Thursday, April 2, 2009

My Industrial Training Experience



2.1 Standalone controllers

In my training period at dataad I was vastly involved with Industrial standalone controllers. In fact the first week of training was completely dedicated to studying the operation and troubleshooting techniques. I was also given comprehensive training of how to use a digital multimeter for this purpose.

2.1.1 Introduction

The industrial controllers are vastly expensive but reliable when compared to domestic controllers. They form the core of any industrial system by measuring, displaying and retransmitting process values.

The most frequently used controllers are temperature indicators where input is a thermocouple, weight indicators where inputs are load cells, Energy indicators where inputs are current and voltage etc… In effect there exists a controller for every conceivable process variable. Some are not just passive but also active. E.g.: PID indicators where PID control, Relay output (on/off) can be programmed into it while some modules support RS485/modbus, RS232, and Ethernet or 4-20mA communication.

In srilanka the companies which deal in these controllers are a very few due to the specialized market segment it caters for. The well known brands are Schneider electric, Gefram, cutes etc… My training company is the registered dealer for many such world renowned brands. They sell, install and give customer support for these products.

The irony is given the high cost of these well known brands the majority go for the low end, cheap SELEC brand products which are of Indian origin (but coming with a one year comprehensive warranty). The sad fact is that the failure rate of selec products is very high.

2.1.2 Display only

Here the input to the unit is a thermocouple or RTD sensor and usually the type can be set within the controller menus.

Fig.2.1

In the training period I was entrusted with the task of providing technical support for these modules (figure 2.1).

The display units, Resolution and format all can be changed within the menu.

2.1.3 Display and relay output

Here the only difference is the addition of output relays. The outputs can be SSR or open collector.

Fig. 2.2

The second row in the display gives the set point that we have programmed through the menu.

One of the most important things that I learnt was how to interface the controller to the control systems using the foresaid two methods.

In event of SSR relays (this is more expensive) we can directly connect the relay (figure 2.1) to the outputs and get it energized.

Figure 2.1

230V

S

E

L

E

C

+

_

Relay Out

Relay

~

Load

In the event of Open collector output internal output comes from the transistor.

Output


Figure 2.2

The other important controllers I used were Timers and Counters. They provide the important function of timing or counting digital pulses.

E.g. To count the number of paper batches folded by paper processing machine.

Figure 2.2

S

E

L

E

C

+

_

Relay Out

Relay

~

Load

24V DC

S

E

L

E

C

24V DC


Figure 2.3

2.1.4 Display and PID control

In order to implement a precision set point we need to set up a complex feedback control algorithm such as PID. The modules that I worked with were PID500 series selec.

The word PID is usually misinterpreted by the technicians with respect to this module. The output is digital not analog. What is done here is that the digital Relay output is Pulse Width Modulated by using the analog PID output thereby precisely controlling the set point. I faced some practical problems from which I understood the importance of PID control in a control system. That is described in the PLC section.

2.1.5 Retransmission and Communication

The aforementioned modules do not have a method to integrate into a larger automation system. But when such integration is necessary we have to go for modules with retransmission and communication functionality.

2.1.5.1 Analog retransmission

Here the output is analog. It can be a voltage of 0-5V, 0-10V or 0-20V or a current of 0-20mA, 4-20mA.These outputs act as inputs to PLCs’.

2.1.5.2 Digital retransmission

Here the output is a digital signal. Generally the physical layer is RS485 and the application layer is modbus.

2.1.6 My experience

Major portion of the workload constituted of testing and troubleshooting returned modules and providing customer support. Apart from that I had to visit a total of three sites for installations.

2.1.6.1 Ayurveda research center

Boilers where we installed the temperature indicator modules along with k type thermocouples

Fig. 2.4

Here we had to check the installed PIC101 modules and refit it as the electrician had wired it incorrectly. They were lucky that no module was damaged. There were 6 modules connected to k-type thermocouples. The thermocouples were submerged in the boilers to measure the boiler temperature.

