RFID Chip Assembly
              For 0.1 Cents?
UPH, and 5 years depreciation, the         process step, introduces a flip chip        consumes less material. Bea...

              Table 1 – Possible cost breakdown for a 5 cent RFID tag

Smart Labels, Tickets And Luggage Tags
On One Machine
                                                       ing ...
Upcoming SlideShare
Loading in …5

RFID Chip Assembly For 0.1 Cents?


Published on

Published in: Business, Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

RFID Chip Assembly For 0.1 Cents?

  1. 1. ASSEMBLY RFID Chip Assembly For 0.1 Cents? by Hugo Pristauz, Datacon Technology While the world is arguing over the “Nickel tag,” a passive RFID tag priced at 5 US cents, and whether it could be on the market by 2008, the suppliers of backend equipment are aggres- sively pursuing various technology avenues to decrease manufacturing costs and enable the ubiquitous use RFID tags. The application possibilities of RFID are endless. Technical analysis shows that broad-based applications of RFID will develop in two distinct phases. Phase 1, which we are in right now, is characterised by only marginal advances because RFID is superfi- Figure 1 - Benefits and trade-offs of conventional and visionary approaches cially tacked onto existing business processes. Phase 2 will bring about a truly revolutionary change in the global business environment, after radical process re-engineering and new RFID-centred business models emerge. This phase will unlock the full potential of RFID and turn it into a new base technology. In addition, analysts say that industry executives are focusing on the price of RFID Figure 2 - Direct and indirect chip assembly production flow tags and when they will come down. and optimised for decades and is advantages, consider the assembly of Price is the significant benchmark known as conventional die bonding high-end flip chip devices like DSPs for RFID tags to be accepted in technology. Based on this know- and ASICs where an assembly yield of R F I D broad-based markets. Today, there is how, equipment vendors derived ad- over 99.95% is a standard target. For talk of the “Nickel tag”, a fully func- vanced die bonding machines which low-end flip chip devices like RFID tional passive RFID transponder for can carry out high-speed flip chip tags, a yield of 98% is standard; an just a Nickel (5 US cents) per unit. assembly. Due to the huge potential optimisation to 99% yield is realistic Back in 2003, the average price point demand for RFID tags, a couple of if all process and material parameters & of RFID tags was ten times this value. alternative approaches (for example, are well controlled; and 99.5% should But now there is evidence of a new Fluidic Self Assembly and Vibratory be a reachable in the mid-term time- C A R D S wave of technology enablers that will Assembly), referred to also as “vision- frame. drive the production cost of RFID ary approaches” in this context, seem tags down. The focus here is the cost to offer solutions to the challenge The following simple calculation of chip assembly. (Figure 1). helps explain the cost pareto of yield loss. A complete chip assembly line The most significant advantages of including adhesive application, flip S M A R T General technology approaches to conventional and advanced die bond- chip attach, curing, testing and slit- RFID chip assembly ing equipment are that they are based ting is currently available at a mar- on mature technologies offering low ket price of less than 1,000,000 US$. The method of assembling a chip investment risk and high-yield op- With 7400 hours of production time onto a substrate has been developed eration. To give a better idea of these a year, a line throughput of 10000 OnBoard Technology September 2006 - page 46
  2. 2. ASSEMBLY UPH, and 5 years depreciation, the process step, introduces a flip chip consumes less material. Bearing in cost of chip assembly is 0.27 US cents interposer. In a subsequent step, at mind that NCP is less expensive than per tag. Assume a further 15 US cents very high throughput, the interposer ACP, the ideal interconnection tech- for the costs of an RFID tag and con- is mounted on the antenna, which nology seems to be dispensed NCP. sider a yield loss of 2%. Then the cost can be done by crimping. Indirect as- The advantages of NCP are: easy, fast, of the yield loss is 0.3 US cents, which sembly is advantageous especially for low-cost processing; high-quality in- exceeds the depreciation costs of the manufacturers who have no previous terconnections; no additional under- equipment. This calculation should experience