Productivity through flexibility

Andy Kellard, Europlacer, Poole

Compact, reliable and inexpensive, the mobile-phone handset has moved from the realms of science fiction to everybody’s pocket in about a decade. Such is the rate of engineering progress that last year’s model is obsolete, and the appetite for more or new features continues unabated. And yet advances in capability are not matched by increases in price. In fact, the opposite is the rule; consumers want more features for less money. It’s true that the mobile represents the extreme end of the trend, but the features-versus-price equation holds true for all electronics products.

For an electronics assembler, consumer pressure results in more complex products, which in turn demand greater component variety, count and density. Added to the demands for price reductions, the weight forces assemblers to find productivity gains. In other words: make more, improved products for lower costs. This is the starting point for capital equipment justification.

As in the heart of a line, a placement machine is the piece of equipment subject to the most rigorous appraisal. As an expensive bit of kit, a purchase decision involves both corporate executives and engineers. Executives will need to be assured that the supplier is financially stable and able to demonstrate a long-term commitment to continually delivering the R&D, equipment, innovation and support that its customers demand. Other ‘comforting’ factors – although difficult to quantify – can include whether the placer is sold via an agent or supplied direct from the manufacturer; or whether the machine is made locally or imported.

At an engineering level, the performance of the placement machine in day-to-day use becomes the primary selection factor. Measures such as speed, feeder capacity, changeover time and product variety all need to be considered in a justification. These various factors combine to determine the ma-chine’s productivity.

However, confusion can arise from the definition of productivity itself and how it is measured. Productivity is often confused as output, most directly determined on a placement machine as the rate at which components are added to the PCB. This is of little use on its own though, because it doesn’t take into account the capital cost of the placer itself. You’ll always find a faster machine, but if it offers a 50% increase in the number of components placed at double the cost, don’t go asking for a salary increase. Output is a valuable (and quantifiable) measure nonetheless, provided it is used in conjunction with the purchase price to calculate cost-per-placement.

In its simplest derivation, cost-per-placement equates to the hourly cost of the equipment (directly related to the purchase price), divided by the placement speed achieved on a product. Although it’s true that this doesn’t take into account other cost factors such as factory and company overheads, it does produce a figure that can be used to make meaningful cost comparisons between alternative equipment. The problem is, however, that simply measuring the throughput speed achieved on a particular product or test board won’t always give a reliable indication of what the cost-per-placement figure will be once the machine under trial has been installed on the factory floor.

It’s a lot like measuring the fuel efficiency of a car. To obtain a reliable indication of what this day-to-day use means, consideration is necessary of how the car is used. In a similar way, determining the actual performance of a placer under real operating conditions demands an assessment of the effect of key-operational functions, such as product changeovers and maintenance, that will inevitably produce downtime. Downtime that, by definition, will significantly reduce the productivity.

In financial terms, the purpose of placing components onto a board is to “add value” in order that a PCB plus assembled components is worth more than the sumof its parts. It is but a simple step to calculate the “profit” generated by the placer once the cost per placement has been calculated.

Let’s look at an example using realistic figures and assumptions (see table 1). Machine 1 is a European-designed flexible placer costing $478,000, fully equipped with accessories and feeders. It is able to handle a range of components from 0402s through to 0.4mm pitch leaded devices. Under ideal conditions, machine 1 is able to place 20,000cph, however, in normal operation, and on a variety of assemblies it is rated at 14,000cph. Changeover time is 15 minutes.

In contrast, machine 2 is a high-speed placer from a Japanese manufacturer, and fully equipped costs $575,000. This placer is more limited in its component handling capabilities, only accepting tape-fed devices. Being a high-speed machine, it is rated at 40,000cph, but this drops to 24,000 in normal factory operation. Changeover time is 45min, with a further 20min required for changeover validation.

The calculation of the cost per placement for machine 1 (see table 2) reveals the following: Assuming a single shift of 8hrs and four changeovers per shift, machine 1 is productive for

(8 x 60) – (4 x 15) = 420min or 7hrs

In this time the machine can place:

14,000 x 7 = 98,000

Which, in a year of 200 shifts, multiplies up to 19.6m components

Machine 1 investment is $478,000. Assuming this is paid over five years, the placer costs $95,600 per year. So the cost per placement is:

$95,600/19.6m = $0.0049 per placement

Repeating the calculation for machine 2, using the same assumptions, yields a figure of $0.0065 per placement (or 33% more than machine 1, as illustrated in table 1). Let’s take this one stage further and calculate the respective profits generated by machines 1 & 2 over their five-year payback period. For this calculation, we’ll assume that each placed component adds $0.0150 to the value of the assembly.

