Contact us


Engineer Blog

The Best Solutions are the Simplest

Kikusui’s Smart Rack series: adding unexpected value

My name is Shiro Arakaki and I work in Kikusui’s Solution Development Division. In this column I will share some of my experiences with product design. But first, a quiz: what do Rose’s Lime Juice, stir-fry seasoning and 2-in-1 shampoo have in common?

I don’t need to tell you that Rose’s Lime Juice is a drink, stir-fry seasoning is a food, and 2-in-1 shampoo is a hair product. You have probably consumed or used all of these items at least once. If you were about to say they are all mixtures, you would be close. Indeed, Rose’s Lime Juice is a mixture of flavored syrup and water, stir-fry seasoning is a blend of several different Chinese seasonings, and 2-in-1 shampoo is a mixture of shampoo and conditioner. That is not the answer I was looking for, however. The answer is: all three save the consumer or user the effort that would otherwise need to be expended to achieve the desired state.

Effort, Work and Added Value

Life is full of hassles and time-consuming situations. You can find countless examples, not only in your personal life but also at work. However, I want to examine the idea that dealing with hassles are all part of performing one’s job. For example, the value of tasks that are literally performed by hand, such as therapeutic procedures performed on the body and the creation of artisanal handcrafts, would appear to lie in the fact that those performing them expend time, that is, in the process. If these tasks are performed in too short a time, we value them less. However, with products whose value is based solely on the end product, that is, the outcome of a process, it cannot be said that value is proportional to time expended. I feel that this is especially true with the industrial goods with which I work. This idea may be easier to relate to if described in terms of agricultural produce or food: the cultivation, ripening or fermentation of such produce takes time. However, there is generally no need to leave mechanical devices to mature in the same way─we are expected to build the devices as quickly as possible. (To be precise, in some cases we do perform continuous-current testing, or ageing.)

Anyone who has experience in manufacturing knows all too well that building a device to standard is a battle against time. It is all about following a defined process, quickly and precisely. With such devices, emotional components, such as whether the engineer built the device with “tender loving care”, are not seen as adding value. In fact, it doesn’t matter whether the device was casually slapped together by some random tearaway─as long as the outcome conforms to specification, the device “passes”. I think Japanese consumers in particular want to be able to appreciate the dedication and effort of the craftsperson in the finished product, perhaps because of the Japanese belief that there are gods and goddesses dwelling in every single object. Unfortunately, however, there is no guarantee that the person who receives the outcome of that labor will appreciate what has gone into making it. That may seem harsh, but that’s life.

In the course of your daily work you sometimes encounter tasks that, while clearly time-consuming, do not translate into increased product value. That said, it is not acceptable to perform such tasks sloppily just because they are time-consuming. A hassle indeed. The three consumer products I listed at the beginning of this article are nothing special at first glance. Some will look at these products and say cheekily, “why not just mix them yourself”, or complain about having to pay someone extra just to blend the ingredients together. On closer inspection, however, we see that there is more to it, because the process of dilution and blending require considerable expertise. This is a classic example of added value that is not readily appreciated.

The Saving of Labor is in Itself a Solution

This brings me to the point of this article. Kikusui has for some time produced bespoke rack assemblies containing power supplies and other instruments. To create these assemblies, we mount the requisite devices on a rack and connect them together. In recent years, however, we have been delivering an increasing number of what are called “Smart Racks”, built to a set specification. (Figure 1) I was responsible for engineering the PBZ SR series of bipolar power supplies and the PCZ-A SR series of AC electronic loads, both of which are examples of such Smart Racks. Cynics will say that I merely stacked the devices on a rack and wired them up.

Figure 1. A selection of Smart Racks

Anyone who has actually built such a rack, however, knows that it is no simple task to mount all the devices and wire up their primary and secondary terminals in a way that enables you to guarantee that the stacked devices will function correctly as a single unit, that is, to specification. It requires extensive knowledge (ironically, the more experienced the engineer performing the task the less likely they are to think of it as knowledge): all too often you wire the devices up only to find that they do not work correctly. This is another example of value that is not readily appreciated, if you ask me.

This was particularly true with the PCZ-A SR series. Not merely a stack of electronic loads connected in parallel to increase capacity, the devices in the PCA-A SR series enable load units to be combined in the desired permutation in accordance with the number of phases and configuration of the AC inputs. This is a job the customer could do themselves if they wanted to, although make no mistake, it is a hassle. But if you have plenty of time on your hands and enjoy this extra labor, you will find Smart Racks to be overpriced, if not superfluous.

Generally speaking, however, clients want to get this reconfiguration step over and done with as quickly as possible because the switchover process itself does not generate any value. And that is the value of the Smart Rack. You are paying for time savings and the knowledge that the devices have been connected correctly (if I may say that without sounding boastful). I believe the Smart Rack has every right to be termed a “solution”.

