Sorry couldn’t resist,  I love the IT Crowd

Ok, what am I fixing and what are my procedures?

I need to remake some of my circuits from my experimentation posts and give them to one of my lovely colleagues to sabotage. I should start with something relatively simple and work up to the complex stuff so…

• Start off with a simple logic gate circuit
• Step it up to a transistor circuit
• Move on to a basic monostable circuit with an LED output
• Finish it off with the binary to decimal counter

4 circuits of increasing complexity… working my way through these and returning them to a working state will prove my competency

I have my circuits but what are my procedures?

I’ve already done some work on fault finding and analysis but I think, considering my past performance, that I need a little help. Dean has highlighted 3 methods of circuit analysis that I can try:

• Top Down – starting at the input, working through the process and finishing with the output
• Bottom Up – starting at the output, working through the process and finishing with the input
• Split – starting with the process and working outwards

Looking at these three methods I can see that they would be suited to different circuits. You might use the top down or bottom up method on a simple circuit but it would take ages to go through a large or complex circuit all the way from input to output. For this type you’d use the split method, identifying the signals going into and coming out of your process section, this narrows it down a little faster.

I’m getting a little ahead of myself here, the first thing I need to do (considering the mischievous nature of my fellow students) is to make sure I’m not going to have my eyebrows singed off first! . If I’m going to hand someone a circuit and ask them to break it there is the distinct possibility that they might put some of my components in the wrong way round, this could result in said eyebrow singing if I just powered it up straight away. Back when I first started to work with electronics I put together a fault finding document, I then added to it for use with prototyping methods. Both procedures included a visual inspection, this has to be the first thing that’s carried out. The next would be a powered inspection and finally using the tools I’ve investigated for analysis using the new methods I’ve learned above.

So my final… awesome… circuit-in-distress-saving  procedure is:

Visual

• Double check your circuit matches your diagram.
• Is everything pushed in properly?
• Are all polarised components in the right way round?
• Are your stripped wire ends the right length? (5-10mm)
• Have any connections snapped?
• Are the values of your components correct? (resistors, capacitors etc.)
• Is your IC in the right way round?!?

Powered

• Is anything heating up? (Be careful!)
• Does it smell of burning?

Tooled Up

• Top down, bottom up or split depending on type and size of circuit.
• Multimeter to check continuity and voltage values.
• Ocilloscope to check signal types and changes.

Time to get building those circuits and choosing someone to break them…

Our old friend, the oscilloscope, has made another appearance. It first made itself invaluable in my keypad safe when my chip malfunctioned.

This time round it has helped me get a look into the signals travelling through my circuits. In the binary counter circuit I could see the pulses from the astable timer, how they separate into the 4 outputs of the binary counter, the wave form diagram is in my experiment post. The oscilloscope basically displays how voltages change with time, it shows as a trace across a CRT screen. This makes it really useful for checking out whats going on inside your circuit, you can see digital signals changing and check the signals timing. There is another device that can be used to look at digital signals, it’s called a logic probe. It has two LEDs that turn on and off depending on the high or low signal. One advantage of the oscilloscope over the logic probe is that you can look at any speed of pulsing signals on the scopes screen but the logic probe cant handle anything over about 5MHz. 

Ah, the multimeter, where would I be without you. A multimeter is an amazing tool that can measure voltages, current, resistance, test diodes and transistors and basically is an all round wonder tool. One thing I have found especially useful has been the ability to test components such as switches. During the keypad safe project I needed to add in a switch But I couldn’t work out which way round it needed to go as it was a SPDT (single pole double throw). By setting the multimeter to ohms you can see which connections work together, on an open switch you should get an infinite reading and on a closed one a reading of close to 0. So by experimenting with which connections you attach the multimeter to you can see which points are connected in which switch position . This also helped when I needed to troubleshoot my binary counter, Dean had to jog my memory as in my frustration I seemed to have forgotten all my analysing skills! There is also an audible continuity function, this is just a tone that sounds when the multimeter detects continuity in a wire or connection, very useful to detect breaks in your circuits. I’ve got a couple of videos of multimeters in action, one showing a capacitor charge here and one showing the changing resistance on a potentiometer here.

Datasheets are also a big help when constructing and troubleshooting circuits. They have all the details for a component, have a look here 4510 datasheet and here 4028 datasheet  for the data sheets I used in my counting circuits. It has the maximum and minimum working conditions, dimensions, logic diagrams and truth tables, with this information it’s easy to work out how to build a circuit and also to work out why it might not be working.

The last tool in my arsenal of analysis is to always keep in mind your circuits inputs, processes and outputs. If you power up a circuit and your LED doesn’t light up/flash then the best thing to do is to start by looking at the inputs. Is your power supply working and in the right range? Are all your connections to the process portion of your circuit correct and continuous? If everything here is working then move on to the processes, is your chip in the right way round is probably the best one to start with, then check the connections, you can look at the signals going into and coming out of your chip with a oscilloscope to see if they are what you would expect (you can also match them up to the signal diagrams in the chips datasheet). Lastly if everything else is working it must be your output. This is alot more productive than randomly testing portions of your circuit and lamenting that ‘It should be working!’.

After painstakingly putting the kit together and testing it, my contrary PIC chip has decided to stop working. Today I plugged it in to find that only some of the keypad keys are working, the LED won’t flash and the buzzer is making a constant low level bzzzzzzzzzzzzzzzzzzzzzzzzzz (very annoying when you’re trying to find out what’s wrong with it). None of my connections are loose, I’ve got no solder bridges and no crossed connections, time to ask for help from a higher power…. Dean, are you free?….

OK, time to run some diagnostics, after discussing my problem with Dean he has reccomended that we use an oscilloscope (such a cool name) to try and see what is going on inside my circuit. The oscilloscope will track any signals being sent from the legs of my PIC to the outputs on my board, the blip on the screen should jump up for a live signal or remain flat for no signal. Or in the case of my key pad it should jump up and down rapidly to indicate the PIC is scanning each of the output rows in turn.

Once I had taped over the buzzer (there’s only so long you can listen to it going bzzzzzzzzzzzzzzzzz without having serious thoughts about destroying it in an inventive fashion) we got down to investigating. Using the circuit diagram (see below) we tested each output in turn, it didn’t take long to see that outputs 7 and 12 had swapped places, the constant high/low signal that was supposed to be scanning one of the keypad rows was what was causing the buzzer to be so irritating. We’re still not sure what caused this to occur as the only thing that had happened to the board is that I had carried it home and back to the workshop, the only explanation we can come up with is that it suffered a static shock from the box I was keeping it in and it scrambled the programme. Right, keeping it in a static proof bag now…

One rapid reprogramme later and I’m back to having a working board, only 3/4 of a day lost…