For through-hole components, such as leaded resistors or the larger values of tantalum capacitors, whether insertion machines are used or the task is undertaken manually, the process is to take the component from its packaging, preform the lead wires if necessary, and insert the component in the correct position:
At the point of insertion, both axial and radial components have the same characteristics, with leads on known centres and at 90° to the board. In order to make insertion possible, the diameter of the hole will be slightly larger than the lead, so that there will be little interference between lead and hole. As a result, the component could spring back out unless it were restrained in some way. This is normally carried out by ‘clinching’, where the lead is bent over sufficiently to hold the component firmly against the board. The design has to allow for clinching, as this creates opportunities for unwanted short circuits to occur.
Not all components are intended for mounting directly in contact with the board: for example power and high voltage components often require a ‘stand-off’. In order to achieve this, the leads are normally formed before taping. In this case, the form can be designed to leave some residual spring in the lead so that it is retained by the board after insertion, and such leads are generally not clinched.
Other components that may be formed before placement include transistors: modern constructions have three in-line pins, but it is common for the centre lead to be bent to give the equivalent of the triangular arrangement of pins on 0.1 inch centres that was originally devised for TO- package with glass-to-metal seals.
Dual-in-line integrated circuit packages have slightly splayed leads1, so cannot be mounted onto tape, and are normally packaged in tubes made of antistatic material. The ends of the tubes may be taped or plugged, or removable pins may be clipped through holes in the tube in order to retain the components.
1 The reason for this 15° splay on DIPs is partly to improve the interference fit with the through-hole which keeps the component in place during soldering, and partly so that, if leads are slightly displaced by poor handling they can still be brought into correct alignment by pressing inwards. It is difficult to devise placement fingers that are able simultaneously and selectively to push in those leads that are splayed out and pull out those that are pointing inwards.
Once the packages have been released by removing tape, plug or pin, they can slide out of the tube, ready for placement. Many machines will be gravity-fed, perhaps with some vibration, in the same way as stick feeders for chip components. The first activity is to locate the ends of the leads in the insertion jaws and then close them to the correct pitch, so that the body is central between the insertion jaws, and the leads are vertical.
As shown in Figure 2, the next move is to push the dual-in-line body towards the board, so that the leads are ‘combed’ to the correct pitch and mate up with the holes in the board. For dual-in-line packages, the stand-off required in order to be able to clean under the package is provided by the lead having a shoulder that prevents over-insertion. Underneath the board, in order to retain the DIP, it is normal to clinch only a small number of leads, usually those on opposite corners.
In SM component placement we look at the way components are presented, on reel or bandolier in tube, or on trays. All of these are to defined standards to allow interchangeability. There are, however, other options associated with components for insertion. Depending on their size and physical characteristics, components also presented loose, in boxes or bags, in boxes with foam inserts, or even separately wrapped. All these methods require some manual intervention and suggest the use of hand insertion rather than automatic equipment.
As described with placement, the issue is one of finding the right component and putting it in the right place with the right orientation. Drawings (or the computer equivalent) that show what the component looks like, where it goes, and what way round it goes, supported by part numbers on the drawing and on the box, are not a total answer. It is desperately easy to pick the wrong component and/or put it in the wrong place.
It is not surprising therefore that much work went into fool-proofing the insertion task. The writer recalls a trip to Productronika 1995, where no fewer than ten manufacturers offered equipment aimed at improving productivity and accuracy. A typical solution, applicable also to SM parts in low volume, is to hold the components in small trays mounted on a carousel, only one of which is made available to the operator at any one time. A laser beam, projected onto the board through a moveable mirror, indicates the position of the component. By making small movements of the beam, it is also possible to indicate the orientation of insertion.
Whilst equipment of this type is well suited to a task where an operator assembles many components onto a single board, frequent volume production practice is to use a number of operators and divide the task between them. The job design is important, splitting tasks which might get confused between different operators, and limiting the number of components at any one station. Typically the work-station is laid out so that the picking hand goes through an easily memorisable sequence, again to reduce errors.
The great advantage of the hand option is that people are generally more dexterous than robots, particularly when it comes to handling parts that are difficult to insert. Some kinds of connectors for example need to be slightly ‘waggled’ during insertion and will then lock into place on board. This kind of operation is difficult to achieve mechanically, and the rates are frequently lower than when using human operators. Hand insertion is also very effective when components have thin wires.
