Annunciation Panel

Introduction

The annunciation boards were designed in 1988 when we started trading as Digital and Software Technology. A general set of boards were designed—processor, inputs, outputs, and sensor interfaces. Annunciation panels were requested by four different industrial sectors.


General Description

The layout was made for 16 inputs then replicated four times for 64 total displays. The bi-colour (green and red) HLMP-2965 LED arrays from Hewlett Packard were roughly 9mm×9mm, with 30mA drive requirements. An application note (I think from National Semiconductor) showed how to drive bi-colour LEDs with a dual 555 timer (LM556) on a 5V supply. The LED arrays had a 2V drop for the red, and 2,2V for the green. The four internal (per colour) LEDs each had a current limiting resistor, so the total drive was probably 120mA, but we designed for less, as the total board would need 64×4×30mA = 7680mA. (With bi-colour, the display is either red or green, not on or off, so the total drive must be considered). The advantage is that if power is lost on a panel, then the absence of alarm lamps does not mean there are no errors. The red and green were easy to see (if you are not colour blind), and if there is no lamp, then there is a problem!

Bi-colour annunciation panel all on red

The above panel picture was from a scan done a long time ago and unfortunately only the red were on (default for fail if not driven). This panel was for spinning machine monitoring and the LED arrays had numbers transferred onto their surface. The output could easily be seen thirty meters away. The panel size was roughly 19” wide to fit into industrial 19” racks. The picture is small to save ink if printed.

Side View

Anunciation panel side view

A side view of the annunciation panel mounted on a white powder coated aluminium sheet.

NC punched front panel

Anunciation panel side view

The front panel was NC punched on an Amarda turret punch. The dimensions were taken from the printed circuit layout, which was designed on an imperial grid (0,1” centers). When the board was made, it was probably pen plotted, as there was a stretch on the Y axis. Unfortunately, with the minimum quantities of both boards and the aluminium sheet, there was not going to be any second run. When the front panels were sprayed black, the offset was not as obvious. The PCBs were also populated before the panels were ready, so we were not going to redo the boards at the display prices.

A quick search for HLMP-2965 arrays in 2011 gave a price of A$3.030550 each for a 320 price break (extended A$1,057.76 on Digi-Key. For the 64 displays, that would be A$211.55 without blank board, drivers, assembly, Australian freight, etc.). Which is why our next version would be based on digital signage .... You’ll have to wait, but large LCD screens are well below A$400.

Printed Circuit Board

Anunciation panel printed circuit board

The boards were made as 16-way ×4 for a total of 64 displays. The wiring here was for our in-house testing and simply wired the boards in parallel for power.


Steam Trap Monitoring

The Spirax-Sarco steam agents received a request from a supplier to one of the tyre plants. We quoted and then heard nothing. Our price was approximately R35,000 (US$63301). That was to monitor 128 steam traps and display the results about 200 meters away in one of the floor supervisors’ office. An excellent idea considering how much steam is used in a tyre plant, and any stuck valve wastes a huge amount of energy.

To monitor each trap requires a sensor, but the condensate is not as conductive as tap water, as it was steam in a previous life. The electronics also have to be rugged—not only because of high temperature and moisture, but people tasked with looking after pipes and fittings don’t have silk smooth hands—more like trolls from the “Lord of the Rings” movie. They prefer an axe to neurosurgical instruments. The individual valves would have sensors to monitor water in the traps, then convert to a digital signal and take it to a remote panel with 24V input cards.

The local processor would send these via RS-422 over a serial port to a similar processor card in the supervisor’s office which would drive two of the 64-way bi-colour panels, as well as interface to a PC for data logging.

Long story, but a while later we bumped into the engineer at the tyre plant, and they said the quote over R120,000 was a bit high! No bull, Shurlock! I felt like what the Chinese must feel like when they are squeezed for price only to find out that the end-user paid at a 400% markup.

Spinning Bobbin Monitoring

A company asked us to design a system to monitor digital outputs to see when a nylon thread had broken from a bobbin. The nylon was threaded through a spring-loaded wire loop, and when the thread broke or ran out, the “trip wire” would trigger. For an operator looking head-long into hundreds of threads, it took a while to detect and repair the break. The system was driven by our 68000 boards runnning a 10ms tick real-time kernel. As far as we were aware, there was no failure in the ten or more years we were peddling custom design.

Hybrid Energy

We designed a system for Hybrid Energy Research for the Port Elizabeth Technikon before 1990. There were digital outputs to sequence relays for different power loading on battery banks, wind turbines and solar panels. The electronics was housed in a 40 foot shipping container converted into a field laboratory. The annunciation panel gave a quick overview of any alarms, which output were being driven and any state-machine status.

Tannery Automation

Before the PWP Tannery automation, several projects were done at two other tanneries. The annunciation panel was installed in the supervisor’s office. The switch settings of hard-wired pushbuttons and multiway switches were monitored to see when the drums were run forward, backwards, stood idle, were filled with water, etc. Data logging allowed the day- and night shifts to be compared for any problems. The day and night temperatures can differ by up to 20 °Celcius, so process variations are highly likely, and much above 40 °Celcius, the collagen is likely to break down, ruining the skin.


1Historical exchange rate for 1988 was ZAR 5.53:US$ 1