-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueIPC APEX EXPO 2024 Pre-show
This month’s issue devotes its pages to a comprehensive preview of the IPC APEX EXPO 2024 event. Whether your role is technical or business, if you're new-to-the-industry or seasoned veteran, you'll find value throughout this program.
Boost Your Sales
Every part of your business can be evaluated as a process, including your sales funnel. Optimizing your selling process requires a coordinated effort between marketing and sales. In this issue, industry experts in marketing and sales offer their best advice on how to boost your sales efforts.
The Cost of Rework
In this issue, we investigate rework's current state of the art. What are the root causes and how are they resolved? What is the financial impact of rework, and is it possible to eliminate it entirely without sacrificing your yields?
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
Cost of Ownership for Dispensing Equipment
December 31, 1969 |Estimated reading time: 8 minutes
It’s not news that electronics manufacturers are looking to reduce manufacturing costs. Saving money on the purchase of capital equipment seems like a good solution. But selecting equipment based solely on price can be deceiving. This article applies SEMI’s guidelines for dispensing equipment to show what equipment characteristics are most important to reduce production costs.
By Glen Gibbs and Al Lewis
Electronics manufacturers are looking for ways to reduce manufacturing costs. Saving money on the purchase of capital equipment - a major component of overall manufacturing cost - seems like a good solution. However, selecting equipment based solely on its price can be deceiving. The true determining factor becomes production cost per part. To make that determination, other factors come into play, such as fixed costs, labor, recurring costs, yield, and the number of good parts produced. In the case of dispensing equipment, an improvement of 0.01% yield can make a dramatic difference. Doing a complete lifetime cost of ownership analysis enables the manufacturer to evaluate real equipment cost, and provides a tool to determine how best to reduce total cost of ownership.
SEMI developed standard E35-0305 - Guide to Calculate Cost of Ownership Metric for Semiconductor Manufacturing Equipment.1 This guide serves to provide a standard metric for evaluating unit production cost-effectiveness of manufacturing equipment in the semiconductor industry. It establishes a well-defined procedure to facilitate an understanding of equipment-related costs by providing definitions, classifications, algorithms, methods, and default values necessary to build a full or constrained cost of ownership calculator. This article applies SEMI’s guidelines to dispensing equipment. Application of this model shows what equipment characteristics are most important to reduce production costs.2
Cost of Ownership Model
Cost of ownership (COO) equals the total cost divided by the total number of good parts, or COO = Total Costs/Number of good parts. The total cost of owning the equipment consists of the sum of four major parameters:
Figure 1. Cost per part in U.S. dollars.
Total cost = F$ + L$ + R$ + Y$.
• F$ is the initial fixed costs involved in purchasing the equipment, such as purchase price, cost of installation and initial training, system qualification, and any movement of the equipment in the factory.
• L$ is the fully burdened labor cost.
• R$ is the sum of all recurring costs associated with the equipment over its lifetime, and includes such things as consumables, maintenance, utilities, floor rent, and specialized support personnel.
• Y$ is the yield cost associated with equipment mishandling of devices. The cost of lost yield due to the equipment is the product of the probability that the device will be damaged by the equipment in question, multiplied by the cost of the device at that processing step, or Y$ = N × P$.
The total number of parts produced over the life of the equipment = L × T × Y × U.
• L is the entire lifetime of the equipment.
• T is the theoretical maximum number of parts processed per year.
• Y is composite yield, or the number of good units produced.
• U is equipment utilization. There are five factors that cause utilization to be less than 100%. These factors can best be described in the formula U = 1 - {(SM + USM + A + S + Q)/H}. All values are in hours per week.
a. SM is scheduled maintenance.
b. USM is unscheduled maintenance.
c. A is assist time.
d. S is standby time.
e. Q is qualification of the equipment (changeover and recalibration procedures).
f. H is total number of scheduled production hours per week.
Baseline Model
For this model, five examples were used with varying cost parameters. Those examples were:
Example 1: Equipment cost of $190,000, a yield of 99.99%, and a throughput of 800 units per hour (UPH).
Example 2: Equipment cost of $190,000, a yield of 99.90%, and a throughput of 800 UPH.
Example 3: Equipment cost of $190,000, a yield of 99.99%, and a throughput of 400 UPH.
Example 4: Equipment cost of $100,000, a yield of 99.99%, and a throughput of 400 UPH.
Example 5: Equipment cost of $100,000, a yield of 99.90%, and a throughput of 400 UPH.
As a baseline, for Example 1, a common surface mount dispensing application of underfill of a chip-scale package (CSP) was chosen. The baseline example is the high-end fluid dispensing system. Table 1 shows the actual cost values used in calculations for the baseline example. Parameters compared include an equipment cost of $190,000, a yield of 99.99%, and a throughput of 800 UPH. Using the values on this chart for fixed, labor, recurring, and yield costs, the total lifetime cost of ownership for the equipment is $799,375.
The cost of ownership model depends on actual production cost and utilization parameters. In this example, numbers representative of what one would expect to find in an electronics board-level manufacturing environment were used. In all cases, equipment life is assumed to be five years. Sample equipment utilization assumptions and calculations for Example 1 are shown in Table 2. During this time, 29,179,112 units are produced. With a total cost of equipment of $799,375, the cost per unit produced is $0.0274.