In the case of machine 1 the annual added value is:

19.6m x 0.015 = $294,000

The annual profit (added value minus costs) is:

294,000 – 95,600 = $198,400 ($992,000 over five years)

Repeating the calculation for machine 2, again using the same assumptions, yields a figure of $745,000 over the five-year period (or 25% less). Although machine 2 offers the highest potential output of the two placers, it is clearly less suited to higher-mix environments demanding frequent product changeovers. Remember that the example considers just four changeovers per shift; further changeovers would increase machine 1’s advantage.

It’s clear that the cost per placement on the faster machine is higher for two main reasons: a higher investment, and longer changeover times compared with machine 1. Higher speed placers tend to be more expensive because they need to be engineered to provide both high-speed and an element of flexibility to cope with an ever-diverging range of packages. The slower speed solution, however, doesn’t require expensive high-speed capabilities. Nonetheless, a lower speed machine can offer the flexibility of the faster machine – and often greater flexibility – while needing a far lower initial investment.

In this example, we have seen how extended changeover times can damage placement machine productivity. Even the fastest pla-cer can’t compensate for extended changeover times by brute speed alone. Dramatic cost per placement decreases can be enjoyed if the changeover time is shortened.

One of the most time-consuming elements of a product changeover is the process of locating the component families particular to the new assembly, and loading them into the correct feeder locations. If it happens that a location is already occupied by another component, then this item has to be removed or re-located to a different position.

A solution to the problem is to use intelligent feeders. These can be situated in any machine position, and the placer detects the particular component’s presence. The machine automatically modifies the placement sequence – into the optimum order – to take into account other component positions. It is a further advantage if the machine is capable of highlighting previously loaded feeders of the component types that are already on the machine, or ready for use but held off-line.

On Europlacer machines, for example, an operator can instruct the placer to show the required feeder locations. In addition, the machine can indicate if these feeders are on the machine or stored on a connected storage system. Each feeder is equipped with an LED, which illuminates to indicate its position to an operator. Conversely, the machine can show all redundant feeders for removal from the machine. If component types remain unidentified by the placer, it will flag these as missing items. A production run can commence with missing components – giving the operator time to load up fresh feeders to add to convenient machine locations prior to completion of the run.

With a little forethought it is possible to improve machine productivity even further by buying a machine with a larger feeder capacity rather than one with a higher placement rate. For example, some board assemblers have calculated their top 80% component types, and these stay on the machine permanently, so each changeover is limited to a maximum of 20% of component types used. This has proved a far more effective route to improved productivity than simply investing heavily in placement speed.

Placement machine comparison is not an easy process. Before setting out on theselection procedure, the user should remind himself that what really counts is how much it costs to place a componenton a PCB. Consider the variety of pro-ducts and components and the numberof product changes under normal con-ditions: all these will influence cost-per-placement. A high-speed machine alone is rarely the best solution in higher-mix environments; a placer offering reasonable speed, good component hand-ling and quick changeovers frequentlyis.

Die Kosten, die entstehen für die Plazierung eines Bauteils, sind in der Baugruppenfertigung der kritische Anteil der Bestückung. Die Lösung könnten High-Speed-Bestücker sein, die mit hohen Raten plazieren. Doch in typischen „europäischen“ Fertigungen mit mehreren Produktwechseln per Schicht können solche Systeme oft auf der Kostenseite nicht mit flexiblen Bestückern der mittleren Leistungskategorie mithalten. Ihr Kostenanteil per Bauteil liegt höher.

Les coûts résultant de l’implantation d’un composant représentent la partie critique de l’équipement dans la fabrication des cartes. La solution pourrait venir des systèmes à implanter les composants «High-Speed»,lesquels travaillent à haut rendement. Or, dans les fabrications typiquement «européennes», qui se caractérisent par plusieurs changements de produits par équipe, ces systèmes ne peuvent fréquemment pas suivre les machines à implanter les composants souples, de rendement moyen. Leur coût par composant est plus élevé.

I costi creati dal piazzamento dei componenti rappresentano nella produzione di gruppi costruttivi la voce più critica nell‘equipaggiamento.La soluzione potrebbe essere degli equipaggiatori High-Speed che piazzano i componenti ad alta velocità. In tipici siti produttivi „europei“ con molti cambi di prodotto per turno, però, simili sistemi, dal punto di vista dei costi, non possono concorrere con gli equipaggiatori flessibili di medio rendimento. Il loro costo/componente è più alto.

Fax +44-1202-266599

EPP 156