The environment was extremely hot and humid and we were in great discomfort. As it was a state institution the employees were not helpful.

Afterwards we had a discussion with the managing director regarding how we could help them increase their efficiency using automation. One thing I learned here was that the labor resistance to whatever automation measure is much more significant than the technical problems.

2.1.6.2 Paper processing plant

This plant processed raw paper into printed processed paper. We had to troubleshoot a machine which used a SELEC counter to count the number of papers folded. The sensor was a simple reflective proximity sensor. The technicians had incorrectly installed it. So we had to give the correct power supply and also set up the sensor. We had to improvise in that occasion to mount the reflector in such a way as to give a reading of one pulse per each revolution of revolving bar.

2.1.6.3 Energy meter installation at polythene packaging plant in wattala

The energy meter that we installed in the electric panel

Fig.2.5

This was my first experience with the working of an industrial electrical panel. Here the factory management wanted to reduce their electricity bill. So they wanted to measure their KWh and KVAR usage.

We had to disconnect the power to the whole plant before working on it. As there was a voltmeter and ammeter installed earlier (with CTs’) we had to take the input from them and wire it to the SELEC energy meter.

After installation we had to explain how to measure the energy unit from the meter to the manager. The method to measure VAR value was very difficult to explain to the factory manager as he didn’t have any knowledge about electrical engineering.

2.2 Industrial sensors

There is a big market for industrial sensors and for some of those sensors dataad is the sole dealer. Unlike domestic sensors these need to be reliable and accurate.

2.2.1 Linear transducers

Fig.2.6

They are used to measure linear distance. This is based on the principle of potential difference through a resistor network.

E.g.: input voltage 24V, then output varies from 0-24V proportional to the position of the transducer head.

On several occasions I had to show customers how to interface linear transducers to an automation system.

2.2.2 Proximity sensors

Proximity sensors have a digital output (contrary to what the name suggests) There are different types used in different occasions. These are usually involved in counting or timing events.

E.g. counting number of paper sheets, number of metal tins, Timing 10 seconds after metal object is discovered and then turning on the conveyor belt.

2.2.2.1 Inductive type

This is only used to detect metal objects.

Fig. 2.7

Metal object

Sensor

Conveyor belt

2.2.2.2 Capacitive type

These can only detect non-metallic objects; the sensing distance is also less.

Fig. 2.8

2.2.2.3 Reflective Type

This has the highest sensing distance though the price is pretty high.

Fig.2.9

Soap

sensor

Reflector

Moving conveyor belt

We had to install this sensor at the previously mentioned paper processing plant at wattala.

2.2.3 Temperature sensors

There are basically two types.

2.2.3.1 Resistance type (RTD)

The operating principle is the change of resistance of material with temperature.

E.g. PT100 – platinum 100ohms resistance at 25 degrees.

This is used when accuracy is the main requirement. Main disadvantages are slow response time and small measuring range (~125 C).

Fig. 2.10

2.2.3.2 Thermocouple

Thermocouples are the most widely used temperature sensors in the industry. Low cost paired with good accuracy, fast response time and durability are some of its features.

K-type is the most widely used. It is constructed out of a Ni Chromel junction. It usually has a range from 0-1000 degrees with precision of 0.25C and accuracy of 4C. I learnt the working principle as well as how to interface the sensor to an indicator module.

2.2.4 Humidity sensors

Fig. 2.11

Humidity sensors are usually installed in cold rooms where we need to maintain an ambient humidity level. Usually the output is a loop current of 4-20mA. I had to give customer support/sales assistance to find the suitable pressure sensors matching the customer requirements.

2.2.5 Weight transducers

These transducers are widely known as load cells. Its principle of operation is the strain gauge action. The unbalanced output is amplified by using a bridge circuit. Output is usually a current. The load cells are generally accurate for 1% FS.

Fig. 2.12

2.2.6 Pressure sensors

It has the same operating principle of the strain gauge. Its output is proportional to the pressure acting on its diaphragm. My experience was limited to helping the customers find a sensor which satisfies their requirements.

