EBME Forums Forums Maintenance Technical Assistance Continuity Testing

AC Earth Bond Test (Rigel 277)

Test Voltage.....6V rms nominal (no load)

Test Current.....100mA, 1A, 10A, 25A (0 - 0.2Ohm range as per IEC60601)

Range....1A, 10A, 25A.....0.00. - 19.99Ohm.

100mA.....0.00. - 4.99Ohm

Resolution....0.01Ohm

Accuracy....1A, 10A, 25A …+/- 5% of reading, +/- 2 digits

0.10A……...…..+/- 5% of reading, +/- 4 digits

As far as my reckoning on the back of a piece of scrap A4 goes, at typical 2m IEC320 lead resistances of around 0.075Ohm and at 100mA test current, the worst-case uncertainty in the measurements in the Rigel 277 indication may be as much as +/-60%, i.e. the indicated earth bond reading could be up to 1.20Ohm (This assumes there is good contact resistance at connections that does not require high "wetting current" to clear oxidation, etc, and no thermal effects in the earth). At best uncertainties of up +/-15% may be evident for earth bonds of 2.0Ohm (BS EN 60601-1 limit) at 25A test current and they may potentially be up to +/-25% at 0.10Ohm.

Perhaps my arithmetic is a bit shaky but I think this demonstrates, that for a tester in widespread use, limited resolution of the tester and errors in the measurement may limit the accuracy and most likely the accuracy & repeatability of earth bond measurement at lower currents and for lower resistances compared to higher resistances measured at higher test currents - even for "state of the art" technology available currently. Look at the published specifications for the Rigel 277, for example, in the operators' manual.

After looking at the information provided by JB it seems to me that the problem of poor contact resistance exacerbates the problem of poor accuracy & repeatability when insufficient wetting current is available - thus spurious "fails" are more likely to be produced at low current. Even at high currents measurements may be more repeatable (due to the wetting current maintaining low contact resistance) but not particularly accurate, as outlined above.

I think RICK was correct about this wetting current business - although a well-trained and experienced colleague I've worked with for a few years has commented on it in the past. How's the training coming along Alan? Have you found the degree course you're looking for yet?

I don't know about shaky...

But I would go by my abacus, your calculation should have been 0.12ohms, not 1.2ohms.

Well spotted - typo not calc - it was 10pm.

Personally, I like the post by Les of Hartlepool the best! Says it all, really.

Anyhow what I was getting to is that in the draft IEC 62353 standard I have access to (the standard that's going to be introduced imminently, apparently geared up to be used with the 3rd edition 60601-1 standard) the accuracy of the respective measuring functions [of measurement equipment, e.g. medical safety testers] within the range marked or declared by the manufacturer shall not exceed +/-10% of the measured value as calculated in IEC61557-1. Do testers out there currently perform to within +/-10% uncertainty for all measured parameters?

They may well do, but are the contacts in their sockets up to scratch?
The ESTs that I have seen perform their 'self calibration' using the earth continuity fly lead as a reference to an earth continuity pin seperate from the testing socket. I have definitely encountered at least one EStester where the Earth pin on the inbuilt socket has accumilated a degree of oxidisation beyond what I would have deemed acceptable and that would be regardless of any poor contact on the equipment's mains lead earth pin.
Therefore, when you calibrate the floating lead, you are not actually calibrating the testing circuit. My question would be; do the calibration houses/suppliers of the testers consider this when they are ratifying the annual calibration of your testers?

Hi Richard,

I am a committee member of SC62A / WG14 who drafted the IEC 62353. Feel free to contact me if you want to discuss any querries you have regarding this standard. As for your comment on 10% as per IEC 61557-1, when using higher current eg 10A/25A on the current safety analysers (Rigel / Fluke / Metron) you will be able to meet the accuracy.(note that the mentioned spec on the datasheet is worst case and does not reflect the actual accuracy of a calibrated unit hence the recommendation of manufacturers for annual calibrations.) Rigel adjust their safety analysers to within 1 digit at 0.20 Ohm which is 5%. This is by far more than required as moving the croc clic - often used) or using a 4 mm probe in the earth eq.pot. point give appr. 0.02 - 0.05 difference due to bulk resistance. (see my presentation at IPEM / June 06)
Hope this clarifies a little
Thanks

Hi John,

Thanks for that if I need anything clarifying RE: standards, etc, I'll get in touch.

The point of my postings, from a lay-persons perspective RE: continuity testing, fundamental design limitations, etc, and as an engineer maintaining medical equipment, is not to infer any problems with the calibration of specific devices or how well they compare but has been to outline the potential problems with low-current testing performance generally - nevermind contact problems and physical issues in the DUT.

