Separation of F-type applied part and remaining parts

[dtec2]

Hi all!

I'm currently having my first read at IEC 60601-1 and I stumbled upon clause 8.5.2.1 which mentions that patient connections of any F-type applied part needs to be separated from all other parts. However, given my current device's architecture (which I will detail below), this doesn't make sense to me. I'd like to hear your thoughts on this!

Basically, my device is a wearable device which is body-worn for several days and has 2 snaps/electrodes (the applied part) which are used to acquire ECG (therefore a BF applied part). The device is powered by a LiPo battery. The device can only be charged via USB (either through a medical grade charger or by being connected to the computer). Given that the medical grade charger has all the protection required by the IEC 60601, I'm only considering hazard situations to arise from the PC connection (e.g., a peak current from the mains, which flows through the computer, goes into my device via the USB port and reaches the patient through the snaps). To protect against this, I have implemented an USB isolator (the ADuM family isolators), which has the recommended creepage distances and air clearances for 2 MOPP.

Hence, given that I'm protecting the only possible source of hazard, it doesn't make sense to me that I need to separate the ECG circuitry from the remaining circuitry. What are your thoughts on this? To be in compliance with this clause, can I consider the entire wearable device as a Type BF applied part?

Thanks so much!
Diogo

 

[Benjamin Weber]

The way you describe the device it seems to be quite small. So considering the whole device as one applied part sounds reasonable.
But be careful: The requirement is about the patient connections, not the applied part. The patient connections are only one part of the whole applied part. In your case I would say the electrodes (the conductive areas only) are the patient connections. Those need separation to all other parts, i.e. metal accessible parts, SIP/SOPs, earth connections (protective earth, functional earth, potential equalization). But also to all secondary circuits in order to not exceed patient leakage and auxiliary currents in normal condition, single fault conditions and special test conditions.
Regarding the ADuM separators: As far as I recall there are different package sizes with different creepages/clearances! I recommend to check if the required values are really fulfilled! Do not just rely on "2 MOPP" stated in the data sheet!

 

[dtec2]

Thanks so much for pointing that out Benjamin - I was missing that point.

So, in my case I definitely need to isolate my ECG circuit from the remaining circuit using a MOPP? The reason why it makes me confusion is that by isolating the USB port, the whole device is "secure", as I don't see a way that an excessive current will get to the patient through the ECG. But I accept this extra security layer.

Can you please give me some advice on what MOPP I can use to isolate my patient connections? I was thinking of using something between the AD8232 (instrumentation amplifier) and the MCU (signaled with an A on the image in attachment), but I'm not sure what I can use. Do you have any thoughts?

Regarding the ADuM separator, I'm indeed choosing one which contains more than the minimum creepages/clearances for 2 MOPP! PS. I haven't yet included the ADuM in my diagram.

Many thanks for your help - it is extremely valuable and helpful!

 

[Peter Selvey]

Just to add to BW's comments: the key point is to realise that you are adding electrodes to the patient. It makes sense then to make sure that the electrodes are isolated from all accessible parts. It's not the secondary that you need isolation from, but the accessible parts.

Consider for example a really bad design that is made with a metal box. The 0V of the circuit is connected to the metal box. The 0V of the circuit is also connected to the patient electrode. That means anybody touching the box is effectively touching the patient.

The best solution is case by case.

If your main electronics is small, then it makes sense to treat it all as patient circuit and then provide 1 MOPP isolation around it (between the circuit and any user controls, chargers, SIP/SOPs etc.).

If the main electronics is complicated and many interfaces to the operator, other devices etc, it can be more efficient to have a small isolated circuit just for ECG (1 MOPP barrier between ECG and main secondary), and then you don't have to worry about 1 MOPP around the main circuit.

 

[Peter Selvey]

Just to add to my comment: electrodes bypass the skin impedance. It's the reason why more care is needed.

 

[dtec2]

Thanks for your help Peter - that was helpful!

I completely agree that isolating a patient connection from user controls, chargers, SIP/SOPs, etc makes complete sense. What I'm struggling a little bit is to understand if that really needed in my case.

