Development of the MX Series of Heat Drying Moisture Analyzers

This is the fourth installment in a series of development stories and summarizes A&D’s development of a heat drying moisture analyzer.

A&D has been developing scales and balances since our inception 34 years ago and has been manufacturing weighing equipment for over 30 years. In the beginning, our low name recognition made it difficult to sell products based on the A&D brand and we often entered into OEM supply arrangements with partner companies as a result.

This was also the case for heat drying moisture analyzers where mass measurement technology is used, and we supplied only the mass sensors for a long period. Technically speaking, heat drying moisture analyzers are composed of a heater and a highly sensitive embedded balance as its mass sensor. This combination of a highly heat-susceptible balance and a heater that reaches a temperature of 800 °C presents many design issues.

Addressing these issues to suppress the temperature conditions that are so harsh for the balance requires a complete design solution. Consequently, we were concerned for many years that supplying the mass sensor elements alone limited the completeness of a product. As a result, about 8 years ago, we decided to take up the challenge of creating a product using 100% A&D technology that produced the ultimate level of performance in the environments where moisture analyzers are used.

When we were planning development, there was a manufacturer with overwhelming share and name recognition in the heat drying moisture analyzer market in Japan. This caused some within A&D to say that market reception and sales would likely be poor if we released a product under the A&D brand, no matter how distinctive it was.

Nevertheless, we held the opinion that the market always recognizes sophisticated, high performance products available at a low price and moved forward on the difficult road of product development. This included establishing our own heating technology, since our department did not possess any.

The first obstacle was the cost of the halogen heater. While there were other heating methods available, including infrared lamps and sheath heaters, we decided it was best to use a halogen heater since we felt that the demand for short measurement times would grow stronger in the future.

There were several leading products in the market that used horseshoe-shaped halogen heaters around the outside of the pan to ensure even temperatures over the surface of pan, but we found these horseshoe-shaped heaters to be extremely expensive. Alternatively, when low cost, straight halogen heaters were used, several heaters had to be lined up to heat the pan surface evenly. On top of this, no which method we used, we found it hard to heat the pan surface evenly and we were unable to eliminate uneven heating. We were also concerned that these complex constructions would drive production costs too high.

Another problem with these existing designs was that the halogen heater and the heated sample share the same space. If volatile portions soil the halogen heater, the halogen cycle might break and cause the heater to burn out or the level of heat generated might drop. Obviously, we were concerned about the problems this would create for users.

Right from the start of development, we thought that somehow a single straight halogen heater could heat the pan evenly. Looking back now, this was a quite bold proposition. At first, we considered using a reflector to reflect the light and devised various ideas for the material, surface reflection, angle, size, and shape of the reflector. Still, we were not able to prevent the single straight burn line that was left on the sample on the pan.

After reviewing the reasons why the pan surface was not heated evenly, we thought that the problem might be that the light (heat) of the heater was hitting the sample directly. Instead of using a reflector, we decided to place glass in between the heater and sample. The glass receives the light, increases in temperature, and gives off secondary radiation. Accordingly, we later named this method Secondary Radiation Assist (SRA).

Since the heat-resistant glass is under the halogen heater, it receives the light before the sample. The glass is thermally conductive so its surface heats up evenly and its secondary radiation spreads uniformly over the entire pan. While making several prototypes, we performed tests with corn grits on the pan. When the corn grits were baked evenly and the color of the entire surface changed, we realized that we were on our way to developing a new type of moisture analyzer.

Heat drying moisture analyzers produce volatile components during heating and are therefore the measurement equipment that is most susceptible to soiling. A by-product of using the SRA is that contamination sticks to the glass of the SRA instead of the surface of the heater. Since the easy-to-clean flat glass of the SRA prevents contamination, lamp exchanges are greatly reduced, resulting in reduced management costs and maintenance time.

The unit also has seven layers of insulation to protect the heat-susceptible balance from the 200 °C temperature of the pan. We were able to perform temperature correction in the optimal locations of the balance and achieve a final sensitivity of 0.001% (10 ppm). Our top-class MS70 unit has a sensitivity of 10 ppm and can be used to manage the moisture of plastic material (resin pellets) before the injection molding, which many considered impossible with heat drying methods. We consider the performance improvements of measurement equipment that resulted from the development of the MS70 to be a good example of pioneering and shaping new markets from new measurement possibilities.

We also developed Rs-Temp, software designed for product usability in basic operations and beyond. It automatically finds the optimal heating temperature for unknown samples in 30 minutes and has excellent graphical functions to visualize moisture rates in real time. This software is included as standard on a CD-R. Also included as standard is sodium tartrate dihydrate, which is provided as a reference standard for moisture rates due to the crystallization water in its molecular structure. We believe that the standard inclusion of sodium tartrate dihydrate acts as our guarantee as a manufacturer and shows our accountability for product performance.

