Surface Mount has introduced a new range of case styles for diodes, transistors and chips. Here is a guide through the maze of new styles: SO means surface Mount. SOT is the case style for 3 or 4 leaded devices:

THE SIZE OF SURFACE MOUNT
Surface mount is the transition between the components we know today and those of tomorrow.
As far as designers are concerned, Surface Mount is only a temporary, intermediate stage to assist in reducing the size of a product to bring it on the market as soon as possible; while the circuit is further developed and perfected.
The aim of a designer is to put all the circuitry into a single custom-designed chip. But going from conventional componentry to final design is often not possible in one step. Sometimes it requires a few intermediate stages, requiring custom chips plus a few discrete components. To do this, it is convenient to use surface mount.
Products of tomorrow will be designed around a single dedicated chip with only the need for batteries, switch and speaker etc.
We can see this already with calculators, watches, LCD games, cameras and many more brilliantly designed products. They contain a single dedicated chip and nothing more.
There are still a lot of other devices in the transitional stage and these are currently using SM technology to reduce the size.
As the demand for a particular product increases, the cost of re-designing the circuit (and the chip) becomes economical. Eventually everything is incorporated into a single chip and the evolution is complete.
This is the way surface mount is assisting designers to bring out new products in steps and stages and as it gets phased out of one design, another is taking on the transition.
It's difficult to know how long surface mount will be around, but just like everything else, it will eventually be outmoded by a radical new approach.
Until this time comes, there will be a demand for engineers to design and work in this field and new openings are appearing all the time. (Some surface-mount repair-work is also required in the industry, but since this area of technology is so reliable, the demand is very small. It mostly involves replacing mechanical components such as switches and motors.)

We really shouldn't be presenting a "hands-on" project for Surface Mount as it has been exclusively designed for robot handling, where "pick and place" machines take components from carriers, spools, hoppers, tubes etc and place them on a PC board with really amazing speed and precision.
But if you intend to design projects in this medium, it's absolutely essential to know how big the parts are, so that you can get some idea of the space they occupy and how the board will turn out.
It's also essential to know how the devices are attached and the process of soldering so you can design the board correctly.
Since normal surface-mount boards are Wave Soldered, it is important to place the components at 90º to the action of the wave so they do not form a shadow and prevent the solder from touching the pads.
This also applies to placing components in the shadow of taller components, and there's 100 other tricks you have to learn.
So, if you want to start, here's your opportunity. There's nothing like experiencing it first hand and if you put a kit together you can at least say you have got off the ground.

