The high-density ASK (amplitude shift keying) environment is one in which there are so many transmissions occurring at a given frequency that they occasionally collide and thus fail to get properly decoded by the targeted receivers. There are a number of potential solutions, but in this discussion we are concentrating on limiting the transmitter On-period. This is accomplished by operating the encoder IC at the maximum data rate and by limiting the transmitter On-time via a monostable multivibrator. Obviously, this is not appropriate for critical applications, but provides an inexpensive solution for many non-critical applications. For a complete mini-system, the HT12E encoder must be matched with a HT12D decoder.
Datasheet link (http://www.holtek.com.tw/pdf/consumer/2_12ev120.pdf)
Key contact de-bounce multivibrator
The most significant problem occurs when the operator keys the encoder for an extended period of time. This essentially jams the frequency so that no other transmissions can be received. To prevent this from occurring, a monostable multivibrator keys the encoder for about 20mS. This limits the transmission to the minimum 4word data burst. To prevent the transmitter from being keyed by rapidly pressing the key button, another section of the monostable multivibrator prevents successive keying until a 1.2sec timeout period has elapsed.
The 74HC221 non-retriggerable monostable multivibrator is ideal for this function. Do not confuse this device with the 74HC123 retriggerable monostable multivibrator that has the same pin-out.
Battery life should be excellent as both the monostable and encoder are CMOS –neither draw measurable current when at rest. Limiting the transmission time also reduces battery load. As a result, the battery may never need replacement. The way the 74HC221 monostable works is that there is no voltage across the timing resistors when the device is at rest –without voltage, they cannot add load. Note that not all monostables work this way –some have significant standby current.
Data format /protocol
The Holtek device has a unique, non-standard data format and protocol. It is optimized for low cost, minimum component count and functionality. Conversely, the UART (universal asynchronous receiver /transmitter) approach requires a close frequency match. This requires expensive crystal oscillators etc. In the Holtek devices, transmit and receive frequencies are crudely matched –perhaps within about 10%. To make data synchronization reliable, each bit gets individually synchronized via timing from the leading edge of a positive going sync pulse that occurs immediately before each data state period. Data remains reliable up to perhaps 30% transmitter /receiver frequency mismatch. The down side is that this technique halves the data rate, or doubles the transmit period.
Each data word consists of a sync pulse and 12bits of address /control data. This is repeated a total of 3 times so that the data burst consists of (4) identical words. At an oscillator frequency of 5.5kHZ, this requires 45mS. Ideally, this allows for 22 transmissions per second. In a high-density environment, it has a practical limit of perhaps 5 transmissions per second in order to avoid excessive collisions.
The datasheet shows how the frequency is set via changing a resistor between 470K and 2M. The frequency is also a function of battery voltage, but there is sufficient latitude between a fully charged and a discharged battery to maintain satisfactory operation. For this exercise, I set the frequency to 5.5kHZ (max) via a 470K resistor. Battery voltage is 3V.
Other means of increasing the transmitter density
Initial experiments indicate that the oscillator may be bumped up to 22kHZ by reducing the oscillator resistor from 470K to 100K and that the frequency seems to track in a near linear fashion (note that the Holtek 12A device runs at 38kHZ). While this is a nice improvement, it is an out-of-spec condition and not all devices may function. Also, it is unknown at this point if the receiver can be tweaked to follow at this speed. So for now, this is ‘Twilight Zone’ speculation…
A more reliable means to increase transmitter density is to add additional frequencies and to color code units so that the transmitters may be matched with receivers. Since there are three or four standard frequency offerings, this is practical. Another means to marginally reduce environment density is to reduce range via reducing transmit power and /or receiver sensitivity.
My guess is that if a number of the above means are employed, the maximum number of transmissions may easily approach 20transmissions per second.
Replacing the HT12E /HT12D chip set with microcontrollers may increase this by an order of magnitude, but such requires significant programming and testing –not a simple solution.
Undocumented words and idioms (for our ESL friends)
Twilight Zone –idiom –a takeoff from Rod Sterling’s popular paranormal TV series of 50years ago