pain to understand, because the standards are a little confused about what theny want and about exemptions.

I´ll try so summarize what i think should be done to make things easier (the last time we discussed this topic in your lab we "lost" three days).

First, you need to specify to specift the equipment that can be connected to the port.

7.9.2.5 ...

The instructions for use shall specify any restrictions on other equipment or NETWORK/DATA
COUPLINGS, other than those forming part of an ME SYSTEM, to which a SIGNAL INPUT/OUTPUT
PART may be connected.

Click to expand...

If you make a restriction, meaning you specify an equipment that complies with IEC 60601 or relevant ISO or IEC standards (this information is not clear in the standard but can be verified in : IEC/TR 62296 - Considerations of unaddressed safety aspects in the Second Edition of IEC 60601-1 and proposals for new requirements) then the normal condition would be the presence of the voltage or current from the connected equipment, i.e., the maximum current or voltage in normal use. If there´s no restriction, then it´s assumed that the normal condition is the full mains voltage, simply because you cannot guarantee which voltage can be present on the other equipment if a fault develops (also take a look at the explanation to Subclause 8.7.4.7 c) in the rationale).

Now, to the tests and requirements.

As temujin said, there´s no explicit requirement for the isolation of SIP/SOP. Problem is, if they have electrial signals or connections, isolation would be needed for compliance with leakage current measurements (see explanation to Subclause 3.115 in the rationale).

What do you need to test? Well, basically , touch current (8.7.4.6 c) and figure 14) and patient leakage current (8.7.4.7 c) and figure 17). But only when required....and when is that? Well, it seems that the third edition only requires testing when you do not make restrictions on the equipment being connected (and this is a little different from the second edition when there could be other exemptions), thus the additional transformer on the tests.

So isolation would just need to be there if you do not declare the equipment to which you can connect yours.

This might seem strange given that that the requirement for the other equipment to comply to 60601 or other ISO or IEC standard is not present in the third edition. That´s why i think there´s still something missing in this explanation. But right now this is my interpretation.

Cheers!

Thanks for the information. So, you mean to say that, if we say that the devices that will be connected to the USB port should not carry voltages above 5V DC, then we don't need isolation for this port. Right? I have one more question. Are any USB cables (host to host bridge cables) which have inbuilt isolation. Atleast, we can specify that these can be used.

Thanks

No, i didn´t say exactly that. I meant to say that, if you say that the devices that will be connected to the usb port complies with IEC 60601 or the relevante IEC safety standard for the equipment, AND the voltages are below the limit levels on 8.4.2, then i think isolation would not be needed - but simply because the leakage current tests would not be aplicable. This approach is similar to the second edition, but do not take into account the other exemptions.

The compliance to the standard is the crucial point, because it guarantees that the other equipment cannot have a hazardous voltage in a fault condition. If it was just "USB port should not carry voltages above 5V DC" as you said, there´s no guarantee taht in a fault the other equipment could develop, say, maisn vooltage, so in this case a isolation would be needed.