2.2.7 Encoders

Encoders are special sensors used to measure rotational speed. Generally it gives a digital pulse output per each 1 degree of rotation. I had to test as well as interface encoders to SELEC modules.

2.2.8 Water level sensor

Another interesting new sensor that I worked with was the water level sensor. This was brought down from china to test its marketability. Apart from having 4-20mA sensor output this had a RS 485 output which is very interesting as no sensor that cheap had this functionality. The range was 2m while accuracy was 2cm. I was given the task of configuring and setting up the prototype. I interfaced the 4-20mA output with a PIC 101 module and varied the submerged length of the sensor. It gave a very accurate reading. But I could not get data using the digital output due to not having enough information about the application layer of the communication protocol. Though they sent some software in the CD it didn’t work and multiple requests for help through email went unanswered. If there was a good manual which described the application layer, we could have written a program to sieve the data. But due to the language being Chinese we were at a loss. The product was low cost but very good quality, the only weakness being the virtual non existence of user support in English.

Fig. 2.13

PIC101

Water level

Sensor

2.2.9 My Experience

I got the chance to interface all the available sensors to controllers and PC. A memorable experience was helping out with the load cell application. The application was a precision tea weighing and packing facility. The managers wanted to automate this process which is usually done manually.

The load cells used for weighing needed to be accurate to 10g (which they were) but the customers were complaining that it showed a deviation of over 200g.

Given that the company was new to load cell technology the engineer found it difficult to handle. Therefore I had to do the necessary calculations and provide the required customer support. I found that the problem was not with the load cell but with the indicator and retransmission module. So I had to calculate the minimum possible error and give my recommendations to the General Manager after much calculations and practical testing.

The used Gefran 40B96 indicators were then recommended to be replaced with more precise MP30 indicators. This solved the problem. This experience gave me an insight into conflict management (as the customers were very angry and rude).

2.3 PC based industrial automation

My company was one of a handful of companies which installs PC based automation equipment in srilanka. They are the authorized dealers for the low priced Advantech Brand which is of Taiwanese origin.

PC based industrial automation is the ultimate tool in industrial automation. It can service any system with any number of IO and implement any control algorithm. I was able to get a comprehensive understanding about all aspects of this broad subject.

2.3.1 Hardware

Fig. 2.14

The Hardware used was ADAM modules (figure 2.14) and industrial PCs’. The signal conditioning and transmitting modules were daisy chained together (power supply 24V DC) and connected to the PC using a RS232 to RS485 converter module. The general application layer protocol was the proprietary Advantech protocol.(some modules supported the open MODBUS protocol but they were expensive)

Fig. 2.15

2.3.2 Software

Adamview is the software used to do the control algorithm programming. It has inbuilt control blocks as well as a dedicated scripting language called basic script. Initially I designed a program to interface a heater to control a small heating application using Adamview as an exercise to get used to the PC based automation design.

My project work included programming Control algorithms using ADAMview such as in the above figure. My contribution to the company in this area was the introduction of the usage of ADAMview solely as the Datacenter. I found the method to interface Adamview to external programming tools such as visual C and Basic. Initially I demonstrated prototypes using VBA in excel and then in VB mobile 5.0. Figure 2.16 shows the basic interface that I worked with. I also created a small demonstration video to help beginners. It can be viewed at http://www.hasala.tk .

HMI (Human Machine Interface)

Basic script Block

Fig. 2.16

Fig. 2.17

Visual C/Basic, Delphi, excel/access VBA (Control Algorithm programming tool)

Data from RS 485 network (Adamview/Labview/OPC server)

Data Center

OLE automation/DDE interface

IO Driver (Adamview/Labview/OPC server)

Figure 2.17 gives the data flow of my method (using ADAMview datacenter)

2.3.2.1 Moulding machine HMI

Industrial PC with touchscreen acting as the HMI

Moulding Machine

Fig. 2.18

Having been shown the SCADA developed for the moulding machine I was dissatisfied with the design of the SCADA. The limitations of ADAMview as a SCADA development software was clearly evident. Therefore I read the Adam manuals and understood the theory behind the software and through intense research finally got the working code for a system with ADAMview only acting as a datacenter. My test control algorithm and SCADA development platform was Visual Basic mobile 5.0.

