Protective Earth Test in IEC 60601

[sankui]

Dear all,

There is a technical issue pertaining to the protective earth test in IEC 60601-1. Hope I can get the answer here.

As you can see from IEC 60601-1, clause 18f) requires a current source with a no-load voltage not exceeding 6V when tested the impedance of protective earth. The CTL DSH-432 which is talking about the measurement of protective earth, states that a circuit to the PROTECTIVE EARTH TERMINAL may have zones of higher impedance, for example due to oxidation of materials. Voltages higher than 6 V prevent detection of such zones because of their ability to flash through.

I have no idea why. I tried to find some literature to support this view, but in vain. In fact, many other standards, such as IEC 60950 for IT equipment and IEC 61010 for laboratory equipment, require no-load voltage of 12 V maximum. I am convinced that for non-medical equipment, the same problem also exists. The same problem means voltages higher than 6 V prevent detection of oxidation of materials.

Can anyone tell me how to understand the origin of 6V in medical standard and 12V in non-medical standard.

Thanks and best regards

[Stijloor]

In Reply to Parent Post by sankui

Dear all,

There is a technical issue pertaining to the protective earth test in IEC 60601-1. Hope I can get the answer here.

As you can see from IEC 60601-1, clause 18f) requires a current source with a no-load voltage not exceeding 6V when tested the impedance of protective earth. The CTL DSH-432 which is talking about the measurement of protective earth, states that a circuit to the PROTECTIVE EARTH TERMINAL may have zones of higher impedance, for example due to oxidation of materials. Voltages higher than 6 V prevent detection of such zones because of their ability to flash through.

I have no idea why. I tried to find some literature to support this view, but in vain. In fact, many other standards, such as IEC 60950 for IT equipment and IEC 61010 for laboratory equipment, require no-load voltage of 12 V maximum. I am convinced that for non-medical equipment, the same problem also exists. The same problem means voltages higher than 6 V prevent detection of oxidation of materials.

Can anyone tell me how to understand the origin of 6V in medical standard and 12V in non-medical standard.

Thanks and best regards

Does anyone have an idea why this is?

Thank you!

Stijloor.

[20110131]

Hello,
Your question seems to be why is there a 6V medical, and 12V non-medical maximum.
In your comments the flash-over answer was also given.
I can give you technical comment which might answer your question.
The ground-bond test which is also IEC/ISO standardised,
has been tested for many years with a common series of instruments called ROD-L ground-bond testers.
The test device applies a nominal 30Amps AC to the chassis,
and verifies the resistance to earth terminal as being below
user specified limits of nominally 0.01-to-0.025 Ohms.
This resistance generates o.3 to 0.6V at the nominal 30Amps test value.
All this technical talk leads to the comment that a supply with
DC or AC 1 to 2 Volts is enough to test the Earth ground, keeping in mind other series resistances.
This then can lead to the no-load of 6 or 12 volts maximum.
A no load of more is not benifical, so is being restricted.
This comment is a general Earth Ground safety resistance check comment,
which is intended to possibly answer or guide an answer for you.
Happy new year.

[Peter Selvey]

The impedance limit for medical (IEC 60601-1) is 0.1 or 0.2ohm depending if a mains cable is fitted, meaning the voltage drop can be 5V at 25A, or higher for higher current equipment. While typical test values are around 1V, over 20+ years I have seen the odd case where measurement voltages of 2-3V are necessary (e.g. large equipment with cabling to other parts).

Although it is theoretically possible to need more than 6V for higher current equipment (test current above 25A), in practice such large equipment uses large gauge wiring so the resistance is well below the limit and the test voltage remains well below 6V.

I don't know of the original reason for the 6V, but from a theoretical point of view keeping the voltage as low as possible makes sense. In the event of a loose joint (e.g. terminal screw not tightened) or poor choice of materials, it is possible to have a small amount of oxidation on the metal parts which conducts only at higher voltages. At the same time it is possible to get faults to earth from parts in the mains circuit with a much lower voltage than the rated mains voltage (e.g. 10V). The combination of an oxidized joint with a lower voltage fault would see that voltage appear on the frame of the device. In medical this is more serious situation than for non-medical equipment, as there are often fluids used around the equipment meaning even low voltages can be significant.

That said, it is all hypothetical and if you studied the full sequence of events leading to harm, the overall probability would be extremely small.

If there are complications meeting the 6V limit, I would suggest splitting the test in two: first, a simple continuity tester with say 5Vdc source to prove that there are no oxidation or other effects, followed by the test at 25A or as required.

[sankui]

In Reply to Parent Post by Peter Selvey

The impedance limit for medical (IEC 60601-1) is 0.1 or 0.2ohm depending if a mains cable is fitted, meaning the voltage drop can be 5V at 25A, or higher for higher current equipment. While typical test values are around 1V, over 20+ years I have seen the odd case where measurement voltages of 2-3V are necessary (e.g. large equipment with cabling to other parts).

