Power-up Bind: At the highest level power-up bind just means using the power-up process as a replacement for pressing a button.
Description: Every time power is applied to the unit with power-on bind implemented, it will enter a bind sequence. In some cases, such as those when power is infrequently removed, this will not appear substantially different from what has been discussed under 2-way button bind. In cases where it is likely for power to be cycled frequently this method is either not acceptable or there will have to be some changes to prevent it from becoming intrusive to normal device operation. The likely change is to reduce the timeout as described above. Power-up bind is typically implemented on one side only, with the other side using a button. Generally the timeout will also be kept small. It is possible to use power-up bind with a long timeout if power is rarely cycled. Examples are keyboard and mice for desktop users (Cypress battery life is typically a year or more), sensors in a building control network, or bridges built into some embedded systems. Obviously there are multiple combinations of power-up on one side or the other, or both, and short or long timeouts. The system architect must carefully consider the behavior that the user will encounter to determine if power-up bind choices are acceptable.
Example 1: Power-up bind on USB bridge, button bind on device. The bridge, since it is USB based, will see frequent power cycles when it is removed from the PC, or when the PC shuts down or hibernates, therefore it will have a short timeout of ~1 second. This is generally not noticeable when compared with the USB enumeration process. The device will use a standard button. To bind, the button on the device is pressed first, and then the bridge is inserted into the USB port. A bind channel subset and reduced PA are also used to generally increase robustness.
Example 2: Power-up bind on building sensor, button on bridge Sensors would typically have batteries inserted once every few years. A long timeout on the device side is used so that the basic 2-way button bind behavior is preserved. The long timeout would not be intrusive to the device operation. The user can start the bind process on either side: press the button on the bridge first and then insert batteries into the sensor, or the reverse. A bind channel subset is used for robustness, and a higher PA is used in case the bridge is a substantial distance away.
Example 3: Power-up bind on mouse/keyboard, button on bridge.This is somewhat less likely, but presents an opportunity to save cost by removing the buttons on the mouse and keyboard. Those that are not typically powered down can probably use the method described in Example 2. Others, particularly those intended for laptop users, have on/off switches to prevent inadvertent activation during transit. This power-up bind method employs a very brief bind sequence on power-up of the device (1 second or less) and a standard button on the bridge with a long timeout. For general use the ~1 second timeout should be short enough that it is not inconvenient when powering up the product, however it does slightly increase the power consumption of the device in cases where power may be cycled frequently. To bind devices the bridge must initiate the bind process, then batteries are inserted into the mouse or keyboard, or the power switch is cycled. If both devices are being bound the process is repeated for each one.
■ Using power on to initiate bind on the device side can eliminate the need for one or more buttons, thus saving
■ This process is not necessarily intuitive for the user, therefore it needs to be described in user documentation (which users do not always read).
■ There is a slight potential for increased technical support calls for users who do not understand the power-on behavior (Short timeouts require sequencing. Long timeouts make the device appear inoperable until the timeout is
■ It potentially increases power consumption since the process repeats whenever power is cycled