EBME Forums Forums Maintenance Technical Assistance Recording,Storing and Analysing Electrical Safety

For a long time I have been having a debate with myself and others about whether we should be downloading and storing test results. There appears to be two arguments.

The first is it's Pass/Fail. If it fails fix it so that it passes, no need to store test results. If it can't be fixed scrap it.

The other is to store pass results, analyse at a later date to ascertain any creepage that may indicate failure of the electrical insulation or leakage that may compromise patient or staff safety.

This is another old chestnut that keeps insisting on popping up every now and then, Joe. I think we all decided years ago that downloading and storage of megabytes of data for analysis was literally a “waste of space”. All very well in theory, but who actually has the time to wade through all that stuff, just to detect a trend towards failure? Just because it’s possible (easy) to download from modern safety testers, doesn’t mean that it’s worth the craic (so to speak). No, Mate. Your first option is the one to go for. Always keep things simple, in my opinion.

Article in March 2007 Scope(IPEM) entitled "International Standard for Routine Electrical Safety Testing of Medical Electrical Equipment" under section "General proposals in the standards" seems to indicate in second paragraph.

Quote " Reference (ideally the first) values for all measurements should be recorded. If reference values for leakage currents are over 90% of the allowable values, then the trend must be investigated".

Does this statement not indicate that some record should be kept and analysed or am I misinterpreting this?

If you intend to do this then you'll also have to analyse the trend in the calibration data for the electrical safety analyser won't you? To asses how calibration of the analyser influences the measurements obtined, nevermind variatons in the device.

Another complication is that if different safety analysers are used for subsequent tests on the same device this means you'll have to analyse the calibration of each tester that tests have been performed each time the analyser is calibrated.

This would enable you to determine how the safety test results for a particular device have been inflenced by different analysers' measurement error or for variation in error between calibrations of the same safety analyser.

Some devices that have been developed for testing to DIN 0751, e.g. Gossen Metrawatt Secutest SIII medical safety tester, are based upon the recurrent test standard that's discussed in the Scope article - these instruments have the facility to include the measurement errors, i.e. measurement error, in the readings obtained.

This facility simplifies analysis of safety trends between calibrations of an analyser (assuming there is drift or adjustments are necessary) and between analysers of the same type, but with slight variations in measurement errors, I suppose.

Hi,
The new Rigel 288 will tackle this problem. Soon to be released feature will include to compare test data on the spot. Comparison is triggered by a threshold which is user configurable. In other words, only test data which has changed over a certain percentage or absolute value (whichever is greater) will be displayed. This ensures that you do not waist time looking at leakage measurements which are always below 5uA or so. The following link has more details on the 288. Don’t hesitate to contact me if you want a demo of the product;
http://www.rigelmedical.com/products/rigel_288.asp

JB,

One problem I have encountered in the past when considering storing test data on test equipment was their limited memories. How many results is it able to store?

Hi John
The Rigel 288 can store upto 10,000 equipment results (ie 10,000 full IEC 60601 test results)
Scheduling software makes it possible to keep more data on the PC and transfer only that data that is needed for the retest. So in theory you can upload 5000 equipment, re-test them and keep both results in the memory.

In my experience the only things that fail electrical tests are detachable mains leads and the odd heating element(wax baths). I don't need stored trends and expensive automatic testers to tell me to replace these items - this would be the conclusion of my risk assessment.
If equipment rarely fails electrical tests, why is there this drive to record the results?

Cheers
Mark

Why NOT retain test results. It takes about 10 minutes per week to downlaod them. If you use a tester with PC link & software it automatically saves the data on the PC in any case.

If something goes seriously wrong with a piece of kit and you get asked to provide maintenance records then you will be bound to be asked "Where are the test results then ?"

As for spending time analysing drift in values I've never met an engineer who did that. If the unit passes you get on with the next task. If it fails you fix the problem re-test then retain a before and after result.

If you get audited the question always asked is "Where's the evidence ?" If you chucked away your test results where's the evidence you ever did the tests?

Also if you've trained a new technician and passed them to work unsupervised it's very usefull to be able to check their test results for the first couple of months just to be absolutely sure they are running the tests correctly and don't need help.

Marc

Joe,
The thing to remember when reading articles from Scope or any other IPEM dcuments is that they are written by Scientists. These guys have the time and freedom to postulate and have a tendancy to overthink situations. As engineers we make decisions based on real life experiences and over the last twnty odd years of safety testing I have never had a situation where we've been able to predict the failure of an e.s.t. by checking the previous records. The only useful requirement for storing e.s.t. results is for legal purposes, i.e. to prove that it has been done.
So, consider this punk! Are you a Scientist or are you an Engineer? Go ahead, make my day.