Then I demonstrated the capabilities of this powerful method by developing a test GUI for the moulding machine. While doing this I had to study two programming languages which were completely new to me, namely Visual Basic for Applications (VBA) and visual basic VB.

2.3.2.2 Data Logging system for garment factory

The application was to develop a system to log the total hours that the JUKI machines were being used. We had to show the cumulative value for each of the 80 odd machines.

The input was taken from the VSD. We had to use the modbus protocol and use that to interface the inverters to the automation system.

Here I faced a big problem in the programming section due to the non availability of multiple timers in VBA. Therefore I had to shift my platform from VBA access to VB which supported timers.

Sadly I was not able to extend the Application to create a user friendly database in access. The company lost interest in this project, and due to time constraints I had to stop my research on implementing an access database using VBA to log the work hours of the JUKI machines.

I also learned advanced Programming using Basic script, which was a relatively powerful tool compared to GUI blocks. The method I introduced was much more powerful even when compared to basic script.

2.4 Industrial Control with PLC and PAC s’

Designing a control system for machinery starting from scratch is a daunting if not an impossible task. But this was the only method available at the introduction of electrical control of machinery. The electronics as well as software had to specially be designed. As the systems thus made were not flexible this had to be done repetitively for all the different kinds of machinery in the factory. Therefore due to the complexity and expense, industrial automation was out of bound to small and medium scale plants. With the advent of cheap microcontrollers and associated hardware this situation changed. A new architecture for industrial control came to be known as a Programmable Logic Controller (PLC). With this automation became a mainstream affordable concept.

The specialty of a PLC is its robustness, flexibility and reliability. Therefore it can cater any system requirement thus there is no need for separate unique systems.

The software used to program a PLC was created such that the electrical personnel associated with earlier relay logic systems could easily adapt to it. This language is called ladder programming language. With time communication technologies such as modbus over RS485/422/232/Ethernet, Profibus over fiber became standard features of PLCs’.

These more advanced PLCs’ are known as Programmable Automation Computers or PACs’.

Apart from being a controller itself, these support the ability to network thousands of such systems into one integrated SCADA and a fully automated control system which watches over thousands of machinery.

Power supply and communication ports

Processing Unit

Digital IO

Analog IO

Fig. 2.19

Figure 2.19 shows the PAC that I worked with during my training.

2.4.1 PLC interfacing

PLCs’ have analog input/output (IO) as well as digital IO. Analog IO can specialize such as k-type thermocouple (TC) as well as generic ones such as 0-10V. We program the PLC such that we get the desired outputs for all input combinations.

The PACs’ that we used (ADAM 5510 EKW) have the ability to interface with the PC. It also has the capability to extend the number/type of IO in PLC by adding extra IO cards or ADAM 4000 modules.

Figure 2.20 shows the PLC that we installed in the moulding plant.

Figure 2.21 shows how the PLC inputs were extended using an Adam 4018+ Thermocouple module.

Fig. 2.20 Fig. 2.21

2.4.2 Software

The software used to program the 5510 PLC is the ladder language. I was able to get some experience in ladder programming with the simple SELEC ladder program as well as with the much more sophisticated KW Multiprog. The projects associated with ladder programming are annexed.

2.4.3 My Experience

I was able to get hands on experience by working on the moulding machine project as well as the selec PLC.

2.4.3.1 SELEC PLC

The first assignment given to me was to design a ladder program to control the water level of a tank. The first program was tested using the selec PLC MM3010. Apart from the control algorithm development and interfacing, I also learnt to design a small HMI to interact with the users.

The SELEC PLCs’ are very cheap but are less reliable and flexible. (Do not have IO expandability or Communication). The display is very crude as seen in the figure 2.22.

