Buying a digital multimeter

[Will]

Good stuff Ed - very interesting ! I don't believe it was contactr failure in my case - the display became more and more unreadable as different sections of the whole display area blackened - Where they were not blackened all seven segments were still registering. The new one still has the slide-in pressurized contacts you describe. Thanks

[Dean Huster]

Will, I can guarantee you that every instrument sold by Tektronix goes through a complete calibration before it leaves the plant in Beaverton. If it doesn't meet specifications, it's usually repaired and recalibrated. Always been that way, always will. Hewlett-Packard/Agilent and Fluke are the same way. You can't have an instrument with pots and variable capacitors in it NOT go through a calibration process.<p>The calibration is a lot faster now than it used to be, that's for certain. Since the advent of the 7000-series in 1972, Tek has used 1% resistors for all the critical circuits rather than 5% types, which improves reliability, increases stability and narrows the adjustment window of the various calibration adjustments so that the exact calibration point is easier to hit. In fact, in many cases, a lot of calibration adjustments are eliminated by using the 1% resistors for silly things like transistor bias and stage gains. Of course, they've always been used for the really critical stuff like attenuators and sweep timing.<p>The calibration was sped up considerably with the use of automated test equipment and especially with the advent of the DSO, where the bulk of the calibration is done automatically by the instrument itself. With a modern DSO, the instrument can be connected to an ATE system through its own GPIB connector and calibration is a hands-off operation.<p>Fluke, like many DMM manufacturers (and I'd slap a big exception on most of the Asian imports) have made their instrument calibration pretty quick and easy, but they still get at least a pseudo-calibration. For instance, Tek, hp and Fluke have been using laser-trimmed shunts for a long time now. Each shunt is manufactured automatically and laser-trimmed to exact values so that ammeter calibration is a thing of the past. Once the basic DCV and ACV functions are set, DCA and ACA are automatically right on, usually as accurate as the voltage functions.<p>The "old" Simpson 260 VOM has five adjustments that have to be made before it leaves the plant. And you can bet that those adjustments are made, setting the basic DC accuracy (2 pots), the AC accuracy (2 pots) and the 10A range (moving a slider on the shunt). For what they charge for a Series 8 now, that danged 260 had better run through a calibration step before it comes to me!<p>So, I still hold to my position that your average instrument isn't born and shipped, hoping to the Almighty that it falls within some hallowed statistical curve and that the customer doesn't notice that his new DMM is off by 2%.<p>One of the previous posts mentioned the ISA - Instrument Society of America. I've found that the ISA (the group was very active in Oklahoma City) is mostly directed at industrial instrumentation, i.e., 4-20ma current loop devices, PLCs, load cells and stuff like that rather than what I consider to be "true" metrology such as the calibration of standards and test equipment. Not to say that the ISA is bad, just not what we in the metrology segment ever messed around with.<p>Oh. Prior to teaching (21 years now), I spent 12 years in metrology in U.S. Navy PMEL and with Tektronix, and that's where my experience with the subject comes from. Granted that someone who worked for B+K will have a different slant on things (their stuff may not have gone through calibration :&gt <p>Dean