SURFACE MOUNT ASSEMBLY
Since there are no leads on surface mount components, we do not need any holes in the board (for multilayer boards, holes are required to join one layer to the next - they are not for the components) and the components can be loaded at impressive rates of something like 3600 devices per hour for a single head machine.
To prevent them dropping off the board before the soldering process, tiny drops of glue are used and this keeps them firmly in position so the boards can be turned up-side-down for wave-soldering.
Although Surface Mount devices are more expensive than conventional components, the higher cost is offset by the savings in board space, loading time and neatness of the final design.
This is where Surface Mount wins hands-down. A Surface Mount project looks much smarter and more up-to-date than a conventional design. The size of the board can be reduced by up to 50% and this makes it very attractive for products that need to be compact such as cameras, video machines, computers, and almost any other item you can think of.
Some new chips are only available in Surface Mount and this forces the manufacturers to take on the new technology.
This is exactly what happened to us. Our new speech chip (our solid-state tape-recorder project) is only available in surface mount and this forced us into the new technology. The same is happening in design laboratories all around the world and in fact, SM is seeing a growth rate of 15% per year with much of the new growth taking place in the high volume, high-technology area.
The range of components is increasing too and already we have Surface Mount coils, inductors, transformers, and other devices you never thought could be produced in such minûte form. There is no limit to the ingenuity of engineers. Eventually standard components will be taken over completely and although this may take a number of years it will certainly occur. We have already seen the demise of the valve, the 1 watt resistor and many other components. It's only when I see them in a museum that I marvel at the leaps we have made in technology.
If you think this won't happen to conventional componentry, you are kidding yourself. These changes are as certain as progress itself. They are driven by economics and economics runs the world.
Surface Mount has arrived and is here to stay. Manufacturers can see the savings in this area and are setting the pace.
And, believe me, the pace is mind blowing. Apart from miniaturisation brought about by surface mount, we have large scale integration and multi-layer PC boards. These have combined to shrink devices from desk-top size to palm size - with 10 times the features and 100 times the power!
Manufacturers of VCR's, TV's, automotive instruments, cameras, computers and TOYS have all opted for surface mount as their preferred form of construction.
Why have we included toys? Because toys represent one of the greatest driving forces in the electronics industry. Many of the recent advancements in electronics have taken place through the medium of toys.
Talking dolls, whistling keychains, swearing keyrings, etc, have all introduced the "dob of electronics."
If you have ever taken a whistling or talking keyring apart, you will find it contains a single chip mounted directly on the PC board and covered with a dob of potting compound. Some of the earlier designs used discrete chips and a few resistors but as the designs were tidied up, this was converted to the "dob of potting compound." This is commonly called COB (Chip On Board).
The reason why manufacturers use toys as the proving ground is quite simple. It provides a huge market that can be supplied and tested in a very short space of time. This means the designers get rapid feedback and the product can be updated and perfected very quickly.
Speech chips have followed this trend. Early chips cost more than $70 each while the latest are surface mount and cost less than $10! You can buy 30 seconds of digitised speech on a keychain or in a credit card for less than $8!
The main problems with early chips was understanding the robotised speech. It was difficult to work out what was being said. But the new chips are so clear that you think it's a tape recording.


HANDLING SURFACE MOUNT
You won't believe anything I say about the size of these components until you see them for yourself.
The size is totally unbelievable and it may take a while to build up enough courage to take them out of their carrier strips.
Normally these devices are not handled by humans and it was never intended for them to be touched at all. The fact that some components have values marked on them is merely a result of pressure from end-users.
There is no real need to have any markings on them as they are handled from start to finish by computer controlled insertion equipment.
Even testing and alignment of the built-up board can be carried out by robot testers and so component marking is a bonus for us.
A very small percentage of surface mount devices are soldered by hand in short-run productions and in these cases the operator works under a low magnification lenses or with the naked eye. To do this sort of work you need to have very good eyesight, nimble fingers and a calm temperament.
Once the initial shock of the size subsides, you can get down to organising your soldering equipment and see if you are going to be able to physically handle the task. You may need tweezers to pick up the parts and something to hold them in place while soldering.
To do this properly you really need three hands but if this is not available, you will have to use some other means of holding the part while feeding the solder and using the iron.
If you have someone that can help you, now is the time to enlist their assistance. I'm not going to discuss the need to be an expert at soldering as the sheer size of the components will keep any absolute beginner away.
However I am going to say that you can forget the cheap and rugged 40 watt soldering iron, the instant heat iron and many of the other so-called electronic soldering irons such as the gas iron, soldering gun and even the 700°F soldering station. They are all far too hot and/or too cumbersome for this type of work. You need a soldering iron or station with a temperature of 320°C to 350°C and a very fine tip. When I say fine, I mean a tip that will almost prick you if you touch it.
This fineness is absolutely essential for soldering the pins of the IC as the lead spacing is half that of conventional IC's. Some of the other components can be soldered with a medium tip, but certainly not the IC.
We will not be gluing the components to the board before soldering as the glue is very expensive and has a life of only a few months. Instead we will be holding each item in place with a probe (such as a paper clip or fine screwdriver) while tacking it in place, prior to soldering.
This is where you can ask for assistance by getting someone to hold the component or add the solder while you solder it in place.
Some components, such as ceramic capacitors, are not identified in any way while those that are marked, (the resistors) require a magnifying glass to read the numbers.
We have placed the components in a carrier strip in the kit and enclosed a note to let you know how they are arranged. Do not take any of them out of the strips before they are required as you will not be able to identify them if they are mixed up. The old motto "look but don't touch" certainly applies.
As we said in the introduction, this project is a TELEPHONE RING CIRCUIT using a mini piezo as the output device and a CMOS Schmitt trigger as the oscillator and driver.
One of the outputs of the chip also drives a LED via a transistor and this has been done to add a transistor to the board.
Some constructors will say the chip is the hardest component to fit while others will have enormous difficulty with the transistor.
In fact, this project would be ideal as a soldering test for advanced students as it will not only test soldering skills but also neatness, placement of parts and identification.
I believe a similar project was passed around a group of 20 workers at a hi-tech plant with the requirement to desolder all the components from the model and solder them back in place.
I understand that all the components withstood 20 solder and desolder operations without a failure.
The fact is, surface mount components are extremely robust if soldered quickly at the correct temperature. They are designed to withstand a 10 second submersion in molten solder or other fluid, but if you subject them to a higher temperature, you run the risk of premature and permanent damage.

