Arguing with the test lab about failing in an EMC test

[MedicalResp]

Hi,

I have a wireless medical device that measures a specific vital sign and transmits the data over to a dedicated software installed on a laptop.
The device is not a vital-signs monitor or life supporting device.
The device's essential performance is defined as the accuracy of the measured parameter.

During ESD gun testing at 15eV (a part of EMC 60601-1-2) the device may in some cases either:
(i) temporarily loses communication with the laptop, but automatically reestablishes the communication after few seconds and continues the measurement.
(ii) loses communication with the laptop, but requires to manually restart the measurement using the software
(iii) shuts down and requires to manually turn it on. Once turned on the measurement continues.

In all above-mentioned scenarios the software alerts the user that the data transmission is stopped and instruct the user if a specific action is needed (for example to turn on the device)

Although the accuracy of the measured vital sign (i.e. essential performance) is not affected during testing, the test lab claims that the device cannot pass the test, since there are cases in which the measurement does not restart automatically after the ESD interference.

In my opinion, their claim in not correct and is not supported by the standard. While it might be arguably considered as a functional fault, this is not an essential performance or safety fault and therefore cannot be a reason for a failing the test.
Before starting an argument with the test lab, is there anything that I am might missing here?

Your inputs will be much appreciated.

 

[VinceTech]

ESD at 15KV. I guess its -1-2:2014

Did you provide the pass/fail criteria in the test plan?

 

[sagai]

Hi,
So, I do not think that software alert could matter too much for the EMC test result.
It is either working or not, difficult to build confidence on.

I know little to nothing of your device, however if the transmission cuts of with no automated mechanism to resurrect it, there is a job for risk analisys to reason the acceptability of.

I do not think that arguing with a certified test lab would go anywhere, they meet number of devices on a more frequent basis than manufacturers could.

Cheers

 

[MedicalResp]

Thanks guys for the quick replies.

ESD at 15KV. I guess its -1-2:2014

Did you provide the pass/fail criteria in the test plan?

Yes. 60601-1-2:2014 (4th ed.) - 15KeV in air discharge.

Just to clarify, since our device works on Bluetooth, we have to comply with the radio freq. standard (EN 301 489) as well. The max. limit for air discharge in EN 301 is 8KeV and we are able to show a continuous communication link during the test.
Our only issue is at 15KeV which is part of the IEC 60606-1-2 4th ed. and as per my understating continuous communication link is not a defined criteria in this standard, but rather the essential performance.

The pass/fail criteria as we have provided is the essential performance: we are using a simulator to simulate the measured vital sign during the tests. The criteria is that during measurement the measured parameter cannot exceed the predefined accuracy limits (e.g. the simulator provides a value of X and our device should measure X +/-10%).

As part of risk management we use the software as a risk measure, to alert the user when the device is not measuring. As I understand, if during operation mode the device provides the essential performance or alerts when it is not measuring and also instruct the user and there are no safety issues than there should not be any issues for the test lab to approve.

Hi,
So, I do not think that software alert could matter too much for the EMC test result.
It is either working or not, difficult to build confidence on.

I know little to nothing of your device, however if the transmission cuts of with no automated mechanism to resurrect it, there is a job for risk analisys to reason the acceptability of.

I do not think that arguing with a certified test lab would go anywhere, they meet number of devices on a more frequent basis than manufacturers could.

Cheers

Please see my answer above.
Although this is my first time dealing with a test lab, my logic is that their analysis and decision must be based on the standard. Maybe I am mistaken (and I have no problem to learn something new), but I have not seen any reference in the standard that supports their claim. As the manufacturer , we define what is an acceptable risk and what is not.

 

[Pads38]

Couple of questions -

Is your device solely for professional areas? or does it include home healthcare settings as well?

Are you using any 'Particular' standard (60601-2-xx)?

Reason for the questions is that each can modify pass/fail criteria.

 

[MedicalResp]

Couple of questions -

Is your device solely for professional areas? or does it include home healthcare settings as well?

Are you using any 'Particular' standard (60601-2-xx)?

Reason for the questions is that each can modify pass/fail criteria.

Yes. the device is solely for professional areas. No particular standard is applicable.

 

[Pads38]

A new Particular standard that I am familiar with is 601-2-40; Evoked Response Equipment. The new version was published in 2016 - after Ed 4 of the EMC standard. As such it includes requirements directly related to the elevated ESD levels brought in by Ed 4.

Clause 202.8.1.101 introduces Immunity Pass/Fail Criteria, with 3 levels being shown. (Very) roughly these are:

A) Basic Safety and Essential Performance maintained
B) As A, but temporary disturbance allowed, no manual intervention allowed
C) As A, but temporary loss of function allowed, manual intervention allowed.

Further tables then detail the tests and which criteria to apply. For the 15kV ESD this allows criteria A, B or C for professional use devices; for home it is only criteria A or B.

I understand that this approach, commonly used for 'normal' non-medical devices, is likely to be used whenever a 'Particular' standard is updated.

As this represents the 'state of the art' I believe you would be able to use this to convince your EMC test house that you should be OK.

 

[3dBUnderThe Limit]

I have been working with EMC for medical devices for many years and I think that the EMC lab has made the correct call. Unless the pass-fail criteria for a particular EP includes transition to a safe state where no data received is acceptable this should constitute a fail (60601-1 4.3).


Unless the root cause can be fixed, I would provide a rational why the loss of data is not critical and include a new pass-fail criterion in the test plan. See 60601-1-2 ed 4; 8.1 and annex I.


With the risk of being slightly off topic:
In dry condition a person can be charged with up to voltage levels well above the 15kV used in the standard, walking across a floor when the relative humidity is low can charge a person to 35kV (Running on a treadmill could be one scenario that can create high voltage levels). So, if the device fails in the ESD test it’s likely that you will see problems in the field.

 

[Jaydub]

If there is clear pass/fail criteria in your test plan, that should be used. If the criteria is not spelled out in the test plan, a criteria A,B,C could be stated, with a description of what happened, without a "pass" or "fail".

 

[VinceTech]

1. Criteria can be determined through manufactures own Risk Analysis .
This is according to IEC60601-1-2:2014 Clause 8.1 '...Before IMMUNITY testing begins, the MANUFACTURER shall determine specific, detailed IMMUNITY pass/fail criteria, based on applicable part two standards or RISK MANAGEMENT,...'

2. You can still decide if the test result is acceptable or not based on the Risk analysis afterwards. This is according to Clause 8.1 'Following the tests, any effects on the ME EQUIPMENT or ME SYSTEM that are observed during or after the application of the test DISTURBANCES should be considered in the on-going RISK MANAGEMENT PROCESS.'