with bare-chip processing fill necessary; few process steps; low help to illustrate the high value of and do not want to invest in gather- epoxy cost; compliancy with low-cost high-yield assembly equipment. ing the necessary know-how. At the substrate materials; compatibility same time, investment costs for the with reel-to-reel applications. follow-up assembly step are signifi- Direct and indirect chip assembly cantly lower. The trade-off here is The feasibility of high-volume NCP higher packaging costs. Also, crimp In the manufacturing of RFID tags, connection quality is still a very con- there is a basic distinction between troversial matter. Therefore, alterna- direct chip assembly and indirect tive methods such as soldering or ad- chip assembly (Figure 2). hesive interconnection methods are being used. Or the connection is fa- In direct assembly, the chip’s bumps cilitated after the interposer is glued are positioned and placed directly onto cardboard with the antenna be- onto the antenna connections by ing applied in a second step via the means of flip chip technology. The printing of conductive ink over both key advantage is a lower packaging cardboard and interposer. cost, because this requires fewer proc- ess steps and consumes less material. Figure 3 - Optimal machine concept requires However, despite the large antenna Flip chip interconnection well-balanced dimensions of the bonding area pitch, high throughput rates neces- technology and process flow sitate negligible indexing timing. production on flex substrate has also Thus – with a conventional flip chip The type of interconnection method been demonstrated in smart card bonder approach - the technology also has to be chosen. Figure 4 shows production, where the high reliabili- is not without its challenges, which the various options that are generally ty requirements of the flip-chip pack- will be better manageable the bigger available. In terms of RFID chip as- ages has been proven. In addition, the bonding area of the antenna web. sembly, the most promising technol- the non-collapse soldering process is However, the trade-off with a bigger ogies are NCA (non-conductive adhe- used for RFID chip assembly. For ex- bonding area is longer travel time sive) and ACA (anisotropic conduc- ample, the GBS process (gold bump for chip transport. A well-balanced tive adhesive). Since the pre-applied soldering) benefits from the low dimensioning of the bonding area epoxy also functions as an underfill- bumping costs of the gold bump and will result in a successful machine er, the interconnection technology is the high quality of the intermetallic concept (Figure 3). very cost-effective, especially in paste interconnection due to soldering. form (NCP, ACP). The epoxy can be As an alternative, various manufac- applied by screen printing or dis- Collapse processes are not feasible turers employ indirect RFID chip as- pensing, whereby dispensing saves because of higher bumping costs, R F I D sembly. Indirect assembly, as a first further costs because the process thicker package profiles, and the Figure 4 - General flip chip interconnection options Figure 5 - Types of chip assembly process flow & C A R D S S M A R T OnBoard Technology September 2006 - page 47
  3. 3. ASSEMBLY Table 1 – Possible cost breakdown for a 5 cent RFID tag Table 3 - The cost budget per RFID tag for the inlay assembly process illustrated in Table 2 Table 2 - An assembly process for 4” x 1” UHF inlays Table 4 – Mid-term RFID tag cost budget for the inlay assembly process illustrated in Table 2 does not need years, a widely shared view of the an inline oven RFID label’s cost budget is shown for final curing, in Table 1. or the place and cure flow, which According to semiconductor manu- eliminates the facturers, RFID IC manufacturing Figure 6 - Throughput of a heated press for parallel lamination in the for less than 2 cents will be realis- lamination of dies heated press. tic for high volumes. Also antenna manufacture and label conversion need for an extra underfill step, un- The place and laminate flow is very technologies are developing toward less pre-applied underfill is used. The widely used in RFID inlay production a cost target of less than 1 cent each. ICA process is also inferior because it for NCP and ACP packages, but has It is beyond the scope of this article requires the extra underfilling step. its trade-offs in line throughput. Fig- to give a detailed picture of the cost Ultrasonic (US) and Thermosonic ure 6 shows the achievable through- map for all combinations of techno- (TS) processes require a mechani- put of a heated press as a