Although it is theoretically possible to need more than 6V for higher current equipment (test current above 25A), in practice such large equipment uses large gauge wiring so the resistance is well below the limit and the test voltage remains well below 6V.

I don't know of the original reason for the 6V, but from a theoretical point of view keeping the voltage as low as possible makes sense. In the event of a loose joint (e.g. terminal screw not tightened) or poor choice of materials, it is possible to have a small amount of oxidation on the metal parts which conducts only at higher voltages. At the same time it is possible to get faults to earth from parts in the mains circuit with a much lower voltage than the rated mains voltage (e.g. 10V). The combination of an oxidized joint with a lower voltage fault would see that voltage appear on the frame of the device. In medical this is more serious situation than for non-medical equipment, as there are often fluids used around the equipment meaning even low voltages can be significant.

That said, it is all hypothetical and if you studied the full sequence of events leading to harm, the overall probability would be extremely small.

If there are complications meeting the 6V limit, I would suggest splitting the test in two: first, a simple continuity tester with say 5Vdc source to prove that there are no oxidation or other effects, followed by the test at 25A or as required.

Hi Peter,

Thanks for your explaination. It makes more sense. Actually, it is also an advice to split the protective bonding test by two steps in CTL decision or IEC 60601-1 3rd edition. But the real problem is that my company is also manufacturing medical equipment (per 60601) and IT equipment (per 60950). We only have one ground bonding tester with no-load voltage of 12V maximum which meets the IEC 60950.

You may have noticed that IEC 60950 also considers the corrosion due to electrochemical action. But this idea has never been reflected in the IEC 60601, including the 3rd edition. I do not know whether this is one reason for more stringent in IEC 60601 regarding the no-load voltage.

I agree that oxidized joint with/without a lower voltage fault can also draw a higher voltage when tested. But according to Ohm's Law (R=V/I), when current is constant, the higher the voltage, the greater the resistance. The important thing is that all the equipment can encounter such a problem (oxidized joint ). This is not the "PATENT" of medical devices.

Assuming that the resistance is 0.3 ohm and the test current is 25 A, then 7.5 V of voltage is needed. From the electrical principles, if the maximum no-load voltage is 6V, above mentioned assumption can not draw a constant current of 25 A. In modern ground bonding tester, alarm will be generated. Even without the alarm, the test current may be automatically adjusted to 20A, the test result of 0.3 ohm (6V/20A=0.3 ohm) will be therefore shown via the display.

On the other hand, if the maximum no-load voltage is 12V, the ground bonding tester will clearly show the 0.3 ohm test result and alarm may also be generated because the setting limit is 0.2 ohm maximum. That is to say, it doesn't matter the voltage is 6V or 12V, the failure's result will be obvious.

I guess the meaning of "flash through" in CTL decision is that the oxidized joint will be undetected and the test result is also POSITIVE when test with voltage exceeds 6 V, say 12 V. But I can not find any evidence to support such opinion although I am a engineer.

Ah, this is perplexing.

[Peter Selvey]

An oxidized joint may not conduct well at low voltages but can conduct with higher voltages (which can be a pain in design trouble shooting!).

This is to say, a joint with oxidation does not follow ohms law, rather it acts like a voltage dependant resistor.

Such a joint may read 1000 ohms at 5V, but the same joint can read 0.1ohm at 10V.

Thus the test voltage is important, and should be as low as practical so that bad joints can be reliably detected.

[sankui]

Thank you again, Peter. Your answer seems to be reasonable. I agree that test voltage should be as low as practical. However, my question still is, if the issue (oxidized joint acts like a voltage dependant resistor) exists in medical equipment, it seems no reason not to consider the same issue exists in non-medical equipment.

Yes, it should be more serious situation in medical device. But an oxidized joint in the protective earth path is not desired in all product standard. Furthermore, IEC60601-1 3rd edition has adopted IEC60950 as the means of operator protection (MOOP). In fact, the patient will be in the same risk (oxidized joint can not be detected since the 12 V test voltage in IEC 60950) if they use personal computer or the like.

[Peter Selvey]

The difference (my guess) is that for non-medical equipment, 12V on the frame of a device is not considered hazardous. Skin resistance, no return path due to normal floors, shoes etc are an additional probability factor that makes 12V safe (noting that it is already an unlikely situation). For non-medical, a low impedance path for electric shock is only assumed for voltages above around 40Vac / 60Vdc, when skin resistance starts to break down (skin is also a voltage dependant resistance).

For medical devices, they assume skin resistance is bypassed and the patient is also earthed by a low impedance path, so that even 0.1Vac is considered hazardous in normal condition. While this is certainly overkill for most situations, it is a basic assumption for medical devices, and sets it apart from other standards such as IEC 60950, 335, 1010 etc.

Sure, with this assumption even 6V is dangerous, but we need the 6V for the test. And realistically, it is a far fetched sequence of events.