When parts have been hand-inserted, the wires may need to be cut to length. Also, depending on the soldering method used, the parts may need to be retained. Some of the options for wave soldering are discussed in the next section. For total hand assembly, wires are often not cut until all the components have been inserted, when a foam pad is placed over the board, the assembly inverted, and the wires cropped in a single pass. Very occasionally a diamond wheel is used to cut the wires, but in small-scale production the use of hand cutters is more common. It is important that leads are properly trimmed to the correct length.
Some semi-automated component locators are fitted with automatic cut-and-clinch, so that the operator is able to pick the correct part from a rotary bin sequencer, insert the component at the position indicated by the laser and then the machine will crop and clinch the leads when an operator signals that a component has been placed.
A good example of the genre is the Contact System CS400E (http://www.contax.co.uk). this has a cut and clinch mechanism with two cutter assemblies, one of which is fixed and the other moveable. Each cutter assembly has an inner and outer cutter, and the component leads pass through the centre home in the inner cutter. As the cutter assembly can rotate, clinching can be carried out in any direction, though it is normal to clinch both leads in one operation.
In the past, axial, radial and DIP components tended to use machines of different design. This was because the requirements for that three types of package are different and manufacturing pressures favoured optimising the speed for each type, even if this meant having separate machines, rather than a single multi-purpose insertion equipment.
As an indication of the potential for machine insertion equipment, the fastest two-head machines are now capable of 40,000 components per hour, but the older equipment most frequently seen is limited to around 10–15,000/hour for axial components and 8–10,000/hour for radial parts. DIP insertion is a more complex operation, running typically at 3–5,000/hour, which is substantially faster than the 500–600/hour of which an operator is capable.
Despite their speed, DIP machines are still the least common of the three types. This is at least partly because manual insertion is relatively straightforward, provided that the legs have been preformed downwards, a task which can be done with a simple jig with rollers, and a human operator is capable of handling many different package sizes “without changing the tooling”! Although there are other standard sizes, most insertion machines are optimised for packages with 0.3 inch or 0.6 inch between rows of pins.
Radial components are usually fed direct from bandoliers of different parts, closely packed at the rear of the machine. A component is cropped from the appropriate tape, then transferred to the placement position. Some of these machines are dedicated to components with specific lead-to-lead pitches; others are capable of being programmed (the ‘variable’ in VCD, the generic name by which these machines are known).
Axial components are supplied on separate reels, but traditionally were first removed and then assembled in the desired assembly sequence onto two carrier tapes, before being re-reeled onto a single reel. This was carried out as a separate operation using a ‘sequencer’, typically a very long machine holding over a hundred reels, and with a wire dispenser/cutter that could be programmed to insert shorting links into the sequence.
The advantages a sequencer gives are that the operation can be carried out off-line and does not impact on the insertion rate, and it is easy to create a stock of kits ready for assembly. The first disadvantage is that, once mounted into the sequences, the resulting reels of components are totally dedicated to a specific circuit assembly. A second difficulty arises when the insertion sequence fails. In most cases the machine logic prevents Component #3 going in Component #2 position and so on to the end of the sequence, but the operator has the problem of how to get a spare component – certainly not from the next sequence! The net result is that most insertion equipment operators keep local stocks of those parts that give most problems during insertion. These are ‘uncontrolled’ from the Quality Engineering perspective, and it is fortunate that most through-hole parts are visibly coded with their identity! A consequence of the problems with sequenced parts is that those manufacturers not needing the highest production rate are likely to use a VCD for axial as well as radial insertion.
As explained above, historically, insertion machines were made to take just one of the three lead formats, axial, radial and DIL, and such components, packaged on tape or in tube respectively, were about the only components that could be handled automatically. And of course the process was quite separate from the surface mount line and not integrated with it.
More recently, adding through-hole components has been integrated with what has become known as ‘odd-form’ placement, originally developed to meet the challenge of mounting more difficult components for reflow assembly.
The technology split has now pretty well stabilised: if a component can be picked up by vacuum, and the components can be accommodated in some way within a tray package, assemblers will choose to use a precision placement machine, even though special purpose nozzles and software may be needed to accommodate such features as holes in cans. This is because such machines are used in mainstream manufacture, and have been developed to the point where they are fast and flexible.
‘Odd-form’ placement comes into play when the component has to be gripped in order to achieve the assembly function – usually this is confined to through-hole parts, but this is not necessarily always the case. What an odd-form machine offers is a flexible way of handling a variety of components, based on intelligent heads, and intelligent lead cropping, with the whole system chock-full of sensors, in order to ensure that parts are correctly inserted.
Some of the ways in which parts are checked before mounting to ensure that they will mate with the holes in the board, and their position is subsequently verified, are shown in the linked animations below.
[ back to top ]