I'm attaching a complete diagram of my device. It only contais two accessible parts or SIP/SOPs: the USB connection and an SD card. I only see hazardous situations possibly arising from these 2 places. Do you agree with me?



Regarding the USB, it may be connected to the mains via regular charging (even though we use a "medical charger") and to the computer (which may be hazardous with a current spike) - either way, both are protected with 2 MOPP. Regarding the SD card connector, I'm not seeing a possible hazardous situation from it. I only see danger in case someone purposely connects some danger current there, which doesn't seem a "reasonably foreseeable misuse".

Taking this into account, I believe that the 2 MOPP from the USB isolator provide all the isolation needed for my patient connections (and do comply with clause 8.5.2.1). Am I wrong with my rationale? Or my rationale doesn't matter at all and I just have to implement what they mention?

I prefer making my device overly secure than to put someone's life at risk, but I'd like to understand the rationale behind each implementation.

Thanks so much!!

 

[Peter Selvey]

For 8.5.2.1, only 1 MOPP is required (F-type insulation is 1 MOPP). The point of the F-Type insulation is to stop the medical device from "grounding" the patient. In many cases it can be overkill, but for ECG it also turns out to be useful for noise, CMRR as well.

F-Type insulation is between ECG circuit and all accessible parts. The accessible parts are more than just the USB/SD card, it is all plastic parts as well, such as displays, on/off switches, metal screw and the like. These other parts are likely to be OK, but it's a pain to prove it, and every design change it needs to be re-assessed.

That's why it makes sense to keep the "ECG circuit" as small as possible. In your diagram there is an ECG isolator (1 MOPP) which if it is physically located using common sense then you don't need to worry about the SD card, screens, switches, case screws etc.

It remains an option to skip the ECG isolator, it just means you need to show 1 MOPP everywhere else.

For ECG, CMRR is really about leakage currents (in the nA range), which when flowing through an imbalance in the electrode impedance of say 10k can create mains noise around 1mV. Having a smaller ECG circuit reduces the leakage and hence improves CMRR, which will be important for 2 lead ECG since there is no right leg drive (noise cancellation).

 

[dtec2]

Thanks for sharing your expertise in this - it is very helpful! I completely agree with you and it is the smartest attitude to add 1 MOPP at the ECG.

Regarding the separation of the ECG circuitry, do you have any suggestion on a good MOPP? We are currently connecting the analog output signal to our MCU's internal ADC. I have only been able to find digital isolators, which would make me need to add an external ADC. Do you have any suggestion on this? It would be very helpful.

 

[mehrnaz]

Just to add to BW's comments: the key point is to realise that you are adding electrodes to the patient. It makes sense then to make sure that the electrodes are isolated from all accessible parts. It's not the secondary that you need isolation from, but the accessible parts.

Consider for example a really bad design that is made with a metal box. The 0V of the circuit is connected to the metal box. The 0V of the circuit is also connected to the patient electrode. That means anybody touching the box is effectively touching the patient.

The best solution is case by case.

If your main electronics is small, then it makes sense to treat it all as patient circuit and then provide 1 MOPP isolation around it (between the circuit and any user controls, chargers, SIP/SOPs etc.).

If the main electronics is complicated and many interfaces to the operator, other devices etc, it can be more efficient to have a small isolated circuit just for ECG (1 MOPP barrier between ECG and main secondary), and then you don't have to worry about 1 MOPP around the main circuit.

Dear Peter Selvey , regarding the helpful notes you mentioned about isolation of type BF applied part from enclosure. I have an orthopedic oscillating saw which is type BF and as this is capable of being sterilized, its enclosure is totally metal.
as you know its blade is also metal and the blade is in physical contact with its holder in enclosure, somehow it is not functional to isolate the blade (which is applied part to patient) from enclosure. so would you please help me how to make this device confirm with type BF applied part requirements?

 

[Peter Selvey]

If the part held by the operator is metal and also in contact with the blade then the applied part should be classified as Type B. To make it type BF , it sounds like the enclosure (held be the operator) would need to be plastic or there would need to be some plastic separator between the motor and the saw. It could get complicated if there are fluids involved.

Is there a reason to use BF? Often this comes from marketing rather than standards/regulations e.g. the competitor is BF so it is desirable to have the same for your device.