ASSEMBLY and SOLDERING in INDUSTRY
There are three types of Surface Mount assembly. The first is placement of surface mount components to one or both sides of the board. The second has both surface and through-hole components on one or both sides. The third type has through-hole components on the top side and surface mount components on the bottom.
The different loading techniques for these boards calls for different soldering methods and the most common methods are: Reflow and Wave.
In the Reflow method, solder is screened onto the pads in a printing operation or individually added by means of a gun. The trackwork has been previously protected with a mask to prevent solder creating shorts and bridges. The components are then added and kept in place with solder paste or tiny dobs of non-conductive glue.
The boards are then passed through an infra-red or convection oven that allows the solder to melt.
Another method of reflow is to immerse the board in a saturated vapour of a boiling Fluorinert liquid.
The vapour, at the temperature of the boiling liquid, gives up its heat, causing the solder to flow. In the wave soldering process, the board is dipped in flux and placed up-side down over a bath of molten solder. A wave of solder is created that rises up to touch the board and complete the soldering process.
These processes sound very simple but in fact involve a high degree of technical skill. For instance, if a reasonably complex board has 100 faulty joints per million, the yield is almost zero and nearly every board has a faulty connection!
With surface mount, the soldering process not only has to provide good electrical connection but since the leads do not go though holes, it has to provide good mechanical connection too.
The design of a surface mount board becomes much more critical than a through-hole board due to the size of the components, the size of the lands, the placement of the components and the consideration given to heat stress both during and after soldering.
This is an entire subject on its own and technical centers can be contacted for more information for those who want to be involved in this area.
Along with the different soldering processes there are a range of soldering faults, where the components have either dropped off the board or begun to stand up due to a number of problems. The most common fault is called "tomb-stoning" where capacitors, resistors and packages stand on end after soldering. This results from improper pad design, unequal solder mass, shadowing of the component, misplacement of components, poor quality solder paste and wrong soldering temperature. Fortunately, we won’t have any of these problems in this project as everything will be soldered by hand.

HOW THE CIRCUIT WORKS
The circuit consists of 6 building blocks and the first is the inverter between pins 1 and 2. This forms a low frequency oscillator with a 1u and 1M + 2M2 resistor. It governs the overall timing of the ring by creating an ON and OFF time. When the output is LOW, the tone is emitted from the piezo. When the output is HIGH, the tone is inhibited and this produces the silence between the rings.
This oscillator has an equal mark-space ratio to give the "rings" the same length of time as the silence.
The second oscillator operates at about twice the frequency of the first (this can be seen by the different value of the resistors as both capacitors have the same value).
The frequency has been adjusted so that it produces two highs during the interval when it is activated. The second oscillator does not produce two full cycles but only one and a half as it is the HIGHs that are required.
During each of these HIGHs, the third and fourth oscillators produce a warble that simulates the 33Hz ring of the "bell." The third oscillator generates the 33Hz frequency and this gates the fourth oscillator to produce a 1kHz tone for the piezo.