As we have proposed and realized the above solutions at a reduced cost, the market has steadily acknowledged our success and made A&D into a recognized brand. As evidence of this, we hold an overwhelming market share in Japan for moisture analyzers with a resolution of 0.01%.

In summary, the development of the MX series of moisture analyzers has shown us that to succeed in existing markets it is crucial to challenge the status quo, pursue the necessary performance standards and underlying needs of products, and propose concrete solutions.

Development of the SV Series of Tuning-fork Vibro Viscometers

A tuning fork vibration viscometer is not the kind of product you hear of very often. This is not surprising since, while there are many viscometers available, A&D is the only company in the world making and selling viscometers based on our unique tuning fork vibration technology. Viscometers in labs and research facilities are typically capillary or rotational types. Several companies make vibratory viscometers for production lines, but these use a rotary reciprocating motion with a high frequency of several kHz. Only A&D makes a model that uses a low frequency reciprocating motion like a tuning fork.

A tuning fork generates sound using the phenomenon of resonance at the same frequency. A tuning fork-type viscometer resonates its sensor plates at a natural frequency like a tuning fork and determines viscosity from the drive force (electromagnetic force) required to maintain constant amplitude. This method gains its high sensitivity from its tuning fork structure.

Technology to resonate an oscillator at the relatively low frequency range of 30 Hz was difficult to develop and had never been used in a product, even more than 50 years after the theory was developed. About 20 years ago, a cement company proposed a viscometer using this method and then about 15 years ago joined with A&D. By that time, however, the person responsible for the development was no longer with the company and work on the technology had stalled.

At that time, the product was still in its early stages so its viscosity measurement range was narrow and its cost was high. As a result, only about 10 units were sold a year. Later, another section within the company spent about 5 years trying to redevelop the technology but failed to complete a finished product. Then one day, the company president came and asked me to complete the vibration-type viscometer and so our section took up development of the viscometer.

Over five years of development had already been done in-house and the product had reached the pre-production stage. However, the development team that had worked on the project up to that point had broken up before the product was completed. It was surmised that pressure of product development might have been too mentally stressful for the development team.

When we first started development, things were a bit like a scavenger hunt. In the remains of the troubled project, we found a few completed items, like die assemblies. In the end though, most of the non-structural, technical elements had to be newly developed. Thankfully, our department was able to apply the technical expertise it had gained from developing digital balances. We were able to adapt the fulcrum used with digital balances and reused balance technology for the voice coil elements of the electromagnetic drive member. One particularly difficult area was the electronics. Even here, we were able to adapt the high resolution A/D converter technology of our electronic balances to achieve a high level of sensitivity. In the mass production stage, we struggled with issues such as getting a resonance point of 30±0.02 Hz, but ultimately we leveraged our electronic balance production technology to produce a new viscometer that had the same high accuracy and low cost as our balances.

The target viscosity measurement range was 0.30 to 10,000 mPas, with a minimum and maximum display of 0.01 mPas and 10,000 mPas, respectively. Technically speaking, the resolution was 1/1,000,000. There had never been a viscometer with such a high resolution, nor one that maintained a low viscosity sensitivity of 0.3 mPas and yet whose measurement was easy and stable.

To put a viscosity of 0.3 mPas in more concrete terms, water at 20 °C has a viscosity of approximately 1.0 mPas, meaning a viscosity of 0.3 is about 1/3 the viscosity of water. This is close to the viscosity of acetone, the liquid with the lowest viscosity. This excellent sensitivity made it possible to measure viscosity at a level that had been impossible up until that point.

After we completed the development described above about 6 years ago, we started selling the SV Series of tuning fork　vibro viscometers as a new type of viscometer. There was quite a bit of skepticism toward this new type of viscometer as a general-purpose viscometer and the series did not sell well in the Japanese market early on. However, there were researchers who were unhappy with existing products and they quickly accepted the new viscometer, thanks to its ability to measure a lower level of viscosity and perform consecutive viscosity measurements over temperature changes and state changes from liquid to solid, which other viscometers had been unable to do up until that time.

The series was also well received in the European and other foreign markets. A manufacturer famous for measurement devices for particle size distribution found them excellent as a viscosity measurement method for base materials that determine Brownian motion, and SV was quickly set up as an optional device. As a result, the SV Series gradually gained acceptance as a viscometer. In the last few years, the SV Series has been recognized as part of the viscometer market, which is attested by the fact that, for example, the series is a recommended measurement instrument for oil viscosity at a leading oil company in Japan.

Our tuning fork vibro viscometer is already listed as a measurement instrument compliant with Japan Calibration Service System (JCSS) viscosity standards and the process for Japanese Industrial Standards (JIS) standardization is underway. Furthermore, we have proposed that the physical value measured by vibratory viscometers (viscosity x density) be described as “static viscosity” to distinguish it as a new concept separate from the established values of kinetic viscosity and viscosity.

