I’ve at all times been intrigued with high-precision devices, because it often represents the most effective of engineering design, craftsmanship, magnificence, and even artistry. Certainly one of earliest and nonetheless greatest examples that I recall was after I noticed the weigh-scale design by the late, legendary Jim Williams, revealed practically 50 years in the past in EDN. His piece, “This 30-ppm scale proves that analog designs aren’t lifeless but,” particulars how he designed and constructed a conveyable, AC-powered scale for dietary analysis utilizing normal elements and with extraordinary necessities: excessive decision of 0.01 pound out of 300.00 kilos, accuracy to 30 components per million (ppm), and no want for calibration throughout its lifetime.
Jim’s mission was a non-production, one-off unit and whereas its schematic (Determine 1) was clearly necessary, tells solely a part of the story, there are a lot of extra classes in his description.
Determine 1 This schematic from Jim Williams’ 1976 EDN article on design of a precision scale teaches many classes, however there’s far more than simply the schematic to know. Supply: Jim Williams
To fulfill his targets, he recognized each supply of error or drift after which methodically minimized or eradicated every one through three methods: utilizing higher, more-accurate, more-stable elements; using circuit topologies which self-cancelled some errors; and offering extra “insurance coverage” through bodily EMI and thermal boundaries. On this circuit, the front-end wanted to extract a miniscule 600-nV sign (least important digit) from a 5-V DC degree—a really tall order.
I spoke to Jim just a few years earlier than his premature passing, after he had written a whole bunch of different articles (see “A Biography of Jim Williams”), and he vividly remembered that design and the article because the occasion which made him understand he might be a designer, builder, and expositor of really distinctive precision, principally analog circuits.
After all, it’s one factor to handcraft a single high-performance unit, but it surely’s a really totally different factor to construct precision right into a reasonable production-volume instrument. But firms have been doing this for many years, as typified by—however definitely not restricted to—Keysight Applied sciences (previously often called Agilent and previous to that, Hewlett-Packard) and lots of others, too many to quote right here.
Proof of that is seen within the newest technology of optical take a look at and measurement devices, designed to seize and rely single photons. That’s definitely a mark of maximum precision as a result of particular person photons typically don’t have a lot power, don’t prefer to be assessed or captured, and self-destruct whenever you take a look at them.
I lately got here throughout one other instrument that takes a easy operate to an excessive degree of precision: the DC205 Precision Voltage Supply from Stanford Analysis Programs. This desktop unit is far more than only a energy provide, because it supplies a low-noise, high-resolution output which is usually used as a precision bias supply or threshold in laboratory-science experiments (Determine 2).
Determine 2 This unassuming desktop field represents an impressively excessive degree of precision and stability in an adjustable voltage supply. Supply: Stanford Analysis Programs
Its bipolar, four-quadrant output delivers as much as 100 V with 1-μV decision and as much as 50 mA of present. It gives true 6-digit decision with 1 ppm/°C stability (24 hours) and 0.0025 % accuracy (one 12 months).
Two different options caught my consideration: it makes use of a linear energy provide (sure, they’re nonetheless necessary in specialty functions) to reduce output noise, presumably just for the voltage-output block however not for all the instrument. There’s additionally the inclusion of an DB-9 RS-232 connector along with its USB and fiber optic interfaces. I haven’t seen an RS-232 interface in fairly some time, however I presume they’d purpose to incorporate it.
The block diagram within the Consumer’s Handbook reveals comparatively little, besides to point the unit has three core parts which mix to ship the instrument’s efficiency: a low-noise, secure voltage reference; a high-resolution digital-to-analog converter; and a set of low-noise, low-distortion amplifiers (Determine 3).
Determine 3 As with Jim Williams’ scale, the core capabilities of the SR205 look easy, and could also be so, however additionally it is the unrevealed particulars of the implementation that make the distinction in reaching the specified efficiency. Supply: Stanford Analysis Programs
I definitely wish to know extra of the design and construct particulars that squeeze such efficiency out of in any other case standard-sounding blocks.
As this unit targets lab experiments in physics, chemistry, and biology disciplines, it additionally features a function that standard voltage sources wouldn’t embrace: a scanning (ramping) functionality. This triggerable voltage-scanning function offers person management over begin and cease voltages, scan velocity, and scan operate, with scan speeds settable from 100 ms to 10,000 s, and the scan operate can both be a ramp or a triangle wave. Additional, for working within the 100-V “hazard zone”, the person should plug a jumper into the rear panel to intentionally and consciously enable operation within the area.
Along with the DB-9 RS-232 interface supporting legacy I/O and an optical hyperlink for state-of-the-art I/O, I seen one other fascinating function known as within the well-written, readable, crisp, and clear person’s guide: the right way to change the AC-line voltage setting. Some devices I’ve seen use a slide swap and a few use plug-in jumpers (don’t lose them), however this instrument makes use of a thumbwheel rotating selector as proven in Determine 4.
Determine 4 Even a high-end instrument should cope with totally different nominal power-line voltages, and this rotary swap within the unit makes altering the setting simple and resettable. Supply: Stanford Analysis Programs
Briefly, this can be a very spectacular standard-production instrument with precision specs and efficiency, with what looks as if a really affordable base value of round $2300.
I give it some thought this manner: In the true world of sensors, sign conditioning, and total analog accuracy reaching secure, correct efficiency to 1% is doable with affordable effort; attending to 0.1% is far tougher and reaching 0.01% is an actual problem. But each customized and production-instrumentation designers have mastered the artwork and ability of going far past these limits.
It’s much like after I first noticed an inventory of basic bodily constants such because the mass or second of an electron, and which had been measured (not outlined) to seven or eight important figures with an error solely that final digit. I felt compelled to do additional investigation to know how they reached that degree of precision and confidence, and the way they credibly assessed their sources of error and uncertainty.
What’s the tightest measurement or signal-source accuracy and precision you might have needed to create? How did you affirm the specified degree of efficiency was really achieved—if it was?
Invoice Schweber is an EE who has written three textbooks, a whole bunch of technical articles, opinion columns, and product options.
Associated Content material
