How to understand the "full scale peak-to-valley output" mentioned in IEC 60601-2-27

[raymond02302]

Dear everybody

Currently, i am conducting tests per IEC 60601-2-27. In clause 50, many subclause mentioned the term "full scale peak-to-valley output" .
ie, about the accuracy of signal reproduction test,the standard says "Adjust the SENSITIVITY of the EQUIPMENT to 10 mm/mV. Adjust the signal generator to produce a peak-to-valley output on a display device that is 10 % of the full scale peak-tovalley output. Increase the output of the signal generator by factors of 2, 5 and 10. The displayed output shall be linear within ± 20 % of the full scale output."

How can i understand the highlighted term, does it mean the channel width or the 10 mm due to the sensitivity.

thaks in advance!

[Peter Selvey]

For a printer strip that has for example 50mm, this would mean a signal that is approximately 5mV peak to peak (+2.5mV / -2.5mV). I say approximate, as most systems will have small sensitivity errors, so the standard usually wants you to adjust the source to give exactly full scale.

On a screen this may be more difficult to define the boundaries of "full scale", as some manufacturers allow the signal to drift well beyond the allocated space. But the designers of the screen layout should be able to indicate what they consider to be "full scale", e.g. point at which two adjacent signals will not overlap. This would be the channel width as you say.

On a related point, the new edition just published (IEC 60601-2-27:2011) has fixed many of the errors in the 2005 edition.

So for example, the 2005 edition says start at 10% of full scale, and increase to 20%, 50% and 100% which makes no sense from a laboratory point of view, because at 10% the resolution is too small.

The 2011 edition says start at 100% full scale, then reduce 50%, 20% and 10% which makes much more sense. Another example is the multichannel cross talk test which is now clearly defined where to apply the signal and where to verify the result.

[raymond02302]

In Reply to Parent Post by Peter Selvey

For a printer strip that has for example 50mm, this would mean a signal that is approximately 5mV peak to peak (+2.5mV / -2.5mV). I say approximate, as most systems will have small sensitivity errors, so the standard usually wants you to adjust the source to give exactly full scale.

On a screen this may be more difficult to define the boundaries of "full scale", as some manufacturers allow the signal to drift well beyond the allocated space. But the designers of the screen layout should be able to indicate what they consider to be "full scale", e.g. point at which two adjacent signals will not overlap. This would be the channel width as you say.

On a related point, the new edition just published (IEC 60601-2-27:2011) has fixed many of the errors in the 2005 edition.

So for example, the 2005 edition says start at 10% of full scale, and increase to 20%, 50% and 100% which makes no sense from a laboratory point of view, because at 10% the resolution is too small.

The 2011 edition says start at 100% full scale, then reduce 50%, 20% and 10% which makes much more sense. Another example is the multichannel cross talk test which is now clearly defined where to apply the signal and where to verify the result.

thank you for your detailded interpretation.
but i am still a little intangible.

On a screen, the "full scale" is the maximum scale designed by the manufacturer or the actual scale measured by us.
for example, the maximum scale designed by manufacturer is 30 mm (channel width) but the actual scale measured by us may be larger than that.
so, which is the full scale, the former or the latter?

[Peter Selvey]

I would use the scale designated by the manufacturer.

For patient monitors, the manufacturer may allow the ECG signal to go beyond the channel so the user can still see the signal in case of overload or other problems with the baseline (e.g. poor electrode contact).

However, I don't think it is resononable to worry about accuracy and sensitivity in these cases. It is just a kind of user trouble shooting feature.

If the manufacturer says 30mm then test for 30mm.

[raymond02302]

thank you