function of logical approaches for inlay assem- cally stable substrate, which is not the number of dies per index (to be bly. Instead the cost budget for a par- easy to achieve with low-cost mate- laminated in parallel) and the cur- ticular turnkey assembly solution is rials like PET at thicknesses of 50µ ing dwell time, assuming 2 seconds presented, which will definitely be a and below. for the non-productive time during benchmark for all winning technolo- index, open and close of the heated gies in the future (Tables 2 and 3). press. R F I D Chip assembly process flow The depreciation costs are based on Assuming 7500-10000 UPH for a flip 5 years and 7400 productive hours As a final consideration, the various chip bonder, depending on the an- per year; operating costs are based options of process flow for chip as- tenna size, the line UPH can essen- on 30$/h including environmental sembly must be understood. The tially be bottlenecked by the heated costs, where each operator can serve & classic flow, which has been used press, especially for production of 5 lines. Yield loss is based on 98% for NCP production, starts with big UHF inlays. This is one major yield and 15 cents inlay costs. C A R D S epoxy application (screen printing reason why place and cure assem- or dispensing), followed by the pick- bly solutions, which are standard for Based on a 2-year-outlook, the au- flip-place (die attach). Curing takes soldering processes, are in focus for thor proposes that equipment suppli- place in a two-stage process: parallel future investments. ers will be able to double the UPH of lamination of a couple of dies in a the assembly lines at constant equip- heated press station (pre-curing) and ment costs, and the yield should be S M A R T final curing in an inline oven. Expe- The 5 cent tag increased to 99.5% with inlay costs rience has shown that the number dropping below 5 cents. This results of process steps in the classic flow Coming back to the question of in the mid-term cost budget bench- can be reduced by going with either whether a sales price of 5 US cents mark for RFID inlay assembly shown the place and laminate flow, which will be realistic in the next few in Table 4. OnBoard Technology September 2006 - page 48
  4. 4. ASSEMBLY Smart Labels, Tickets And Luggage Tags On One Machine ing pre-selection processes. Within a fraction of a second the SL/ST can replace a defective RFID transponder before it is attached to the product. The machines also guarantee the highest produc- tion throughput with standard transponder mate- rial. 1% rejects in the transponder reel reduce the output of the machine by less than 1%. From the extremely versatile and fast single-track version to the high-capacity four-track version, Melzer caters for every volume requirement (10,000 to 40,000 pph or max. 65 m/min to 260 m/min respectively). Melzer, which celebrates its 50-year anniversary in 2006, is also specialised in machinery for inlays based on wire embedding technology. The compa- ny provides customised solutions when standard Melzer offers Smart Label/Smart Ticket (SL/ST) machines do not fit the needs of the customer. production line for smart labels, smart tickets and smart luggage tags; laminated contactless cards; Melzer maschinenbau eGovernment products (inlays for ePassport holder Ruhrstr. 51 – 55 pages or eNID cards); and smart cards and dual in- 58332 Schwelm terface cards. Among the features of the machine is Germany the ability to automatically sort out rejects in the Tel. +49 2336 9292 transponder reel fully automatically at the highest Fax +49 2336 9292 85 speed possible. No need for expensive pre-selected transponders or time, space and labour demand- Batch Counting Option Added To Automatic Card Inspection System System, which enables users to automatically batch approved and/or reject cards into groups of a se- lected pre-set size. The M930 is a smart self-learn- ing system that trains on the desired card image in minutes and totally bypasses limitations in operator training. It is a 100% reliable system according to R F I D the company and fully inspects both sides of cards at a rate of 36,000 cards/hour. The Batch Counting Option will make continuous operation of the sys- tem possible, without delays for packaging finished and inspected cards, and is available with all new & systems and as a retrofit add-on. C A R D S Spartanics 3605 Edison Place S M A R T Rolling Meadows, IL 60008 - USA Tel. +1 847 394 5700 Fax +1 847 394 0409 Spartanics announces that a Batch Counting Option has been added to their 930 Plastic Card Inspection OnBoard Technology September 2006 - page 49