The output of this oscillator drives the base of the buffer transistor and also one side of the piezo. The other side of the piezo is connected to the output of two buffers in parallel and this provides good pull-down capability when the left side of the piezo is high.
The only fault in the design of this circuit is the drive to the left-hand side of the piezo. We should have included driving buffers to give it the maximum swing and thus the maximum output. But since we did not have any left over, this is the best we could do.
When the piezo is driven from a pair of buffers on each side, it sees a voltage swing of nearly twice the rail voltage and this gives it the highest output.
The tone is also passed to a LED via a transistor to give a visual indication of the operation of the circuit.
A 47R resistor has been included in series with the LED to limit the current. It is essential to include a resistor as the LED drops a fixed voltage (called the characteristic voltage drop) when it is illuminated and the transistor drops a fixed voltage across the collector-emitter terminals when it is turned on.
The voltage drops are 1 .7v for the LED and .5v for the transistor. This adds up to 2.3v and thus we must include a resistor to drop .7v from the 3v supply rail. By making the resistor 47R we allow a maximum of 1.4mA to flow.
Without this resistor the power rails would be pulled down to 2.3v every time the LED is turned on.
This would cause (a) a very high current to flow through the LED and (b) faulty operation of the circuit as the power rails fluctuate.
The 1u electrolytic across the power rails reduces the impedance of the battery and provides uniform rail impedance during the life of the battery.
A lithium battery has been used as it produces 3v so that we only need a single cell to provide the minimum voltage for the chip.

THE GATING DIODES
Between each of the oscillators is a diode called a gating diode. Its function is to turn the oscillator on and off when required. Here is how it works. We will use the second oscillator as an example, as shown in the figure below.
When the first oscillator (between pins 1 and 2) is HIGH, it is equivalent to connecting the anode end of the diode to the positive rail and this will have the effect of charging the electrolytic. The diode will be able to supply more current to the electrolytic than can be bled away by the 560k resistor and thus the capacitor remains charged. This means the inverter (between pins 3 and 4) will not change state and it is thus 'jammed."
When the first oscillator goes LOW, the diode is effectively connected to the negative rail and ceases to have any effect on the second oscillator.
By turning the diode around the other way, the oscillator will be blocked or jammed by a LOW from the previous oscillator as the diode will bleed away any charging current so that the capacitor will not rise higher than about .6v.
This is a very handy way of gating or controlling an oscillator by the use of diodes.
 

CONSTRUCTION
The object of this project is not to rush,  but take it slowly and produce a neat result.
Start by creating a clear space on the workbench and get all the necessary tools and equipment ready. Make sure all the parts are in the kit by checking it against the parts list and lay everything out neatly in readiness. Look at the carrier strips so that you know what’s inside. Clean the tip of your soldering iron on a wet sponge and open out the paper clip supplied in the kit to form a probe to hold the parts during soldering.

When the cell is fitted, the LED will begin to flash and the piezo will produce a sound similar to a phone ringing.
If it doesn't, you may have a fault and if this is the case, you can count yourself lucky as you will be able to go over the project and diagnose the fault with the assistance of our "If It Doesn't Work section." This is where you will start to learn about electronics and the project will have great benefits.