Effect of Yield
Changing yield by even a small amount can have a major effect on cost of ownership. Using the model and keeping all other values constant, we decreased the yield from 99.99% to 99.90% (Example 2). Figure 1 shows that a small decrease in yield can increase the lifetime cost of the equipment from $262,234 to $1,061,609. Per-part costs are increased significantly from $0.0274 to $0.0364. Relatively small increases in yield can mean large savings or extra costs. To decrease yield and maintain throughput, the number of defective parts had to be increased, showing the effect of a slight yield reduction.
In dispensing equipment, many features that add to the equipment cost are the same that contribute to improved yield, such as throughput advancements, fluid management, and process control. Examples of this are automated calibration that shifts to pre-calibrated motor speeds between dispenses; variable valve speeds; flow-rate verification; low-fluid-level detection; and other software interfaces, process control, and monitoring features. Other factors, such as using non-contact jet dispensing for a greater process margin, also can increase yield significantly.
Equipment Speed
Equipment operating speed (raw units per hours produced) has a near-linear effect on production costs. Features such as the use of jet technology instead of needle dispensing, high-speed fiducial recognition, and sophisticated software features to optimize dispense patterns can yield 3 to 6× throughput improvements. In Example 3, equipment speed was reduced from 800-400 UPH (Figure 1). Although reducing the throughput, or speed, also reduces yield, the number of defective parts was reduced to maintain the same yield. Total units produced decreases from 29,179,112 to 14,589,477, and the cost per part increases from $0.0274 to $0.0539 - almost double.
This assumes that the entire production line can increase UPH by the same amount. If the dispensing equipment is the slowest equipment on the production line, then the effect of increasing the dispensing speed is amplified by the improved UPH of the entire production line. If dispensing equipment is not the slowest equipment on the line, the increased speed has a negligible effect. This analysis only considers the effect of usable speed-change on a dispenser’s cost of ownership.
Effect of Initial Cost of a Dispenser
The baseline case assumes that the initial equipment cost is $190,000. For that price, one could obtain a well-equipped dispenser with the latest technology. Based on actual experience, data confirms a process yield of 99.99% on that equipment, and a throughput of 800 UPH. What happens when we try to save money by purchasing a less-sophisticated piece of equipment at a lower cost?
Assuming we can find a fluid dispenser for CSP underfill that costs $100,000, let’s calculate the cost of operation. Comparing the low-priced machine, Example 4, to the $190,000 machine changes more variables than just equipment cost (Figure 1). At that price, it is reasonable to assume that the equipment would lack the process control and advanced features found on the higher-priced system. Therefore, speed would be slower and yield lower. For purposes of this comparison, the same parameters from Example 3, a 400 UPH throughput and a 99.90% yield, will be used, but for the low-priced equipment.
In this example, total costs equal $695,725, while total units produced are 14,589,477. This equates to a cost per part of $0.0477. Compared to the $0.0274 per-part cost of the base model machine (Table 1), which costs $190,000, a speed of 800 devices per hour (DPH), and a throughput of 99.99%, the cost of ownership for the less-expensive machine is 74% more than the expensive machine. This gives the lower-priced machine a yield of 99.99%, which most likely is higher than that machine can produce. A more realistic yield is probably 99.90%, because lower-cost equipment usually does not include the process control required for higher yields. In Example 5, using a yield of 99.90% and a throughput of 400 DPH for the $100,000 machine, we see that the total cost is $825,792, and the price per part is $0.0567 (Figure 1). The cost of ownership is 107% more for the less-expensive machine.
Reducing Cost of Ownership
From Figure 1, we see what percentage of total cost is spent on each of the four factors: fixed, labor, recurring, and yield. When yield is reduced, what factors increase and what can be reduced to decrease COO? In addition to yield and speed, other dispensing variables include the cost of disposables used in the dispensing operation. What are the costs for dispensing fluid, needles, and pumps? Equipment that conserves fluid by measuring it properly and ensuring it is the correct temperature and viscosity so it is not wasted might be considered to reduce COO, rather than add to it. This is also true for other disposables required for equipment operations.
Conclusion
Cost of ownership is based on a multitude of variables and factors. Equipment cost is just one of them. Before purchasing any equipment, it is wise to use a model, such as SEMI’s, that is adapted for dispensing equipment.2 This analysis raised other opportunities for cost reduction. Recurring cost and labor are also large percentages of the total cost of ownership. Equipment that requires less user interaction can also reduce the cost of ownership. While it may require more sophistication and homework to apply cost of ownership analysis, it can help drive a good decision with long-term positive benefits.
REFERENCES
1 SEMI E35 - Cost of Ownership for Semiconductor Equipment Manufacturing Metric, Semiconductor Equipment and Materials International, San Jose, Calif., Dec. 2004.
2 Asymtek Cost of Ownership Model Calculator can be found under Tech Tools at: http://www.asymtek.com/support.htm.
Glen Gibbs, general manager, Asia Pacific, Asymtek, may be contacted at (304) 776-4802; e-mail: ggibbs@asymtek.com. Al Lewis, director of application engineering, Asymtek, may be contacted at (760) 930-3379; e-mail: alewis@asymtek.com.