Developing a Micro-balance for Production Lines

A new model with sensitivity of 1 μg has been added to the AD-4212 series of production-line weighing instruments. In this installment, I will continue the discussion from Story 2 and further discuss the development of production-line weighing instruments. Product development for the AD-4212 series moved forward with the minimum weighing value progressing in the following order: 0.1 mg, 0.01 mg, 1 mg, and 10 mg. One of reasons that the 0.1 mg model was developed first was that we had been selling a balance for production lines with a sensitivity of 0.1 mg, considered the level of analytical balances, for about 20 years as a special order item. In places where weighing at a sensitivity of 1 mg or greater is performed, the items to be weighed and the weighing jigs tend to be fairly large and heavy so we decided that ordinary precision balances sufficed and the demand for a more compact balance was low.

At the time, the new mass sensor developed for the HX Series of general-purpose balances was being used in production lines. At the time it was developed, the HX Series of general-purpose balances had excellent characteristics compared to most previous general-purpose balances, including a fast response time. Furthermore, the mass sensor that was developed at the time is still being used today in the HR Series of analytical balances, around 20 years after its release.

The HX Series was developed near the end of the “bubble economy” in Japan and boasted unusual level of high performance for the time. It was a top loader (referred to as a general-purpose or precision balance) that had a pan on top of the mass sensor, and it had a quick response speed. At the same time, it had a capacity of 100 g and a minimum weighing value of 0.1 mg, which put it in line with an analytical balance. However, because it had so many added functions, the balance was quite expensive and consequently did not sell well as a general-purpose balance. Nevertheless, thanks to its fast response and high repeatability as a production-line balance, it was adopted by fields with production lines that were said to require a weighing time of one second or less. The HX Series was held in high esteem because its fast response allowed for highly accurate weighing in moving production lines, something that production sites had been given up trying to achieve until then.

The HX Series was the first new product that I developed for the first time from the mass sensor a few years after I joined A&D. As I mentioned before, 20 years ago our sales department was quite demanding, and the product specifications ballooned during the planning stages, just like the economic bubble that Japan was in at the time. As an example, this was the first time that A&D included an internal calibration weight with a general-purpose balance. Other requests included a display section separated from the weighing section, a weighing section that could be operated via a cable, the ability to place the display section behind the weighing section, the display of weighing values on an analog display, and a communication port as standard, which was rare for a general-purpose balance at the time. As the R&D leader at the time, developing a device with these specifications as standard was a huge responsibility that required creative thinking and hard work.

In the end, we released the HX Series while meeting all the demands of the sales department, but its sales numbers as a general-purpose balance were dismal. I learned that, despite the many advanced functions available, users aren’t asking for much from weighing instruments. In other words, the HX development taught me that weighing instruments are just that, nothing more and nothing less. At the time, I struggled to meet the demands of the sales department as a product leader, but by not running away from unreasonable (or so I thought at the time) demands and making proactive proposals and responses, I feel that I was given an excellent opportunity to grow as a designer. Therefore, I am thankful for the various things brought to my attention by the sales department at the time.

This experience taught me the inevitable conclusion that users want balances that are compact, quick, and durable for production lines. For example, we ran counter to this conclusion about compactness by installing a calibration weight inside a production-line balance. Users of production-line balances that have a capacity of 100 g and a minimum weighing value of 0.1 mg do not weigh items of 100 g. The tare loaded as jigs are 100 g or less and the items weighed are several grams at most. If you think about how often calibration would be required, you realize that an internal calibration weight is not needed. And as an added benefit, you could make the weighing section that much smaller.

Let’s go into more detail about this. Many balances have a defined sensitivity drift of ±2ppm/˚C. If the capacity is 100 g and the temperature changes 5 ˚C, using the formula 100 g x 2 ppm/˚C x 5 ˚C, the drift is ±1 mg. For a 100.0000 g display, the change may be in the range of 99.9990 to 100.0010 g. If the weighing value to be confirmed is 1 g, this ±1 mg change drops to 1/100th, or ±0.01 mg. A balance with specifications of 100 g x 0.1 mg cannot detect this level of change. In other words, if the mass to be measured is within 10,000 times of the least significant digit, calibration itself is effectively no longer required, other than during installation. In production lines, if the items weighed in the line are managed separately and the weighing values are checked periodically, there is no need to stop the line for calibration.

The AD-4212B-23 has the SHS mass sensor that was developed for the GX/GF Series of general-purpose balances that followed the two generations of the HX Series. Fig. 1 is a graph of data that shows its 1 μg level of performance. The graph shows the weighing values (zero/span) obtained by raising and lowering a 1.25 g weight automatically over four days from July 8 to 12, 2011, along with the contemporaneous humidity, barometric pressure, and temperature (room temperature/internal temperature of balance).

The repeatability is indicated by the line connecting points (triangles) that represent the reproducibility (σn-1) of ten adjacent span values (capacity minus the zero point). The conclusions gained from the graph are listed below.

1. The repeatability of the span values is 2 μg or less. (The thick red line at the bottom, scale on the right axis)

2. Compared to the span values, which showed little change in the graph, the zero points changed greatly (5 mg or 5000 μg) over the four days in accordance with changes in humidity and air pressure.

3. The stability of the span values can be explained by the small temperature change of about 1.1 ˚C over four days. Furthermore, the temperature of the room was managed by air conditioning, and the area where the weighing instrument was located was separated by a partition, which meant that the air blown from the air conditioner did not influence the results.

4. The cause of the spike in repeatability at 10 AM on July 10th was determined to be the effects of a Magnitude 7.1 earthquake at 9:57 AM off the coast of Minamisanriku.

The results above showed that the SHS developed for general-purpose balances has a sensitivity of 1 μg and a basic performance level with an average repeatability of 1.6 μg. As the developer who proposed the SHS, I must say that I was surprised by the results. The SHS was developed for general-purpose balances, so a sensitivity of 1 μg, 1/1000th of the originally planned minimum weighing value of 1 mg, was amazing.

The data for Fig. 1 was obtained in our balance room. However, other data clearly shows that while repeatability is good at night when no one was moving in and out of the room, during microgram weighing blown air from air conditioning, temperature ripples, wind pressure and vibrations from disturbances caused by people moving in and out, and temperature changes from body temperatures cause instability in the display value and result in weighing errors. Details about the AD-4212B-23 were presented at the Sensing Forum held by the Society of Instrument and Control Engineers in October so I encourage people who are interested in learning more to access the published content on our website.

In the miniaturization field, which is a specialty of Japanese companies, the need for precision weighing is expected to range from research fields to production locations and grow further in the future. As a manufacturer of weighing instruments, A&D promotes the development of simple products that take the location of use into consideration, while at the same time focusing on technical support such as the evaluation of installation environments and working to expand into new markets. By providing a complete solution, we aim to improve our customers’ satisfaction with our products.