By Gael Coron, System Design-In Tools Manager

Effort, low risk, low cost, low power, small factor form and level of re-use are the typical parameters on project managers hit lists when it comes to RF development. For wireless applications, project managers can find many RF modules on the market that can solve such issues. But is this really the case?

Less and less industries are ready to sacrifice one of these features. The competition is more and more intense and final resale price is lower and lower; in contrast the number of features in the final product is increasing.

Many developers would appreciate having all these features in parallel however most fear that they have to sacrifice one or more parameters so as not to endanger the design goals and time frame.

This paper deals with:

• How RF modules can accelerate the Marketing Introduction Plan?

• Whether RF modules can minimize the risk and the effort?

• How RF modules can be cheaper than a proprietary RF design?

Firstly we need to define what a Drop-in RF module is. The Drop-in RF module is an: integrated module that eliminates the needs for external components and physical adjustment, and is also fully tested and sometime certified as FCC and or ETSI, greatly reducing. time to market. Obviously the complexity of using a Drop-in module depends on the integration. Drop-in RF module manufacturers have to find a trade-off between being easy to use, functionality and cost.

1. How Drop-In modules can accelerate the Marketing Introduction Plan?

Having the right product at the right time is one of the most important factors of a success story. Business cases, marketing analysis reports or sales feedbacks are useful tools to define the new product specifications. However, even if the product offers advantages regarding the competition, it will never be a success if it comes to market too late. A typical RF development takes more than 9 months just to pass certification and launch the production. So the time to market is a critical point.

The strategy from some companies, mostly based in the Far-East, is to develop a wireless product and to test the market with it. If it corresponds to a need they will produce it in high volume to reduce the manufacturing cost for a short time of period. But with these conditions, the investments on time, resources and cash must be limited. If the product does not meet the success, the loss will be negligible.

The use of Drop-In RF modules can be one of the best alternatives. With most of them, no external components (excluding the antenna) are required making integration into the final product straightforward.

Depending on module manufacturers, the integration level differs with corresponding strengths and weaknesses. The main differences between Drop-In modules are to embed or not:

• An antenna switch
The antenna switch provides one port antenna connection. However, the insertion loss can be few dB. That means that the data communication range theoretically decreases by 2.

• A microcontroller
Two different approaches exist with a microcontroller on board. In the first one, the microcontroller is programmed with a radio protocol to target a specific application (for example radio modem). In this case the integration is high but the flexibility does not exist. It's impossible for the developer to change the Radio protocol, the only way to use these types of modules in different final applications is to adapt/develop the high level protocol accordingly with the low level protocol embedded into the drop-in module.

The second approach used by other manufacturers is to provide a standard 8 bits microcontroller connected to the RF; however no radio protocol is embedded. This solution seems to be a good compromise between integration and flexibility - is it really the case? For developers this approach has real disadvantages; first, they must develop software according to the connections between the microcontroller and the RF. In some cases, some microcontroller resources can be used because of this connection (for example interrupts). Companies prefer to always use the same microcontroller series, they have already invested money in development tools and do not want to re-invest in new tools.

The form factor is also a key feature to accelerate the integration of the RF module into the final product. Market trend is to have products as small as possible. In some cases project managers are ready to sacrifice performances for size. The wireless headset market is one of them. The constraint on the ear-piece imply the need to have a compact radio module, the performances can be secondary.

One way to reduce the size and to facilitate and accelerate the integration is to remove pins connector and to replace it directly by soldering pads (or half pads). In this way a few square millimeters can be saved. In addition this solution offers advantages when use the drop-in modules with SMT (Surface Mount Technology) components and avoids the need for hand insertion.

To reduce the Product Introduction Plan, the project manager can also take into consideration the manufacturing time (assembly and test). For this reason, Drop-in module manufacturers have made a big effort with the packaging and delivery form. Tape and real or blister tray are the both most popular delivery form because they are suitable for high-speed automated manufacturing processes.

2. Do Drop-In modules minimize risk and effort?

Risk and effort are two important factors to evaluate before starting the development of new designs. A new application with an RF link can be very risky and with a development time of 9 months minimum just for the radio communication. Also due to differences between Europe, US and the Far-East the risk and the effort for development can be multiplied by 2 or 3 if the final product is to be sold worldwide. For example, due to difference of the unlicensed frequency bands between Europe and US (433MHz / 868MHz in Europe and 915MHz in the US).