Also it's to comment that my colleagues and myself are probably going to be stuck in the difficult position of intuitively being aware of and of physically observing problems but may be obliged to follow guidelines (DB6801 supp1), recommendations (IPEM), upcoming standards (62353), that may push us towards test regimens (1A or less) using instrumentation and techniques that some feel may not be up to what's required for earth bond testing and that many will feel uncomfortable with.

There are certainly disadvantages, RE:accuracy and repeatability of earth bond resistance measurements, taken by safety testers in current usage when testing at 1A or less, that need to be weighed up against the likelihood of damage to a DUT - which is rare in my experience of maintaning medical equipment, although I have experienced a couple of problems that I've attributed to 25A testing.

However when routine safety testing, in volume, the last thing I'd want to be doing is routinely trying to sift out the incidences of bad connections, between the DUT and the tester, from potentially faulty earths - especially not if I know, from years of experience, a device can tolerate 10A or more, routinely, without the connection problems that may be observed at lower currents.

I hope, if you read my earlier postings, you noticed that I conciously quoted "worst-case" prior to all my comments concerning accuracy and resolution (although resolution is likely to be an irreducible effect RE: calibration, eh?). I'm aware that calibrated intruments used at higher currents are likely to meet the +/-10% requirement after calibration and should not pose any problems in use when used at currents significantly greater than 1A.

However, as you've inferred yourself, John, it may be very difficult, if not impossible, for some testers in widespread use, currently, to meet the +/-10% using 1A test current or less. This has consistently been my argument in earlier postings.

In future, if I'm put into a position where I'm required to test at 1A or less and record the results, I'd like my continuity test to be able to give me an accurate figure for resistance (thus be able to accurately resolve to at least 3 sig figs between 0.00Ohm and O.30Ohm) that's repeatable - especially if I'm going to stand there fiddling with a mains cable looking for small changes in earth resistance that might indicate a dodgy earth!

Otherwise we may just as well have a Go/No-Go type of test indication that, besides not giving any clues or feedback to what's happening if there's a problem, in my opinion, can lead to ignorance of what's being measured and complacency on the part of those doing the test.

I hope that when IPEM publish their guidance notes on safety testing and IEC their 62353 standard that they include rationales that are meaningful and intended to educate those who carry out safety testing rather than expecting them to just take it at face value. Shame we'll have to pay for the privilege though.

Cheers.

Hi richard
You hit the nail on the head. All test equipment currently on the market (with exeption of the PrimeTest 300) have problems when testing at currents less than 1A. As per our research you need at least 5-6Amps to overcome common issues (wetting contacts).

The IEC 62353 clearly states 'use a current of at least 200mA at 500mOhms - max. voltage of 24V'. this does not exclude any current used at present. IEC 62353 also states that IEC 60601-1 tests can be carried out if the safety of the test engineer can be guaranteed. We can go into a big debate about IEC 62353 and future encertainties but this is most certainly not the intend of the initial post.

Do not fear the future. People already do what they feel necessary based on experience and risk assessment. Nothing much will change in the future. Just a little more help and consistency across Europe.

Feel free to contact me should you like to discuss IEC 62353.

John

Just for Colin Tate's benefit I'd like to mention that when non-medical equipment is connected to a medical device (by common mains distribution or functional connection) that the system, combined, becomes a medical system.

A non-medical device with or without an accessible earth must be tested to it's own non-medical safety standard or comparable methods, since it must be safe, and also CE marked, to be allowed to connect to medical devices.

It must also be tested as part of the interconnected medical system (to meet 60601-1, 60601-1-1 or as the manufacturer stipulates on installation, or as is indicated in the certificate of conformance, or type-testing certificate, if provided).

Incidentally a mains powered device that has no protective earth but a functional one (or even a device that is internally powered), that is connected to a Class 1 system and is exposed, will be tested to Class 1. There is no way to avoid this since we're obliged to probe accessible, conductive, parts for continuity/leakage under various test conditions.

So Colin, irrespective of whether parts of a medical system are medical or non-medical, if we can't use medical device testers that give reliable and meaningful continuity readings, at 1A test current or less, then no, we don't need 1A continuity test current or less when testing medical devices & systems (or any other Class 1 device for that matter, in my opinion).

Not unless there's an explicit warning or indication on the equipment (whether its got a functional or protective earth, or not) that a higher earth continuity test current is damaging or until testers are improved upon to give more reliable, meaningful, results at 1A or less (irrespective of whether the source problem is with wetting currents or poor contacts).