I will now describe my experience developing the PCZ-A SR series. I hope my comments will give you some idea of the thinking and process that led me to this product and of the added value that you do not read about in the product specification. While I don’t like singing my own praises, I hope this account of the development process will be useful.

The Smart Rack Series

While the capacity of many Kikusui power supplies and electronic loads can be boosted by connecting them together in parallel, Kikusui did not initially ship any standard models parallel-connection ready, instead building customized racks on a one-off basis at the request of our customers. Five years ago, however (in 2012/13), we planned and developed our first ever “Smart Rack” versions of the PBZ series bipolar power supply and the PLZ-4W series DC electronic load. I was in charge of building the PZ Smart Rack, although in this article I will describe the process of developing an AC electronic load version (the PCZ-A SR series, which comprises several PCZ-A SR series electronic loads connected in parallel) the following year.

The proposal to release a PCZ-A SR version of the Smart Rack was made in April 2013, just as I was congratulating myself on having completed the development of the PBZ Smart Rack. As is always the way, I was summoned to a meeting room without warning and nominated as project designer. The brief was:

  • That year, Kikusui would develop high-capacity versions of its AC electronic loads.

  • The new loads would basically comprise multiple PCZ1000A’s connected in parallel and allow easy switching between single-phase, single-phase two-wire and three-wire, three-phase three-wire, and three-phase four-wire configurations.

  • The project obviously needed to be low cost.

  • The load would be available in 3kW, 6kW, 9kW and 12kW versions. (The 12kW model was subsequently scrapped.)

  • The device needed to be ready by the end of the year.

After this very basic briefing, I got to work on the specifications straight away.

The PCZ1000 AC

The PCZ1000, which is the basic component in all of the products in the PCZ-A SR series, is a 1 kW AC electronic load that can be operated in constant current, constant power and constant resistance modes and was released in 2000. While a 2007 upgrade saw the PCZ1000 getting extra features (including the ability to be connected in parallel to up to four other units) and being rebranded as the “PCZ1000A”, the fundamental design was retained, so if we’re being honest, it is not all that new. (Figure 2) However, when one considers that the PCZ1000 continues to be manufactured and sold up until this day, you can say that the design has been refined exhaustively. In other words, this is a very reliable unit that you can use with confidence.

Figure 2. The PCZ1000A

Not surprisingly, the main attraction of the PCR-A SR series that I would go onto develop would be the ability to switch between input configurations. I considered what functions we should include.

If the 6kW model was to switch between single-phase, single-phase two-wire and three-wire, three-phase three-wire, and three-phase four-wire configurations, it would need to use all six PCZs to create one single phase two wire output, two sets of three PCZs to yield two single-phase three-wire outputs, three pairs of PCZs to create a three-wire three-phase output in delta configuration, and three pairs of PCZs to create a four-wire three-phase output in a Y configuration. The device also needed to be able to perform switching of the signals transmitted via the dedicated 25-pin parallel connector.

To enable convenient switching between these different configurations, I was tempted to add a selector switch so that the user could switch between single phase two wire, single phase three wire, three-phase three-wire and three-phase four-wire outputs, simply by flipping a switch on the front panel. However, this would require the switching of a large number of connectors using relays, which would require their own control circuit and power supply. While not impossible, this approach would make the device extremely costly.

In situations like these, I find myself asking “what are we actually trying to achieve here?”
Just how many of our clients actually needed this kind of switching function? Rather than developing a model with a complex switching function, wouldn’t it be easier to release a range of dedicated models, one for each of the respective configurations?

The received wisdom is that you should perform a proper market survey and determine whether there is a business case before developing any new devices. However, surveying the market for such niche products is inherently difficult, and even if market research does yield some kind of outcome, it is ultimately a case of, “if there’s no return, we won’t do it.” I believe that when dealing with small markets of this type, the more you research the market, the more you come to the conclusion that you’re better off not attempting anything.

At the same time, you need to consider the gut feeling that salespeople and developers get when interacting with clients. Sometimes salespeople and developers feel that there is demand for a product and that it would be well received, even if clients don’t come out and say it, and that blossoms into a new product. It all comes down to intuition. Strictly speaking, you are supposed to look at actual demand figures before moving to the development phase, but it also seems a little boring to change your stance on the basis of figures. Getting a perfect quiz score is no fun when you’re answering questions you already know.

The market for measuring equipment and power supplies is extremely small compared to the consumer sundries market. Perhaps measuring equipment is a bad market to choose if you’re trying to turn a profit. In fact, I doubt that a shrewd businessperson would get into this market. Does that make Kikusui stupid? I actually think that this attitude is what makes Kikusui the company it is and what makes this job so interesting.

But I digress…
As you can see, my thinking was quite disorganized, but I concluded that as we had decided to build a switchable device at low cost, I should abandon the costly automated switching approach, and pursue manual switching instead. I would consider the future of the product after monitoring its progress after release.

