Is it mandatory for Medical Devices to work properly after dielectric strength tests?

[onurk]

Hello,

We are developing a ME device with BF type AP.
The device is powered from 240 VAC.
The device passes all the tests about 60601.
But after dielectric strength test some electronic parts get damaged. So the device does not work.
There is no problem about leakage or isolation.
The test laboratory tests dielectric strength by connecting one lead to both L, N of the AC plug and other lead to Appled part of the device.
Is it mandotary for Medical Devices to work properly after dielectric strangth tests?

Thanks.

 

[Tidge]

It has been some time since I was directly involved with 60601-1 (3rd edition), but from memory: generally the ME Device is required to maintain essential performance (in addition to providing basic safety) when subjected to a single-fault condition. I seem to recall that after some of the more destructive tests that if the device provides an indication of loss of function (with no impact to basic safety) and the risk profile of the device tolerates this circumstance, that this might be acceptable. I would also appreciate a response from someone with more direct, recent experience with a NRTL.

 

[Pads38]

I would suggest that the test house is doing the test wrong.

If you apply a dielectric test voltage all the way from 'front to back' the stress will be created across more than one insulation, with the division of the overall voltage being set by things like the capacitance involved. You will not be able to tell which part is getting the most stress.

For this, it is essential that you develop an 'Insulation Diagram' (See Annex J - and google for background info). This will set out the various insulations that are required to be tested. Remember that dielectric tests are only applied for solid insulation - you analyse others using creepage and clearance rules.

Send the repair bill to the test house!

 

[Peter Selvey]

Pads38 is correct, dielectric strength is intended to test solid insulation only, which is typically implemented in wiring, opto-couplers, safety capacitors (Y1, Y2 types) and transformers. In principle this should be done on the parts themselves. If it is unavoidable to do it on the parts, 8.8.3 (c) has instructions such as shorting out circuits to avoid damage. It's also worth to note that if the parts are certified, you don't actually need to do test.

In practice, the test for 1 MOP (one means of protection, typically 1.5kV) can safely be done in an assembled device, but the test for 2 MOP (typically 4kV) is rarely possible to do on a fully assembled device. This is not only because it will stress parts not intended to be stressed, but because the parts that are intended to provide the insulation are not likely to be stressed. So you end up with broken equipment and an invalid test.

I recommend to claim damages to the test lab, it's a big mistake to do the test this way. This is not to be nasty but just to make sure they don't it again to other customers.

 

[onurk]

Thanks all for your answers.
We have already draw an insulation diagram. Because of its detailed with our markings and brands I can not post it in here.
I will try to explain:
- AC power input socket is CE certified and on the socket it is written " HIPOT tested"
- Cables from socket to Power supply have drawing that explains isolation to earth 1MOOP (enclosure class 1 to the cables) and isolation to BF applied part has 2MOPP.
-The power supply that we use also have 2 MOPP input to output, 1 MOOP input to earth.
- Our BF applied part has 4kv insulation tested between its cables and its lead.
We can test all insulations separately. Once the whole assembled device tested end to end, some parts have stresses that can cause expensive damages.

Our conflict with our LAB is about "how we can test our device after humidity preconditioning?"
Because of its enclosure is resistant to humidity, the insulations are not affected.
Should we disassembly the device after humidity preconditioning and test insulations separately?
And at which environment should we test? (inside humidity room or normal environment)

 

[Pads38]

I load the humidity chamber with a complete device plus samples of things like the transformers. That way I can do the necessary dielectric tests without any disassembly - some on the device, some on the components.
If the chamber is big enough then the tests can be done inside (a 'walk in' chamber). The one I use is smaller so the tests have to be done outside. For best validity keep the test objects inside until the very last second.