IF IT DOESN'T WORK
If the circuit doesn't produce a sound similar to a phone ringing, you will have to work out where the fault is coming from by reading this section. There are possibly over 50 faults with a circuit as simple as this as any two components could be swapped, any of them could be faulty due to overheating or the board could have a short between the tracks.
To locate the problem, here is the approach:
The first thing to do is measure the internal impedance of the battery. This is done by setting the multi-meter to 500mA range and placing the probes on the battery for 1/2 second. The needle should rise to about 200mA or more to indicate the cell can supply driving current.
Next, measure the current taken by the circuit by placing a piece of plastic under the cell so it doesn't make contact with the board. This plastic can be used as a switch to turn the project off when not required. Measure across the plastic with a multimeter set to 50mA (or 500mA to be on the safe side). The current should be about 1 to 2mA and you can change the range to 5mA to get an accurate reading.
If it is considerably more than 2mA, you have either damaged one of the components or created a short.
Make sure there are no solder bridges between tracks or under the chip by inspecting the board carefully. Next cut the negative track going to pin 7 of the chip so that half the circuit is removed. Re-measure the current to see if the remaining parts contain the fault. Refer to the circuit diagram to identify which components are in this section and if the fault persists, make another cut in the trackwork and "home-in" on the fault. This will save you removing any of the parts and testing them, as soldering and desoldering will create more problems than it solves.
If this doesn't find the fault, you will have to read on. In this type of project we start at the back-end and work to the front. This is because we have a LED and piezo to act as output devices to let us know what is happening and the extent of the fault.
We start with the LED and its driver transistor. If the LED does not light, the fault could lie in either of these components or the chip.
To locate the problem, take a voltage reading at output pin 12 of the chip.
The needle of the multimeter should flicker to correspond to the ringing of the circuit and if not, the fault will lie in the chip or one of the four oscillators.
If the needle flickers, go to the collector of the transistor. Here, you will see the needle sit at slightly above 1v (due to the characteristic voltage drop across the LED plus a very small drop across the resistor) and fall to slightly less than 1v when the circuit produces ring pulses. If the LED does not produce a glow when this occurs, it has either been damaged or is around the wrong way. If the needle does not flicker at the collector, the transistor has been damaged.
You can also measure the voltage at the base of the transistor. The reading you will get will only be about 100mV as the needle will not have time to rise to 650mV during the ring. To get an accurate indication of the signal you must measure it with a CRO.
If the sound from the piezo is not similar to that of a phone ringing, the fault will lie in one or more of the 4 oscillators.
Start at the first oscillator, between pins 1 and 2. Place the positive probe on pin 2 and set the multimeter to a low voltage. The needle will go high for about 1 second and low for the same duration. When the output is LOW, a tone is emitted from the piezo. If the output does not swing up and down, measure the voltage across the power rails (pins 7 and 14) of the chip.
If voltage is present, and the current consumption is about 2mA, the fault may lie in a damaged Schmitt inverter, a leaky 1u capacitor or an open 2M2 or 1M resistor.
You cannot measure across the 1u while the circuit is operating as the resistance of your meter will prevent it charging to 2/3 of rail voltage and the oscillator will not change state.
The only thing you can do is measure the output voltage. If it is HIGH, the input will be low (assuming the gate is working) and one of the feedback resistors may be open circuit. Set your multimeter to 10v range (assuming a 20k/v meter) and place the probes firstly across the 2M2 resistor and then across the 1M. If you detect a very slight movement of the needle, the circuit is working and the resistance of the meter is taking the place of the resistor. Replacing the appropriate resistor will fix the fault.
If the circuit produces a ring-ring-ring-ring without a pause, the gating diode between pins 2 and 3 may be faulty or not making contact.
If the output of the first oscillator is correct, the output of the second can also be detected on a multimeter by probing pin 4. This pin will give two High's during the ring tone and if a fault exists, you can diagnose it in a similar manner to the first gate.
The output of the third and fourth oscillators are more difficult to detect on a multimeter as the frequency is too high for the needle to respond.
The solution is to remove the piezo from its output terminals and place it between pin 6 and the negative rail. Here you will hear a series of clicks to correspond to the 33Hz oscillator, gated by oscillator (b).
Placing the piezo between pin 12 and the negative rail will produce the ring sound except the output will be lower than when connected to the output terminals.
If you find one of the inverters has been damaged, you can use the one between pins 8 and 9. You will have to do a little rewiring, however if the chip is not drawing excessive current due to it being damaged, the change can be made. This just about covers everything and the project should be working perfectly by now - I hope so.

CONCLUSION
If you did not have any success, the best solution is to buy another kit and start again. The main problem will be soldering. Next time you will learn from your mistakes and the project will work first go. For those who have tasted the joys of success, try some of our other projects using surface-mount technology.