The HMI

Fig. 2.22

2.4.3.2 Moulding Machine

The company had undertaken a contract to redo the control system of a plastic moulding machine. The moulding machine had been bought from abroad after been dumped due to been outdated.

The client company wanted to redo the system to facilitate PVC moulding (instead of plastic). Our company had almost finished the project when we started training. A lot of problems arose after that, starting from the Software connectivity, output cards to control algorithm. I had to help in all aspect of troubleshooting. I contributed in a significant way, by redesigning the HMI/OPC server, developing a PID algorithm for heater control and testing of external current to voltage converter.

The project was done in very difficult circumstances as the factory was situated in Ragama (2 hours travel time) and the factory was poorly maintained (high humidity, no clean water and no proper restroom or lunchroom). The air was polluted by PVC fumes (which was a byproduct of their machines). After finishing work we were always exhausted and drenched in sweat. Despite all obstacles we were able to do a pretty good job.

Fig. 2.23

Complex pneumatic circuitry controlled by Electric signals from the PLC


Figure 2.23 shows the complexity of the PLC machine that we debugged

2.4.3.2.1 Basic structure of our PLC based automation system

The OPC server is the intermediate layer which helps the PLC and HMI talk to each other. It acts as the interpreter. We need OPC as there are different HMI software and Different PLCs. If the HMI is manufactured to talk with a specific PLC directly, then it won’t be able to talk with any other (different make), so in order to be able to use any PLC with any HMI, the industry has defined the OPC standard. The HMI software manufacturer should present his specific OPC server software to the clients, and by using that, the clients can connect any PLC (Which supports Modbus) to that HMI.

Fig. 2.24

HMI software

OLE interface

Object Linking and Embedding

COM port (RS232)/modbus

RS485/modbus

OPC server

PAC (ADAM 5510)

ADAM 4018+ (separate data acquisition module)

2.4.3.2.2 Problems faced

As this was the first project using PACs’ for me as well as the company we faced a number of frustrating problems.

2.4.3.2.2.1 OPC non connectivity

The first major problem that we faced was the inability for the OPC server to automatically connect to the PLC at system startup. So every time we have to reset it at startup. After much studying we were able to find the culprit as the assigned modbus addresses.

Due to the higher number of IOs’ the modbus addresses had got mixed up which resulted in an abnormal behavior of the system. So I had to find the modbus to OPC mapping formulae by going through ADAM forums. After that I made a excel sheet and checked the reassigned address mapping.

2.4.3.2.2.2 Uncontrollability of heater temperature

The second problem which arose was the uncontrollability of the temperature of the heaters. They were extremely powerful heaters. Though the ambient temperature for the process was 300C, the temperature soared over 300C without stopping even after turning the digital off the signal. They had to be controlled to be within 10C of the set point. In order to counter this effect cooling fans were fixed but that too did not have a significant effect on the stability. I was given to solve this problem. I decided on pulse width modulating the digital signal. Then I designed a PID algorithm for it. This algorithm would have given optimum stability. Unfortunately they did not implement my algorithm; instead they turned off the heater about 10C before the set point and simultaneously turned on the fans. Though this works for the tested range, the performance of this primitive algorithm over the full dynamic range is unacceptable. Given different set points this system has a big chance to become unstable unlike my strategy.

2.5 Industrial communication system

My company had undertaken a subcontract for a CEB communication system. I was involved in that project from the very beginning. The idea was to use GSM data call technology to send system fault messages to the Main control room of the CEB. The GSM modems were to be interfaced to the autoreclosers in distributed grid substations throughout the country and all fault conditions were to be reported to the main control room via the GSM communication links. The initial testing was conducted at the wattala substation.