Before starting, the project manager needs to evaluate the global risk of the development but also the risk per function. RF development is far from straightforward even with manufacturer's guidelines (Integrated Circuit manufacturer). Typical application, reference designs, application notes or technical notes exist to help RF engineers to develop a wireless link but most of the time are not sufficient or far from the final application. Drop-In modules are one of alternatives found by chip manufacturers that actually reduce the customer support requests and accelerate the design flow. From the Project manager's point of view the Drop-In solution is a good way to reduce the development risk and focus on the actual application (the know-how of the company).

In addition the RF board must respect the country regulations and depending on the result it may be necessary to have a whole new design just to pass the certification. Depending on the Drop-In module the risk can be decreased even further by having a certified module (ETSI for Europe, FCC for the US or SDT for Japan), in this case the risk has been taken by the Drop-in manufacturer and not by the user.

One of the main problems for RF developer is having a complete RF laboratory as this is a huge investment. Certification laboratory and Spectrum analyzer, network analyzer and FM generator are the minimum RF equipment for any wireless development. Companies who are not really involved in RF business and where the "know how" is not in the RF do not tend to have fully equipped RF laboratories with for example a network analyzer. In these conditions, the developer can spend a lot of time and effort trying to optimize the RF link, keeping in mind the objective which is to pass the certification. Most of the time this objective is a real nightmare for people who do not have an dedicated RF background and experience with the relevant regulations. One again, the Drop-In module is a way to avoid bad experiences with RF by offering a complete and optimized solution. It can be optimized in cost, in performance or cost and performances. However, the important feature is that the Drop-In module is optimized in R&D effort, the effort is focused on the application and not in the RF which is a "simple" link between two or more systems as a communication bus.

3. How Drop-In modules can be cheaper than a proprietary RF design?

As with any product, the fabrication cost depends on the Bill of Material (BOM), the subcontractor (worker cost) but also on the annual or monthly volume. Drop-In modules in contrast to Evaluation modules stand for straightforward integration into high volume applications. So, during the Drop-in module development all the parameters have been taken into account in order to achieve the goal.

The Bill of Material needs to be optimized to have the minimum number of components but also to calculate the ratio between components. For example a capacitor is 10 times cheaper than an inductor. Optimizing the component values is also a good way to reduce the BOM cost; it's cheaper to use the 3 inductors with the same value than to use 3 inductors with 3 different values. It's also easier for the logistic department to manage one part instead of three parts.

Forgotten most of time,, the assembly subcontractor has an important role during the development and needs to be selected before the start of the development. Assembly are responsible for purchasing the parts and so can give some advice regarding cost and lead time. For example, in china 0603 components are cheaper than 0805 components. So if the Drop-In module is dedicated production in the Far East it's better to use 0603 components.

RF tests also have a cost and need to be taken into consideration. The assembly manufacturer or the test subcontractor can help by giving the best way and the faster way to do that. The investment on test set-up depends on the volume. A test time of 4 seconds per module is more expensive (NRE cost) than a test time of 70 seconds. But if the volume is important the investment on the test set-up becomes advantageous. The worker cost per module decreases.

RF modules based on the same IC (Integrated Circuit) are all similar and the price will depend only on the annual volume. A Drop-In module from IC manufacturers will be a low cost solution, their objective is to sell transceiver ICs and the Drop-In module is a way to do it. The margin will be on the IC and not on the module even more so if they offer the design for free. By adding all the purchase orders from all their customers they can increase the volume and so reduce the manufacturing costs.

This is all far less costly than using an in-house proprietary RF design where test and assembly costs must be borne, as well as development and certification costs.

The drop-In module seems to have a place in manufacturing and is a good alternative to expensive, long and complicated development of wireless applications. This indicates that the use of a drop-in module can be a better solution than a new design for RF applications destined for high volume production as well as trial runs.

XEMICS is located in Switzerland. Phone: +41 32 720 5192; Fax: +41 32 720 5770;