EMF Consulting and Safety: On the Importance of Instrument Calibration
EMF Consulting & Instrument Test Calibration
When quoting a job or project to a customer, a question that comes up from time to time is:
“Are your meters calibrated?”
An understandable query, but the answer may surprise you.
EMF consulting is a specialty field that most of my clientele have very little familiarity with and it is wise to try and ascertain if you are getting good data for the money paid.
However, my extensive technical experience covering over 25 years in several interdisciplinary fields (including 10 years experience and certification in the calibration of instrumentation) has given me a rather unique perspective and stance about the relevance of this question.
My short answer is that as long as it is a digital meter of semi-professional grade or better from a reputable company and is in proper working order from the date of purchase, observing calibration intervals (which are usually a holdover from the old analog days) is not as important as some may think PROVIDED that the meter is known to be initially accurate within a certain % depending on the application. Very often, you can perform a cross-check with other meters and as long as they are in close agreement, that indicates proper operation.
The only exception to this is if:
- The particular meter/model is known to have serious accuracy / unpredictable performance issues such as one model I discuss on this site.
- Damage from impact, fluid intrusion or a corroded battery has or is suspected to have caused operational issues to the sensors / internal electronics.
This may sound unprofessional or even contradictory to other post on the importance of using a high quality meter, but I have pondered this issue for a long time and I can give a number of strong reasons for this position:
1) Unless you are doing actual research grade work, (such as in academia or R&D) are calibrating other instruments, or you are performing mission critical work (such as an infusion pump dispensing medication to a patient in a hospital) you simply don’t need the absolute best accuracy to accomplish the mission of the instrument.
Close is good enough for almost any application of EMF surveying.
For instance, if an EMF meter reads 1 milliGauss, (mG) but the true NIST traceable reading is 0.9 or 1.1, being off by 0.1 mG or 10 % should not be the deciding factor between buying a home or not.
2) Readings of magnetic fields, radon and RF naturally fluctuate over time, so what is more important is looking at an average.
3) Different meters have different frequency weightings so no two meters will be in complete agreement anyway.
4) You cannot always trust a calibration sticker. A dirty secret in the industry is that a lot of money is spent on calibrations that are incorrectly performed or not even done at all. I know this as a fact because I have not only had calibration technicians tell me, but I was IN the instrument / radio calibration business myself.
Just because an official looking piece of paper says that a meter is accurate, you are still trusting the honesty of the person / company that certifies the equipment and often there is not a clear “chain of custody” that holds anyone responsible or that the right / qualified individual actually performed said calibration. I am not saying calibration isn’t worth doing, nor am I accusing any calibration company in particular, just that you should maintain a healthy amount of wariness about this issue.
A controversial subject analogous to this is the autograph authenticity industry – try Googling that and I believe you will find some of the stories very eye-opening and sobering as to the real value of an expert’s opinion when volume vs. quality directly impacts the bottom line. And frankly, it is easy to forge a very official looking calibration certification paper for most instruments.
5) Different types of measurements can be very difficult to get accurate because of the physics involved. Not everything is as simple as knowing the speed of a vehicle, for instance. (something which can be measured with excellent accuracy)
For example, received and transmitted RF power is notoriously difficult to nail down due to reflections, constructive / destructive interference, etc. so that getting within 3 dB (or 50 % accuracy) is considered reasonably good and the reading can change depending on your antenna gain, frequency weighting or just moving the instrument a few centimeters to one side.
In plasma physics, if your measurements are within a factor of 2 of calculated values, that is considered valid. In other words if you expected to get 5 and get 8 instead, (a 50% + error) that is still in the acceptable ballpark of even high level research data.
6) Calibration at annual intervals is an old and in my professional opinion for most instruments, unnecessary “feel good” tradition that no longer matches the actual requirements of modern solid state electronics that have a fair degree of drift compensation and stability compared to older, analog devices.
If you have a new instrument made from a reputable manufacturer and after only a year it is found to be significantly off, then I would strongly suspect it was defective / improperly calibrated to begin with.
And this is not just my opinion; a number of modern instruments now manufactured have recommended calibration intervals that are either upwards of 10 years + or are even guaranteed for the life of the instrument barring serious damage such as that from a fall / water immersion, etc. which would probably place the meter in a category of “inoperable and uneconomical to repair” so it would be replaced with a new one anyway.