We were given one autorecloser unit to test in our workshop. We tested the connection with the cable as well as through the GSM network. Though it connected to the PC through the cable link the GSM data call link did not work. In order to solve this problem I contacted one of my former lecturers working at dialog and got advice. Finally it was found out that the problem was not with the GSM service provider but with a system variable which had been incorrectly configured. My research indicated that data call technology is not reliable nor cost effective communication solution for this application whereas GPRS being ideal. The CEB engineers had been misled by some shady consultants.

Communication Port (Serial)

Fig. 2.25

Figure 2.25 shows the autorecloser Remote Terminal Unit(RTU) That we tested on.

Throughout the testing phase in wattala I and my training mate had to explain the technology behind GSM data calls and autorecloser to the technical officers. The surprising thing was that a lot were not even aware of the operating principle of an autorecloser. The project as a whole was a novel refreshing experience for me.

The telit GSM modems which were used in this project had an integrated python interpreter engine, giving the users the ability to use it for simple control tasks. Therefore I had to learn the rudiments of the python language and test the control functionality of the modem.

Fig. 2.26

Figure 2.26 shows the Telit GSM modem that e programmed to be used with a autorecloser RTU.

I also learnt how to download programs to the modem as well as how to use the modem as a standalone controller.

E.g.: to send a sms (using AT commands) to a pre determined number if some specific event occurs.

Python language was a pleasure to learn due to its simplicity and robustness. The script we write is interpreted through the python engine.

In the process I had to learn the basic AT commands used in a system. In fact we studied it by using the modem as DCE and HyperTerminal as DTE. The PC acted as a dumb terminal while all actions were taken by the modem.

The two were connected using a null modem serial cable. After doing those tests we connected another modem to the system and tested the wireless communication.

Fig. 2.27

Figure 2.27 shows the architecture of the Telit modem which we used.

I exploited all the architectural functionalities to get the maximum performance from this sophisticated modem.

The project was a success and culminated in a big profit for the company. Due to this success, the company is to be handed the task of developing communication software as well as importing the modems in the next CEB project.

Figure 2.28 shows the system architecture of the communication platform used to service the CEB substation to control center communication.

Fig. 2.28

Telit GSM

Telit GSM

PC

Serial port

PC

Serial port

GSM substation

AT SUBSTATION

AT CONTROL CENTER

2.6 Motor speed control using Variable Speed Drives (VSD)

The three phase induction motors are the most widely used motors in the industry. My company had the dealership for two of the world renowned brands of VSD, namely cutes and Gefran.

In order to help troubleshoot a VSD I had to visit a garment factory in Ratmalana with a technical officer. There was no troubleshooting to be done as the machine has not worked only due to being incorrectly wired. There was no motor theory involved rather it was a case of connecting the correct terminals, shorting the correct pins for the required functionality and setting the menu options. We entered the motor parameters and finally set the correct RPM. Then we demonstrated the operation of the VSD to the technician.

The VSD that we installed is shown in the Figure 2.29

Fig. 2.29

2.7 Expert standalone communication module

An integral part of a Data Acquisition system is the ability to integrate with a larger system. In that respect a RS485 to Ethernet converter/controller is of paramount importance.

Therefore I was given the task of designing a system to facilitate communication between Ethernet, RS232 and RS485 networks. The Ethernet controller EX9188END had to be programmed to get the required functionality.

Firstly I had to configure its communication (RS232) to connect to the PCs’ HyperTerminal and execute remote commands.

After that I downloaded the demo programs to the controller and executed them. So now I had a working stand alone controller. Finally I started work on the final stage. That was developing my own programs. At this stage I faced a number of problems. The first was the type and version of compiler compatible with the demo programs (as I hoped to initially edit the demo programs before going to more advanced stuff). After much corresponds with the company I finally got the required compiler Turbo C++ 4.5.

Then after I set the correct path to the headers I started getting an unusual linker error. I corresponded with the EXPERT Company and posted to all the support forums but did not get a sufficient answer. As I didn’t get an answer to my repeated queries the project had to be abandoned.

The Expert module that I worked with is shown in figure 2.30.

