Back to Resources

Developing our Next Generation Digital Glass Level Sensor

Stephen Tawn* discusses the development of our Glass Level Sensor to bring it into the digital age.

The progress and thirst for digital electronics has shrunk the market for analogue components resulting in more and more parts becoming obsolete. The effects of this have been felt by many, and this includes our long proven glass level sensor, which has prompted the development of a new approach. There are of course alternative analogue methods with different components, but as that would require a complete redesign it was felt our glass level sensor should join the digital age.

The working principle remains unchanged, measure the resistance of the glass between static pins and calculate the glass depth. Our probe arrangement remains unchanged, there are no moving parts and there is no need for any cooling which keeps maintenance to a minimum.

Early prototype testing of the new concept

The clever part is the use of a high frequency AC current to measure the glass resistance. A normal resistance meter would put DC into the glass which can, in certain circumstances cause the generation of oxygen bubbles. An AC current avoids this but does complicate the electronics and to complicate matters further, consideration must be given to electric melting which not only has relatively high voltages in the glass but may also have thyristor control with high switching frequencies.

To overcome all these factors a vast amount of development was focussed on just the drive current for the measuring circuit. This resulted in a power board that generates a clean sin wave of extremely high frequency while maintaining a constant current.  This current must be maintained within a wide span of load resistance; the load resistance being the resistance of the glass which  will change with glass temperature and glass composition.

The measurement calculation is a comparative measurement. This means the glass level sensor can work on all types of glass and ranges of temperature without the need for ongoing calibration. The stable drive current allows us to simplify the tricky bit of converting the resistance measurements to a glass level. But before the processor can crunch the numbers, further consideration has to be given to possible high voltages in the glass. High being anything above 5V which may damage the processor.

Further PCB development

A means to electrically isolate the probe pins was developed that would still allow the high frequency AC signal to pass through to the processing board where they are converted to a DC voltage and then processed into a digital value.

The advancements in microprocessors have made their implementation both cost effective and, more importantly, easy to apply. Using software rather than vast complicated analogue circuits has allowed us to easily generate the glass level value accurately and more consistently using our own bespoke software to process the data. The measurement values are converted to a ratio. The ratio is then processed to calculate the glass depth.

The glass depth is displayed on an OLED screen on the control box and outputs are available for all standard industrial control equipment including 4 to 20mA, 0 to 20mA, 0 to 10V and 0 to 5V.

The new control box assembly

Software error checking ensures that the system is operating correctly. The processor monitors the drive current to ensure the accuracy of the calculations. Any error states operate a volt free relay for remote monitoring. Errors also make the signal outputs default to maximum (signalling a high glass level) to prevent the furnace from over filling.

The entire development of digitising our glass level sensor has been completed in house including circuit board design, software engineering and manufacture. The only critical part that is outsourced is the PCB printing, manufacturing of which is to our design.

Keeping the process in house has allowed us to retain full control and eliminate the reliance on third parties. This gives us the ability to future proof the product and easily adapt to component availability and design.

Also, the transition to digital will allow us to develop the product further to meet the ever changing and advancing technologies used in furnace control and automation.

ABOUT THE AUTHOR:

Stephen Tawn is Project Engineering Manager at Electroglass Ltd.

22/12/2023

T: +44 (0)1268 565577

F: +44 (0)1268 565594

4 Brunel Road, Manor Trading Estate,
Benfleet, Essex, SS7 4PS, England