A&D has considerable experience pioneering and expanding Japanese technology overseas. Considering that this tuning fork vibro viscometer was developed from unique technology developed in Japan and that the vibration method can measure a new physical quantity (static viscosity), we believe it best to quickly establish standards in Japan first. After that, we would like to proactively promote this new method to the rest of the world and let people know about the measurement abilities of our tuning-fork vibro viscometer.

Development of Mass Sensors for Precision Balances

In 2000, we started selling the GX / GF Series precision balances. Thankfully, the series has been well received by users and continues to sell today even 10 years after its debut. Three years ago, we developed and released the FZ-i / FX-i Series as a line of more economical precision balances. The GX / GF Series and the FZ-i / FX-i Series incorporate mass sensors, the Super Hybrid Sensor (SHS) and the Compact Super Hybrid Sensor (C-SHS) respectively. This time, I would like to talk about the background market demands, development processes, and technical solutions concerning these sensors.

In the past, the leading challenge in developing balances was the ability to measure a minute weight value in a stable manner, and it was recognized that balances required a long time to display a stable weight value. Nowadays, however, manufacturing is highly sophisticated, and it is increasingly necessary to achieve both high-level quality control and high productivity, two seemingly irreconcilable propositions.

It took us around two years to develop a mass sensor that could meet such demands, which became the SHS. When we were developing the SHS, it was generally believed that balances could not be used in production lines. This was because precision balances (e.g. 3 kg capacity × 10 mg readability) of the time would take 2 to 3 seconds before the weighing value became stable, but most production line systems required an operation cycle of one second or less. Meanwhile, it was said, and still is even today, that mass measurement is the most sensible way to judge and control quality precisely and at a low cost. This is attributable to the fact that mass measurement realizes a level of control that is beyond the capability of a CCD camera. Examples include monitoring internal defects of tablets or porosities of precision die-casts, or controlling minute applications of oil or grease.

The development of the SHS involved a hybridization of a high stiffness spring material used for load cells and a high-resolution electromagnetic component used for electromagnetic balance sensors. There was no precedent for the combination of Roberval-structure spring material and an electromagnetic balance component, and many engineers here at A&D were in fact against the idea. However, if we had gone as far as to integrate everything from the Roberval structure, fulcrum flexures, tension flexures and beams, as was proposed by several manufacturers already, there would have been various obstacles, such as difficulties in production and repair, and limited choices in processing method. Consequently, we would have had to procure expensive materials and then force the extra costs on end-users in the form of more expensive goods.

Hence, we deliberately did not integrate the structural materials that break easily, such as the fulcrum and tension flexures, and left them as independent elements. Moreover, we adopted a unified Roverbal structure, which had a strong history with load cells using strain gauges and was easy to mass-produce, and thereby cut down the production cost as much as possible. By these means, we successfully established the SHS technology as our unique mass sensor. After some improvements that we made later to the SHS, we have now commercialized high-sensitivity balances with (1) a high stabilization speed of 0.5 seconds, (2) a high resolution of 1/1,000,000 (1 kg capacity × 1 mg readability), and (3) a readability of 1 µg. Furthermore, we also developed the C-SHS, whose production cost is even lower, and incorporated it into the FZ-i / FX-i precision balances. With the C-SHS, we were able to reduce the cost of a weighing sensor by three quarters compared with old electromagnetic balance sensors made before 2000 and accentuate user-friendliness in the field, including ease of maintenance.

Apart from the precision balances, the SHS has also been applied to moisture analyzers of the heating and drying method, the AD-4212 Series, which are special weighing instruments to be used in automated machines in production lines, and accuracy testers for pipette volumes. The scope of its applications continues to expand.

These are our basic observations of the development of the SHS. To tell you the truth, however, when we started the development, we were being pressured by the trend toward the sophisticated integration of sensor components as proposed by our competitors. We were also criticized by the sales division, who said that A&D would fail to stay current with the market and technology if it employed a hybrid sensor. We were in a tight spot as engineers.

I remember that we felt we had no way out and were agonizing every day to find solutions. One day, while I was in a bus during a trip to Europe, I came up with the idea of intermediate integration, which marked a clear difference from full integration. Building on this idea, we kept creating further solutions and finally arrived at a new principle, which was the SHS as it is today.

We repeated testing with trial samples for about two years, and we were very happy when we achieved target specifications. Fortunately, the move from the trial samples to the ramp-up of the GX / GF Series as new precision balances went fairly smoothly, as we were able to utilize our existing infrastructure of highly advanced production engineering as well as gather our in-house elemental technologies of hardware and software.

Based on this experience, we again realized that when working on a difficult theme such as developing a new sensor, it is important to face reality and accurately grasp the situation in order to inspire ideas, and then work continuously to realize those ideas.