Fig. 2.30

2.8 ADAM 6051 Ethernet to RS485 converter

This project gives another method to interface a RS485 network to a PC. The biggest advantage is that we can log on to our SCADA from anywhere in the world (given that we have access to the internet)

The SCADA is actually a java web applet embedded in the HTML code. This set up in the web server which is directly connected to the controller.

ADAM 6051

Web Server

Internet

RS485 network

Fig. 2.31

The project was very interesting as the applications were vast. We had to download the required libraries from the ADAM forums as they did not come with the software CD.

We initially coded the GUI and tested it. It worked with ease. Then we went to the second phase that is to get the system IO value displayed in the SCADA. After compiling, this gave a linker error. Though we tried hard we could not find the bug (Due to not having studied the java language). We had no clue on how to include pre compiled java libraries (binaries) in to netbeans IDE. We emailed Advantech support but they too were not supportive.

Though we could not complete this project we got exposed to SCADA development environment and java programming, this significantly enhanced our awareness of these new technologies. The web server part helped us to learn to setup a basic web server and configure an internet router.

Figure 2.31 shows the architecture of the communication system

Figure 2.32 shows the basic components of the system that I setup.

To the web server

Ethernet Router

ADAM 6051 CONTROLLER

Fig. 2.32

2.9 Research and development (R & D) projects

All these projects were designed and implemented by me (with no external help) with the help of a HNDE trainee who was training alongside me.

2.9.1 RTD to 4-20mA

Up to the present the company had outsourced the special interface circuits needed for the automation systems. The general Manager requested me to do this in order to see if it is possible to build all the circuits inside the company. This was the first R & D project and was completely implemented by myself without any external help. The aim was to construct a low cost RTD to 4-20mA converter.

Initially I studied the operating principle of the RTD sensors. The main component is a precision resistor (in PT100: platinum resistor with exact resistance of 100ohm at room temperature) which has linear variation of resistance with respect to temperature.

The sensor is given a reference current and the resulting current is usually measured by the data acquisition system and retransmitted.

I went through a lot of manuals and finally found an IC which had the requirements that I was looking for. It was XTR 101 from Texas instruments. I used its reference current source of 1mA to convert the temperature into voltage and used the current retransmission capability to send a direct current output varying between 4-20mA.

Fig. 2.33

I had to import the IC from United States while buying the other components (caps, resistors) from Pettah. The initial prototyping was done on a breadboard. Then I faced the problem of testing. As there was no multimeter with an accuracy enough to measure a mA current, I had to improvise. I used a known (200ohm) resistor as the load and sent the output loop current across it. Then I measured the voltage across the resistor. I got the current by dividing voltage by 200.

The graph obtained was approximately linear, throughout the operating temperature range of 100 C. The experiment became an overwhelming success. After completion, the circuit was handed over to the general manager.

The prototyped circuit is shown in figure 2.33.

2.9.1.1 The problems faced

The non availability of precision resistors (1%) was a major problem. The errors were in the range of 10% thus the errors were not acceptable. The breadboard had internal short circuits. Therefore the circuit failed at the first attempt and only worked when we removed the short. Figure 2.34 shows the completed product.

Fig. 2.34

2.9.2 Thermocouple to 4-20mA converter using microcontroller

This project was also entrusted to me by the General Manager. Unlike the former project this was more complex and costly. The first phase of the project was to develop a proper interface for the Thermocouple. I searched all the available literature to find an IC which satisfies the requirements. A big problem was that almost all signal conditioning ICs’ needed the Cold Junction compensation be done manually. After much research I found the MAX 6675 chip manufactured by maxim as the most suitable one. This had an inbuilt micro sensor (thermistor) to sense the room temperature and do the required Cold Junction compensation. Then I had to order the chip online. It took 3 weeks to arrive. Finally I had to construct an interface board. A big problem faced was the ICs’ dimensions being small as it is surface mounted type. It was impossible to prototype on the breadboard. We had to create a schematic and give it to a PCB maker to get a proper interface board. This process of constructing a PCB board was a new experience to me.