Designing the PCZ-A SR

With the specification defined, it was time to start the design process. First, however, I needed to consider the following issues:
(1) Requirements when connecting nine units in parallel rather than five
- changes to the driver circuit
- suppression of oscillation
- changes to firmware
(2) A build that would allow clients to connect the units themselves in a failsafe manner
(3) A Smart Rack capable of holding nine units
(4) Other issues

I immediately set about testing the requirements listed in (1). I planned to round up a bunch of PCZ1000As and perform tests to see whether more than five could be connected in parallel, but as I couldn’t get my hands on that many PCZ1000As, I had no choice but to simply source 10 interface circuits and determine whether communications worked correctly. The comms test worked fine with 10 units, and the results suggested that even 20 would be no problem. I had thought this part would be a challenge, but it turned out to be quite anticlimactic! Happily, I was able to connect the devices in parallel without any hardware modifications. I had overcome the first hurdle.

Next, I needed to make the firmware work with up to nine units connected. I contacted our software developers and asked them to make the requisite change. While I thought this would be time-consuming, the modification only involved editing the value for the number of units that could be connected in parallel and was finished in no time.

I would not be able to assess oscillation suppression until the unit was built, so I skipped this step and moved on to a review of the rack. Rather than mounting the units on a standard 19-inch rack, I decided to go for a more compact and elegant rack, as I did with the high-capacity PBZs I developed the previous year.

I also considered the second requirement, that clients would be able to connect the units themselves in a fail-safe manner, to be crucial. I came up with a device (which I will refer to here as a “switching bar”) that allowed the user to switch between single phase two wire, single phase three wire, three-phase three wire, and three-phase four wire configurations simply by changing the position of the shorting bar placed across the terminals. After discussing my idea with the rack supplier, I decided to build the 6kW model first. This took a considerable length of time, and I had to consult with the rack supplier multiple times before I arrived at a setup that would ensure a fail-safe connection.

Because the cables would be connected by the customer, I added color-coded, silk screened markings and applied a sticker with information on wiring configurations understandable at a glance to ensure that the connections would be performed correctly. I also added markings that showed clearly how the switching bar (attachment) should be positioned in order to yield the desired configuration. (Figure 3) To facilitate assembly, wiring, and maintenance, I positioned the switching bar, where many wires congregate, behind a door for easy access.

After several rebuilds in order to overcome various issues, the switching bar was finally ready. Next, after ordering racks for the 3kW and 9kW models from the rack supplier, I began electrical testing using the case for the 6kW model.

Figure 3. Input terminals

Firing Up the System

With the build complete, it was finally time to fire up the system, that is, perform electrical testing. This is a both an enjoyable and nerve-wracking experience, because a system that is fine on paper might not work, or, even worse, break down. Experience with this kind of testing is also a source of expertise (another form of value that is not readily appreciated). I was most concerned about whether the device would still perform normally with more units connected in parallel, or whether the extra units would create oscillation or other malfunctions. I was quite nervous.

My fears proved unfounded as the device worked without issue and was extremely stable, perhaps because I was testing it at the low frequencies of 50 and 60Hz. All that remained now was to methodically collect the necessary data. The fact that the device not only operated correctly under initial testing, but also performed without problems after subsequent modification and adjustments, brought home to me just how well-designed the PCZ1000 is. I am grateful to the employee who designed the PCZ1000.

Next, I finished the design of the 3kW and 9kW models. Last but not least, I had our manual vendor put together some easy-to-follow instructions.

I started developing the PCZ-A SR in April 2013 and shipped my first bespoke 6kW model nine months later in January 2014. While that is longer than the four or five months it usually takes to build a bespoke project from scratch, I was working on other projects while designing and evaluating the PCZ-A SR, so it was quite a challenge.

A Nifty Product

The need to loosen and tighten screws when switching between modes means that operating the PC2-A SR still requires some effort. However, the switching bar makes it possible for anyone to switch between configurations correctly and without having to think, which is a very convenient feature if I do say so myself. (Figure 4) My design was praised by the heads of other divisions and I have been informed that it is rated highly by clients as well.

Figure 4. The switching bar (attachment)

Just between you and me we have shipped around twenty PCZ-A SR series electronic loads as of the time of writing (July 2017), and still receive a large volume of enquiries about this line. I am grateful that it has been so well received. While hi-tech devices that are packed with the latest technology are obviously most deserving of the name “solution”, I believe that products like the Smart Rack, which, while nothing special at first glance, are appreciated more over time, definitely belong in this category too.

When developing this system, I initially wondered who would actually use what struck me as an obscure device. But looking back now, I am pleased to have been involved in the project and am quite attached to the PCZ-A SR. AC electronic loads are essential when testing the power conditioners that are in so much demand right now. As the developer of the PCZ-A SR, I invite you to try out this system.

Shiro Arakaki
Acting head of System Technology Section, Solutions Development Division

[Major achievements in product development]
PBZ-SR series
PBX-BP series
PCZ-SR series
Surge testers/power fluctuation testers
Bespoke testing systems

Contact us