AV 9000 Checklist Item Under Test:
The integrator shall provide a list of the calibrated equipment that will be used on the performance tests in the Staging and Commissioning Checklists, along with the calibration date and serial numbers for each.
A measurement is only as good as the person and equipment used to take it. The best practice is to have all test equipment calibrated to a traceable standard. This removes all questions. If a recently calibrated meter says it sees 1.0 V between its leads, there is indeed 1.0 V between its leads. The question arises, however, regarding how often equipment needs to be calibrated (or at least verified if it has no adjustment). That is up to you, as the user of the test equipment.
If your organization has a quality management system in place, these processes will be documented. If the processes ever let a team member down (a device measures 0.8 V across something that is 1.0 V… and therefore the test was not valid), there are measures in place to make sure it will never happen again, such as getting the test gear calibrated every year instead of every two years (continual improvement). As an AV professional, I must be able to trust my test equipment.
We were working on a system that sounded absolutely great. I knew it was going to pass the sound performance test with flying colors. The integrator knew it too. They had completed their commissioning checklist weeks before and felt very confident in the system. Despite this, I measured a dismal 27 dB signal to noise ratio. It should have been greater than 60 dB based on what I heard, and it had to be greater than 55 dB to be accepted by the client. I asked the integrator about it, and he confirmed he measured 68 dB or better from all sources with his meter.
We looked at the calibration stickers on both my meter and his meter. His meter was calibrated a month before, and mine was coming up on two years. His meter “won”. Not only had it been calibrated more recently, but it corresponded to what we heard in the room. My meter was way off and needed a tune up, as did my calibration procedure for that particular meter. Two years was clearly too long between calibrations.
So, how often do you need to calibrate the test equipment? If it is heavily used and lives are at stake, perhaps every month would be a good idea. If it is used moderately, perhaps every year is acceptable. If, after several years of calibration, no adjustments have been made to the device, perhaps every two years is satisfactory. The important thing is that:
1. The test equipment does indeed get calibrated. If it can’t be calibrated, at least verify it is measuring true. Many people don’t realize they have to do this at all. They just buy the equipment and trust it blindly. Owning test equipment comes with additional costs of ownership, including calibration.
2. Create a calibration schedule and stick to it. Put the calibration date on a sticker on a device, and make sure it also has an expiration date. If the device is used after the expiration date, the test might not be valid, and it needs to be re-cablibrated.
If you are ISO 9000 or AV 9000 certified (or if you just want to be fancy), all test equipment must be calibrated or verified to a traceable standard. That doesn’t necessarily mean that every device in the organization has to periodically go out to a calibration company. This can be very costly and disruptive to operations (I hate when test equipment leaves me for weeks…I miss them…I have problems…I digress). This just means that each device needs to have a traceable link to a known standard. So, what if you only get a generator and a voltmeter calibrated by a service, and then calibrate or verify all other test gear with those calibrated devices? You just provided a traceable link to a known standard, saved a ton of money in time, services, and lost opportunity, and most importantly, verified that you can trust all your devices are measuring true. If you do this periodically and print your own dated calibration labels, you are well on your way to satisfying that aspect of the ISO 9000 and AV 9000 standard. Also, no one will be able to question your measurements. That’s a nice safety blanket.
In reality, most digital test equipment holds up very well. The biggest problem we see is untrustworthy measurements immediately before the batteries fail. If their performance is verified once every year or two, you will be in good shape. The biggest concern is the mobile device measurement apps being used to commission systems. If they are calibrated, they work remarkably well, but they must be calibrated. If you roll your own calibrations, this won’t take much time at all.
In fact, my iPad and iPhone with their acoustic accessories are a very important part of my testing arsenal, and I verify their performance every six months. However, most apps, out of the box are way off the mark. I had an operator try to convince me that the subway he took to work, typically measuring 100+ dB-SPL, A-weighted (dBA) on the platform, was only 60 dBA “because his iPhone app said so”. When I asked to see the calibration records for the free app he downloaded on said platform, I just got a blank stare.
Make sure you can trace the performance of all your test equipment to a known standard periodically. Owning the test equipment is important; making sure it works accurately is even more important.
Originally appeared in Sound & Communications Magazine, 12/1/13