MAX 6675 this is a very small chip

Fig. 2.35

SPI output

Fig. 2.36

The soldering was also done by the PCB manufacturing company.

The second phase of the project was the interfacing of PIC microcontroller and LCD display. This was a pretty straight forward task as the output from MAX6675 is Serial (SPI). After interfacing we programmed the PIC to display the temperature.

The final stage was the development of the 4-20mA retransmission unit. This phase is still in progress. For this we have imported a very sophisticated IC MAX 5661, the problem we are facing here is the difficulty to solder. This chip is very small (smb) but has 40 pins. So we need to make a board to interface this. We hope to finish this and give an accurate 4-20mA output by the end of April.

2.9.3 Thermocouple to 4-20mA converter using XTR101

Due to the higher cost of the PIC system we have proposed a low cost (without display) Thermocouple to 4-20mA transmitter. That project is still in the design phase. The biggest obstacle we are facing is the need to manually configure thermocouple cold junction compensation. The precise diode needed for this is not available locally and has to be imported.

2.9.4 RS232 to RS485 converter

The final project was to develop a converter. During research it was evident that there was no expensive technology behind the conversion from 232 to 485. We only have to change the voltages from 232 voltage to 485 voltage and then use a 485 transceiver to convert the protocol. Despite this the imported converters are unbelievably expensive. During training I was able to design a prototype circuit. The construction and testing is underway.

2.9.5 Microcontroller lectures

We were given free microcontroller lecture sessions on Saturdays (sponsored by the company). Here we got a comprehensive idea about using microcontrollers for Automation.

2.10 Marketing session

I had to interact with customers on daily basis be it to answer customer calls, give customer support or test returned modules. Apart from all those encounters one stands out as a memorable experience.

I had to go with the technical officer to ARPICO at Nawinna. There we were to show the factory engineer the SELEC energy meters and promote it. It was considered to be a low key demonstration (in fact I was wearing a very casual dress).

Going there the factory manager spoke with us for some time and then without any warning what so ever, he took us to a full profile board meeting of the directors of the Bogawantala plantations. We were indeed very embarrassed as we were not prepared for such a high profile meeting. The whole board of directors started barraging us with questions. As the technical officers’ communication skills was below nil, I had to answer all the questions. The fact that we were marketing Indian and Taiwanese brands were against us as they perceived those to be of low quality. The biggest obstacle here was the fact that almost all did not understand the complex electrical concepts which are a part and parcel of such presentations. Somehow or another I stood my ground and was able to sway their opinion. In the end when we left I was able to create a favorable impression about our company as well as the products. It was an unforgettable experience for me as it gave me an insight on how to improvise and do persuasive presentations.

2.11 Problems faced

2.11.1 A problem with no tangible solution

A customer returned 3 selec timers stating that they were faulty. According to him the problem was that the module didn’t reset when the electricity supply is switched off manually but reset when electricity supply is cut off from the CEB. There was no explanation to this strange phenomenon or a method to test it. Having searched for an answer I came to the final conclusion that the abnormal behavior occurs due to the instantaneous surge that occurs when electricity supply is switched on (from CEB) after a black out. The electronic circuitry of these modules is over sensitive to this surge and this result in the complete erasure of the EEPROM memory. However there was no way to test my theory as high voltage testing equipment were not available. Figure 2.37 shows the faulty module.

Fig. 2.37

2.11.2 Burning out of a RS 485 to RS 232 converter module

Due to my over enthusiasm to test a PC based heater control system I mistakenly connected 230 V AC instead of 24 V DC. This resulted in the burning out of a converter and digital IO module. This was the darkest moment of my otherwise awesome training period. Rather than negligence it was my over enthusiasm which caused it. From this I learnt that whatever the need, must test at least twice, before giving power to any equipment.

2.11.3 Interaction with staff

As we were trainees we were not taken seriously by some clerical staff. Initially this created some problems. But after some time they understood where we stand and took us seriously. In fact we became quite chummy by the end of the training period.


No comments: