The ever helpful Steve has aslo been building a binary counter but he has done it in a very different way to mine, he has used a chain of 4 linked D type flip flops.

To give you a brief explanation of what I mean by D type flip flop I have to go back to my logic gates:

This OR gate looks perfectly normal but what if we…

…link it’s output back through one of it’s inputs? This turns it into a digital logic with feedback device. Any digital circuit using feedback is called a multivibrator, the example I used above is a bistable multivibrator as it is stable in one of two states, you can also get monostable (one stable state) and astable (no stable state) just like the 555 timer!

The logic circuit for a D type flip flop looks like this:

The latch in the truth table looks a bit odd though… We know that a high input to a NOR gate will gove a low output, the circuit will ‘latch’ in the state after anytime that the inpit is high.  When the input is low the output could be either high or low depending on the circuits prior state.

It’s quite quite simple but a little longwinded to explain fully but if you fancy finding out exactly how they work have a look here.

So a D type flip flop is a bistable multivibrator, a digital logic device using feedback with two stable states. 

As I said before Steve has been using these to build binary counting circuits so how does that work? First you need to start by looking at the input and output of a flip flop. The input can come from something like a 555 timer astable circuit, this high, low, high, low wave form goes into the clock input and then is fed back through the data input, this results in the second wave form from output Q. The second wave form is half the frequency of the first, the flip flop divides it by two. This is really important for making the counting circuit.

When you feed the output from one chip into the clock input of another it divides the frequency in half again so the following signals will look like…

…this. Look familiar? It’s the same as the pattern of the outputs from the 4510 chip Each output fed into each subsequent chip divides the signal forming the pattern of a binary counter. 

So whats the difference between the two? As I have mentioned in my post on counting circuits there are two types of counting circuit, ripple and synchronous. The arrangement of flip flops I have described above is an example of a ripple counter. It is vary simple and does work but it has it’s disadvantages. The first is there is a tiny delay as the original clock signal ‘ripples’ through the circuit and the second is that if you are using it to trigger a logic circuit it can give glitchy signals, for a ripple counter to count from 0111 (7) to 1000 (8) it will go through 0110, 0100 and 0000 before it settles on 1000, this happens too fast for us to see but not for a logic circuit. 

A synchronous counter is like the one I built in my previous post.It has a much more complex structure which does the job of making sure that all the outputs change at exactly the same time on each clock pulse. It also avoids the tiny false counts so you could use them to trigger a logic system.

CMOS gates are sensitive to static electricity and can be damaged by high voltages, they may also assume any logic level if they are left ‘floating’. Say if you have a switch connected to your CMOS chip through to ground, when the switch is pushed ground is connected to the input pin. Unfortunately when the switch is open the signal to that pin is open to interference and static, this is a ‘floating’ state. Floating is bad, very bad. It can cause damage to your chip and send misleading signals to your circuit. To get around this you use a pull up or pull down resistor.

A pull up resistor does exactly that, pulls a signal up. It holds the pin in a high state stopping it from floating and so getting rid of any unwanted signals. Pull down resistors do exactly the same thing just pulling the signal low.

This diagram shows an input that is floating, the switch is open so the pin is subject to static and interference:

This one shows the positioning of a pull up resistor (for a pull down the switch and the resistor would swap places):

Most useful

Rapid look through equivalent chips with Dean and it looks like I can use a 4028 chip instead. A 4028 is officially known as a BCD (binary coded decimal) to decimal (1 of 10) decoder, this should perform the same function as the decade counter would in my counting circuits.

First I want to see if I can build a binary counter running off pulses from an astable 555 circuit, here goes…

The circuit diagram is very simple, the pulses generated by the astable 555 circuit feed into a 4510 binary counter chip, this converts the input signal into the 4 output signals needed to generate the binary count, it’s much easier to see in a diagram:

The chip uses it’s network of tiny logic gates to interpret the input from the 555 and turn it into a binary count. I find it almost easier to understand looking at these wave forms rather than truth table below, but that’s just me, I get on better with visual representations :):

The 4510 is a BCD chip, this means that it will count a binary sequence representing the decimal numbers 0-9 hence binary coded decimal counter :) 

So I’ve calculated my resistor values to give me a good frequency of pulses, I’ve hooked up the 4510 according to the pinout below found in it’s data sheet (more on data sheets here) time to test.

Clang! not working! Hmmm… everything is in the right place, the two chips are hooked up properly what could be wrong… at this point enter Dean, saviour of all malfunctioning circuitry…  

I hate it when it’s something really simple that I haven’t spotted, my reset switch was pushed into the board the wrong way round, *smacks head with palm of hand* and this is after Steve showed me how to use a multimeter to see how components should be connected up.

Now that’s fixed here’s the circuit working:

An extension of this would be to link up lots of timers together. To do this you connect the carry out (pin 7) to the carry in/enable (pin5) of another timer, this second counter now counts up 10s. By cascading them in this way you can count much larger numbers.

Note (07/04/11) There are two types of counting circuit, ripple and synchronous. The linked counters above form a synchronous counting circuit, a ripple counter is made up of linked flip flop chips. More on them here along with a more detailed look into ripple and synchronous counters.

Now I can have a go a converting the digital binary signal I have to a decimal format…

I get to have a go using the 4028 chip :) All you need to do is to hook up the 4 outputs from the binary counter to the 4 inputs of the 4028, then, carefully reading the pinout (they’re not in order as you can see below!) attach your LED outputs.

If you were using a divide by 16 binary counter you’d need to add in an extra step to limit the counter to 0-9. To do this you need to connect an AND gate using outputs B and D as the gates inputs. This would have the effect of triggering the reset every time the counter got to 10 (B and D going high together) keeping it counting from just 0 to 9.

The truth table for the 4028 is really simple, once you get past the volume of 0s and 1s the information is easy to extract:

Right, that’s the board put together, safety checks done and schematic double checked, power up

Crud! Crud! Crud! another no go! what could be wrong this time? The astable is still going, the signals to the binary are still working but the signals going to the 4028 are really weak… once again, Dean to the rescue. This time I really have had a proper brain has temporarily left the premises moment. I’d got the leads coming off the vss and vdd pins going to the opposite of where they should be. There is no excuse for this (apart from said stepping out of brain) as the data sheet clearly shows the right connections. Must slow down a bit…

Anyway now it works :) 

You can take this circuit another step forward by replacing the 4028 with a seven segment display driver. It’s a very similar set up, the 4 outputs from the 4510 go into the display driver (4511) and the outputs then connect up to the inputs on the seven segment display. I may have a go at this later on but for now I think it’s time I revisited competence…

I’m going to take a look at the 3 types of circuit you can make with a 555 timer, astable, monostable and bistable.

Monostable – one stable state

This is my monostable circuit and I’ve calculated that I need a 100K resistor with my 100µf capacitor to give me a time of 10 seconds…

Taking into account the tolerances of the resistor and the capacitor that’s pretty good going. If I wanted to get exactly 10 seconds I could use a POT or a variable capacitor.

Here you can see the capacitor charging using a multimeter (more on those here):

So a monostable circuit generates a single pulse when triggered, without the trigger the circuit produces a low (zero) voltage which is it’s stable state. I pulled this brilliant diagram from a site that’s been really helpful, it shows the wave forms of the signals involved in this circuit:

(http://www.kpsec.freeuk.com/555timer.htm#monostable site accessed 18/03/11)

I had one problem with this circuit in that my LED output kept burning out after 3 or 4 triggers so I decided to investigate…

My multimeter showed that 6.5 v was getting through to my poor little LED, it shouldn’t get more than 1.7V so my resistor was really not limiting enough of the current.

Using Ohms law I worked out which resistor would be best:

R = (VS-VL)/I

VS – supply voltage

VL – LED voltage

R = (6.5V-1.7V)/2ma

R = 4.8/0.02

R = 240R

So by adding another 240R to my 470R resistor I should have enough resistance to stop the burnout. As there isn’t a resistor with a value of 710R I used the next highest value which is 750R.

Monostable circuits are often used to trigger digital logic devices. Mechanical switches can ‘bounce’ when they are closed, causing voltage spikes that would confuse an IC into thinking it’s gotten a high signal. By putting a monostable in between your mechanical switch and your end device you can stop confusing signals.

 Astable – no stable state

An astable circuit is sort of like an electronic metronome using a combination of two resistors and a capacitor you can configure your circuit so you output, say an LED, will turn on and off at a set frequency. It has no stable state because it keeps changing on its own.

I wanted to make my output LED blink on and off for 5 seconds each, 5 on and 5 off. To do this I needed an (initially) scary range of calculations.

The first is for working out the frequency of your circuit, this is how many times per second your circuit completes one cycle of it’s wave form:

Frequency is measured in Hertz so a frequency of 2Hz would mean that your circuit would complete its wave form twice per second.

The next two are for working out the high (on) and low (off) times:

T high = 0.69(R1+R2)xC

T low = 0.69(R2xC) 

So… you start by working out the resistor needed for the low time first:

T low = 0.69(R2xC) 

5sec = 0.69(R2x100µf)

R2 = 5/0.69×0.0001

R2 = 72,464 or 72K (I tried it with a 68K)

Then you can work out R1:

T high = 0.69(R1+R2)xC

5sec = 0.69(R1+75K)x100µf

R1 = 5/(0.69×0.0001)-68,000

R1 = 4,464K (I used a 3K4)

So…

And here it is in test:

Note (25/03/11) Cool siren effect using two astable circuits and POTs! I think I’ve just made a device for communicating with dolphins… or for just really annoying dogs…

The effect of using two astable circuits feeding into the speaker and controllable from the two potentiometers gives you an enormous range of (really annoying ;)) sounds:

Bistable

Right, we’ve had one stable state, no stable states and now… yes you guessed it two stable states!

If this sounds a little like theintro to a cheap game show then I’m setting the right tone…

A bistable state circuit can be used as the buzzer system on a quiz show. The contestants each have one trigger, to trigger the high state and the quiz master has the other to reset it.

This one is the easiest one by far as there are no resistor values to calculate!!

Note (25/03/11) Mnemonic for remembering resister colours:

First, the Darlington driver or Darlington pair. You basically feed the emitter of one NPN transistor to the base of another, this multiplies their individual current gains so massively boosting the signal. Darlington pairs can be used like this to convert analogue signals to digital ones. If you had a potentiometer connected up to the base of the first transistor and an LED being driven from the other you can see this effect in action. Instead of the LED getting brighter or dimmer as you vary the resistance it would only switch on and off as the base current in the first transistor became high enough or low enough to affect the output of the second.

Most transistors have a threshold of 0.7V, when it the current through the base reaches this it ‘switches on’ and completes the circuit.

Note (20/03/11) Here’s a cool little circuit I put together to demonstrate a Darlington pair:

Note (20/03/11) And a look at how potentiometers work. A Potentiometer or POT is an adjustable resistor. It has 3 connections, one connects to the power source, one goes to ground and the last runs across a strip of resistive material. It acts as a wiper running across the strip which at one end has a very low resistance and the other which has a high resistance. the wiper is connected to a knob that the user can interact with.

On to capacitors – a capacitor is a bit like a bucket but it collects and stores electricity not water.

If you were to put a voltmeter across the capacitor in the circuit and plot a voltage/time graph it would look like this:

If you were to hook up a few more components to the circuit above you can convert the analogue input of the capacitor charging to a digital output:

The Darlington pair in the process part of the circuit is boosting the signal being taken as the capacitor is charging, this means that the LED will only be on or off, if the pair was replaced with a single transistor the LED would brighten or dim.

In this circuit the capacitor is acting as a delay device, as the capacitor fills the voltage increases until the required voltage to activate the transistors is reached to then turn on the LED.

So…

…this arrangement will act as a timer! You can control the time by using bigger or smaller resistors or bigger or smaller capacitors.

The equation to work out the timing is:

T=RC or Time = Resistance x Capacitance

To get a 30sec time you can work out the components you need like this:

(I’ve chosen to use a 10µf capacitor if the resistor or calculated from this was huge or tiny I could adjust this value)

30sec = R x 10µf

R = 30sec/10µf

R = 30/0.00001 (always convert your values! 10µf is actually 10^-3)

R = 3,000,000Ω or 3M

This is a bit big… I could change the capacitor to a 100µf and use a 300K resistor instead

Note (20/03/11) Polorised and non polerised capacitors: Capacitors with values over 1µf tend to be polorised, meaning that the way round you connect it in a circuit matters. If you connect it the wrong way round you can cause the dielectric inside the capacitor to break down. This can short circuit the capacitor, damage other components in the circuit (by sending them too much current) or your capacitor could even explode!

Instead of using a Darlington pair you can feed this signal through a 555 timer…

I’ve found out a bit more on NOT gates, they are actually called inverters, not surprisingly as they ‘invert’ the input into the output. When you connect up two inverters in series you get a ‘buffer. All this does is take the output of the first NOT gate and feeds it into the input of the next one. The two cancel each other out so you the final output is the same as the first input:

Possibly pointless? Not really, it does have a practical use. Logic gate circuits are signal amplifiers (remember logic gates are made up of those lovely signal boosting transistors…), a weak signal can be boosted using this combination of two inverters. The logic level isn’t changed, you still get a high output from a high input but the final signal is stronger than the initial one. The symbol is the same as for a NOT gate just without the little bubble on the end.

With additional inputs logic gates can process more complex information, for example a 2 input gate has 4 output possibilities, a 3 input gate has 8 output possibilities etc.

The equation for working out number of possibilities is…

Number of possible input states = 2ⁿ 

Where n = number of inputs

This means that your gates can control much more complex systems.

I had a go at prototyping a logic gate circuit using a NAND gate chip but first I should probably say how to interpret/draw logic gates in circuit diagrams. You can do it in one of two ways either by drawing in the IC (labeling it with the chip number) and numbering the pins:

Or by using the logic gate symbol and labeling the inputs and outputs. The label on the symbol below indicates that it is 1/4 of a 4011 chip, so only using one of the 4 NAND gates contained in this chip.

(Please excuse the clumsy editing, I couldn’t find how to edit the circuit symbol in circuit wizard…) 

Here’s my circuit in action:

I also found out that NAND and NOR gates have a cool capability, they are universal gates. This means that by arranging NAND and NOR gates in specific ways you can get them to mimic any other gate:

So to make an OR gate you need to connect up the gates in your IC like this:

(All truth tables can be found here)

There are also complementary output gates, this basically means that you have a gate that can produce inverted and non-inverted outputs. So you could have an AND gate that could give you an AND gate out put or a NAND gate output:

So why would you need a gate that can do this if you could just hook up a NOT gate off the output lead? For two reasons, 1) it saves space and 2) adding in a seperate NOT causes a time delay so if your inputs were alternating high and low your non-inverted output and your inverted outputs would change at slightly different times. 

Again this is getting a bit far afield from what I’m aiming for but would be interesting to look at in the future.

There are loads of logic families but the two most famous (and the ones I’ll likely be using) are TTL and CMOS.

TTL stands for transistor transistor logic, and uses bipolar transistors to construct gates and amplifiers. You can fit more gates in and they’re cheaper to make but they draw a lot more power and need a specific 5V power supply to work. Most TTL IC chips use the 74xx or 74xxx part numbers e.g. the 7400 is a quad input NAND gate (there are 4 NAND gates contained withing the chip).

CMOS stands for complementary metal oxide semiconductor, this is a type of technology used to make MOSFETs (metal oxide semiconductor field effect transistors, what a mouthful, you can see why they shortened the name to CMOS!). They’re a bit more expensive to make but they draw less power are operate over a wider range of voltages. One problem with them is that they are very susceptible to static electricity so you have to be careful when handling them (I’ve already had one issue with static messing up my circuit). Chips in the 40xx series are part of the CMOS family.

TTL chips are not really used anywhere outside of schools now as CMOS chips can do the same jobs without the drawbacks.

Note (07/03/11) – B series CMOS chips have ‘buffered’ outputs to increase the voltage gain from input to output (Have a look at buffers here), this means that they have a faster response to input signal changes. 

You take a Boolean expression of a logic circuit (the bars over some of the outputs mean that they are inverted)…

Simplify it using regular algebraic rules and Boolean logic rules…

And convert the result back into a logic circuit…

This subject is huge and requires going much deeper into the subject than I have need for in this project but I found this site a great resource when I was trying to understand the basics: http://www.allaboutcircuits.com/vol_4/chpt_7/1.html

It’s amazing that a system of logic developed in the 1800′s could have such a huge significance in developing cutting edge technology today.

Op Amp - this IC takes a weak signal and gives it a boost, it’s made up of several transistors, resisters and capacitor. It gives a much better performance than using a single transistor. The op amp is in fact an analogue IC, it takes the input voltage through the two input leads shown below and produces an output that (and here’s where it gets technical…)  is a multiple of the difference between the two inputs, so not just high and low.

the actual gain involved can be controlled using resistors, one of these control circuits is called an inverting amplifier. The values of the resistors shown below can be used to work out the gain using this equation – Gain = R2/R1, so for instance if the resistor values were R2 = 10k and R1 = 1K the gain would be 1000/100 = 10 so a signal of 1V would produce an output of 10V.  

(As this is analogue and I’m working on digital, I don’t think I need to go into any more detail to op amps at the mo.)

555 Timer - This popular little IC can be used in analogue and digital circuits, from my first look at them it seems there in everything. It’s well worth getting to know what each of the pins does on a 555 as it makes it easier to understand what’s going on inside. I pulled the descriptions below straight out of one of my text books:

Trigger - when you apply a low voltage to pin 2, you trigger the internal timing circuit to start working.This trigger is know as an active low trigger.

Output – The output waveform appears on pin 3.

Reset - If you apply a low voltage to pin 4, you reset the timing function, and the output (pin 3) goes low. (Some circuits don’t use the reset function, and this pin is tied to the positive supply.)

Control – If you want to override the internal trigger circuit (which you normally don’t), you apply a voltage to pin 5. Otherwise, you can connect pin 5 to ground, preferably through a 0.01µF capacitor.

Threshold – When the voltage applied to pin 6 reaches a certain level (usually two-thirds of the positive supply voltage), the timing cycle ends. You connect a resistor between pin 6 and the positive supply. The value of this timing resistor influences the length of the timing cycle. 

Discharge – You connect a capacitor to pin 7. The discharge time of this timing capacitor influences the length of the timing intervals.

(Ross, Shamieh, McComb,  2010, p.153-4) 

There are 3 main ways to use a 555 chip, the different configurations producing different outputs. They are…

As an astable multivibrator (oscillator) - astable just means that the output has no stable state, it doesn’t settle down but keeps changing on it’s own. You can use this for flashing lights,making an alarm or triggering a logic chip…

As amonostable multivibrator (one shot) - monostable means this circuits output has one stable state, it generates a single pulse when triggered. This state is really useful for triggering logic devices.

As a bistable multivibrator (flip flop) - bistable means (you guessed it) this circuits output has two stable states, it switches from one state to the other when it’s triggered. You can make a quiz show buzzer system using this state, it’s in it’s low state as the question is being asked, a contestant hits a button to answer triggering the high state, this is then reset in to the low state by the quiz master.

4017 CMOS Decade Counter - This cool little chip will count from 0 - 9 when it’s triggered, by joining lots of them together you can count up tens, hundreds, thousands etc you have to be careful though… the output pins are not in order down the side of the chip, you have to check the configuration in the ‘pinout’ diagram below.

4000 CMOS Series Logic Gates - I’ve kind of started looking into these gates here, but there are a couple of cool experiments I can do to physically show how gates work. I’ll put these in a later post.

So a brief history of logic circuits, the only place to start is with relays.

Relays are electromagnetic switches that physically open and close. The pic above is a broken relay that I took the case off, you can see the coil of copper wire and the connectors at the front. They were initially developed by an American physicist Joseph Henry (1797–1878), and  were eventually used for things like transmitting Morse code, telephone communication and, as it is an on (1) off (0) device, early logic circuits. If you fancy a deeper read…

have a look at this article it’s brill http://history-computer.com/ModernComputer/Basis/relay.html  

Relays were replaced by vacuum tubes in the 1940′s – 50′s, the benefit being that there were no moving parts, the vacuum tube was an entirely electronic switch. A vacuum tube can switch electricity on or off, or amplify a current, the first digital computer, ENIAC, was made using vacuum tubes. They had their problems though, vacuum tubes look and act a lot like an incandescent bulb, they generate a lot of heat and have a tendency to burn out. Another problem was size, you can imagine that if your computer at home needed at least one vacuum tube for each logic circuit it would be massive! ENIAC had a whopping 18,000 tubes!

Take a minor detour if you wish to my cool stuff page on Nixie Tubes, they’re like vacuum tubes but used to display numbers, one of these days I’ll get around to building that clock…

OK, the next step was when the vacuum tube was replaced by the transistor.

This picture should give you a start on the advantages of transistors over tubes The transistor was invented in 1947 and it revolutionised electronics. Without them your phone wouldn’t be mobile, there would be no such thing as the laptop (or desktop or even a ‘fit in your house’ top).

Transistors amplify current and can also be used as electronic switches so they perform the same job as the vacuum tube but with less power, more reliability and less heat. Transistors are made of semiconductor materials such as silicone, these are materials that hover somewhere between holding onto their electrons and letting them move about freely. So they act as conductors under some conditions and insulators under others. Have a look at the symbol for a transistor below and you’ll get an idea of how a transistor works as a switch:

When the current going to the base is low (0v or 0) you can imagine the end of that arrow as sitting away from the emitter lead so there is no connection, when the current going to the base is high (+5V or 1) it’s like the arrow moves over and completes the circuit allowing current to flow. So there’s the switch function, but what about amplifying current?

Here’s a little circuit experiment showing this effect (for more info on potentiometers have a look here):

So the tiny base current passing through LED1 is being amplified by the transistor allowing a larger current to flow through LED2. You get a dim light from LED1 due to the tiny base current and a much brighter glow from LED2 because of the stronger collector current.

For a more detailed read on transistors have a look here http://www.kpsec.freeuk.com/trancirc.htm

Transistors can work using digital signals, this makes them perfect for constructing logic gates. The diagram below is a (simplified) circuit of a NOT gate:

A NOT gate is really just a transistor, with a couple of resistors and diodes thrown in to protect the transistor and keep the current flowing in the right direction but you get the gist of it. It’s even better when you look at an AND gate:

It’s clear that you can build any logic gate you need just by arranging resistors in specific ways.

As I mentioned above, if you had to build a complex arrangement of components each time you wanted a logic gate you’d go mad, so the next step was the emergence of the IC.

 Using transistors for complex circuits was great till you got above a certain size, then errors started creeping in. The sheer number of components meant that shorts and broken connections became more frequent and the circuits became less reliable the bigger they got, advanced circuits contained so many components and connections that they were almost impossible to build. In 1958 an engineer named Jack Kilby had the idea of building these circuits out of a single block of semiconductor material, effectively removing the errors caused by broken connections, this was the beginning of the IC. Inside each IC is a collection of microscopic discrete components made out of the same bit of material, they are arranged in such a way as to act like any number of logic circuits.

The best article I have found that explains this in much more detail is here: http://nobelprize.org/educational/physics/integrated_circuit/history/

Inside a logic gate IC you’ll usually find an arrangement of 4 gates of the same type e.g an IC NAND chip:

This shows the arrangement of inputs and outputs on this chip, when you need a NAND gate you just hook your inputs up to the correct pair of pins and connect what ever you want driven by your output up to the corresponding output pin.

Wow, that was quite a journey, now to look at the sort of chips I’m likely to come across in my travels, what they do and how to use them…