News Detail back to listing
A Cautionary Tale about Hot Chocolate
- Feb 8, 2018
- Latest News
And always keep a-hold of Nurse - For fear of finding something worse
Hilaire Belloc, Cautionary Tales for Children (1907)
One Friday afternoon, many years ago, I got a request for help from a company manufacturing automatic drink dispensing machines.
The machines were of the kind where stacks of paper cups charged with powdered coffee, tea, soup or chocolate are loaded into a carousel. On command, the carousel moves round, drops an appropriate cup down to a delivery position where hot water is added.
The company had manufactured these machines for some years. However, it was conscious of the need to improve its designs. It had therefore engaged an electronics consultant who proposed design changes: principally replacing the ‘old’ motor and gearbox that drove the carousel with a ‘new’ design.
In the course of time, the ‘new’ product was tested, went into production and a number were sold. However, before long, complaints started coming in about the ‘new’ machine from one particular country where the favourite drink dispensed was neither tea nor coffee, as in the UK, but hot chocolate. For some mysterious reason the ‘new’ machines didn’t seem to like delivering large quantities of hot chocolate.
The electronic consultant was asked to sort out the problem and failed: asserting that there was nothing wrong with his design. The company refused to pay the balance of what they owed him until he had identified and corrected the problem. Heated words were exchanged between the parties, and the consultant took the company to court for failing to pay the remaining few thousand pounds due to him. The company counter-claimed on grounds that the ‘new’ machines were not fit for purpose. As the months passed, the legal costs began to rise and soon dwarfed the original sum of money at issue.
The date for the court hearing was fast approaching and the company realised that they urgently needed the services of an independent electrical expert to identify what might be going wrong, and to write an expert report that could be used in court, if necessary. This is where they thought that I might be able to help them. The deadline for the expert report was not quite the day before yesterday, but the following Monday: for submission to the court later in the same week!
I have carried out many different kinds of electrical failure investigations, but at that time I knew very little about automatic drinks machines. It was therefore with some trepidation that I agreed to visit the factory and examine the questioned machines.
I visited the factory on the Saturday and was first shown the ‘old’ machine design, in which the carousel was powered by a very simple, but substantial, induction motor driven from the mains and fitted with a 150:1 reduction gearbox. This gave the carousel a rotation speed of about 20 RPM even when the machine was heavily loaded with cups. I was then shown the ‘new’ design, which differed little from the ‘old’ except that the carousel was driven by a much smaller geared DC motor fed from an electronic power supply designed by the consultant.
To me it seemed as if the ‘new’ design had added unnecessary complication to a well proven design without any very obvious benefit to the customer. “Why did you change from the ‘old’ to the ‘new’?” I asked. I was told that the consultant had claimed that replacing the ‘old’ motor with the ‘new‘ would reduce energy consumption.
The alleged reason for change mystified me: on the one hand, it was obvious that the bulk of the energy consumed by the machine went into boiling water to fill the cups. On the other hand, very little energy went into driving the carousel, which, on average, only operated every few minutes and then only for a couple of seconds at a time. If only a very small proportion of the energy consumed went into driving the carousel, why go to all the bother and cost of changing the motor?
The usual complaint was that a subject machine - always a ‘new’ machine - would refuse to deliver drinks for no apparent reason. An error message would appear, indicating that the machine had jammed. It then required a service engineer to visit, open up the cabinet, check for blocked cups and reset the mechanism. A day or so later the problem would recur, the service engineer would return and restart the machine. It was puzzling that no cups had ever been found blocking the mechanism. It was a case of No Fault Found: infuriating to the customer and disquieting to the manufacturer, whose ‘old’ machines had a well-deserved reputation for reliability.
There were some clues as to the possible cause: firstly, it had been reported that the heavier the load of cups stored in the machine the more often the machine failed, especially if the proportion of those cups filled with chocolate powder was high; secondly, some customers found that by never loading their machines above half capacity they could avoid the breakdowns.
It was part of the control logic that if the if the carousel took more than 2 seconds to pass a particular point, this was assumed to indicate a paper cup jam, and, as a safety measure, power was cut off from the motor: from then on, nobody got any more drinks out of the machine until it was re-set. Much effort had been devoted within the company to trying to discover why the ‘new’ motor was running so slow that the trip time was exceeded, but to no avail.
During my visit, an ‘old’ machine was tested in my presence. It was loaded with 300 cups of hot chocolate powder - the heaviest possible load – and each time the appropriate button was pressed the carousel rotated at a steady speed, estimated at about 20 RPM, until the stack of cups reached the delivery point and the bottom cup dropped down. However, when the ‘new’ design was tested, the carousel ran well above 20 RPM on no load. If the carousel was about half filled, the speed dropped to about 20 RPM, but the operation became somewhat unreliable. At full load, the carousel turned so slowly that the control circuitry detected it as a jammed cup and tripped the motor.
Coming new to the problem, I reasoned from first principles that if an ’old’ motor and gearbox was replaced by a ‘new’ motor and gearbox roughly the same torque would have to be delivered at 20 RPM in each case. In other words, it should not matter whether the ‘old’ or ‘new’ motor and gearbox arrangement were used because the performance would be roughly the same.
But why were the ‘new’ motor and gearbox not doing their job properly? Somewhat up a gum tree, and not knowing what avenue of enquiry to follow up next, I asked to see the drawings for the ‘old’ and ‘new’ motors to check if they really were equivalent designs.
It turned out that the’ old’ motor and gearbox combination was designed to deliver full torque at 20 RPM, whereas the ‘new’ combination was designed for 55 RPM. In practical terms, too little torque was being delivered to the carousel in the ‘new’ design and at much too high a speed. Had the step down ratio of the gearbox been increased by a factor of 3.75 to 1, the torque would have been multiplied by a factor of 3.75 and would have been correctly delivered at 20 RPM.
Instead of changing the gear ratio – not difficult to do because motor manufacturers usually sell a range of matching reduction gearboxes to suit requirements for different output speeds - the consultant chose to introduce a large series resistance into the DC motor armature circuit in order to reduce the speed of the carousel when loaded. As a result, the carousel ran much too fast on no load, erratically at about 20 rpm at half load, and rarely, if at all, at full load. This explained the failure of the ‘new’ machine to function when heavily loaded with chocolate powder.
In my report, I recommended that the company went back to using the original motor. The company took my advice and the ‘new’ motors were quickly replaced with ‘old’ motors. The case was settled and, as far as I am aware, there were subsequently no more hot chocolate delivery problems.
Nowadays, when faced with this kind of problem I always ask to see the drawings, bills of material and the revision history. Sometimes, as in this case, the information is very revealing.
This was one of those occasions where the company would have been well advised to stick to the original design ‘for fear of finding something worse’.