Analog design engineers frequently setresistances and voltages to make analog circuits meetspecifications. By using potentiometers to control the parametersof circuits, engineers can calibrate or vary critical functionssuch as the output of a laser diode in a fiber optic communicationscircuit, the contrast of an LCD, or the position of a motor in aprocess control system.
Potentiometers come in two types: mechanically controlled anddigitally controlled. Mechanically controlled potentiometers, orpots, are adjusted by physically turning a screw. A digitallycontrolled potentiometer (DCP) is adjusted by giving it a digitalvalue that it converts to a wiper position. In many applications, aprocessor or digital controller monitors the analog function'soutput and calculates the digital value that is applied to theDCP.
Figure 1: A mechanical potentiometer is athree-terminal device that functions either as a two-terminalvariable resistance or as a three-terminal resistive divider. Adigitally controlled potentiometer (DCP) has the same threeterminals, plus additional terminals for digital control of theDCP's functions.
The terminals of a mechanical potentiometer are called CW(clockwise), CCW (counter clockwise), and wiper. The correspondingDCP's terminals are called VH (voltage high) orRH (resistance high), VL (voltage low) orRL (resistance low), and the wiper VW (wipervoltage) or RW (wiper resistance). While a mechanicalpot is a three-terminal device, a DCP is an integrated circuit witha minimum of eight terminals.
As mentioned earlier, a screw controls a mechanicalpotentiometer's wiper position while the wiper position of a DCP,such as Xicor's XDCP products, is digitally controlled. While amechanical pot's wiper position is remembered mechanically, a DCPuses a digital memory to store the digital value that is decoded todetermine the wiper's position. Therefore, in addition to resistivematerial and a wiper, a DCP also has a memory and controlcircuit.
Figure 2: Like a mechanical pot, a DCP consists ofan end-to-end resistance and a wiper. In addition, it includes amemory to store the digital value that determines the wiperposition, and a control circuit that decodes the digital value andapplies it to the wiper. The DCP shown here uses a three-wirecontrol bus where the digital value is changed by incrementing ordecrementing the wiper rather than loading a complete value.
The DCP control and memory section is implemented in CMOS and isbiased with a 3- or 5-V digital or logic supply. The device iscontrolled through one of three serial buses:
- Two-wire (similar to I²C)
- SPI (Serial Peripheral Interface)
The control signals for the three-wire bus (Figure 2 ) areUp/Down, Increment, and Device Select. The Up/Down control input isa level-sensitive signal that establishes the direction of wipermovement. The wiper is moved on the falling edge of the Incrementcontrol input in the direction established by the Up/Down signal.The Device Select control input is like an address line and enablesor disables the device.
The control inputs for the two-wire bus are Clock (SCL), abi-directional Serial Data line (SDA), and Address lines (ADDR).The control inputs for the SPI bus are Clock (SCK), Serial In (SI)and Serial Out (SO) data lines, and address lines (ADDR). Thetwo-wire and SPI serial interfaces have protocols that areexplained in the data sheets.
A DCP uses a series resistor array and switches to implement theRTOTAL resistance and wiper position (Figure 3 ).Polycrystalline resistors are connected in series between theRH and RL terminals to form a series resistorarray. Solid-state switches, implemented by NMOS or CMOStransistors, are connected at each end of this resistor array andto points between the resistors. The other end of each switch istied to the wiper output.
Figure 3: A DCP uses a series resistor array andmultiple SPST switches to implement the potentiometer function.When a digital value is loaded into the DCP's memory, it is decodedto select one of the switches. That switch is closed, thusestablishing the wiper position. Only one switch will be closed atany one time.
When a three-wire interface is used, the wiper position can bemoved up or down one switch position by incrementing ordecrementing the value in the DCP memory. When using one of theother interfaces, any switch can be selected by loading acompletely new digital value into the memory.
A DCP can be used in a circuit as either a three-terminal deviceor as a two-terminal device. The most common three-terminalapplication is as a voltage divider circuit (Figure 4 ). Plus and/orminus voltages are connected to the DCP VH /RH and VL /RL terminals. Adigital value selects a wiper position by closing a specificswitch, and thus selects a voltage on the wiper terminal.
Figure 4: The most common application for a DCP inthree-terminal mode is as a voltage divider. The wiper selects avoltage somewhere between the voltage on the lower terminal and thevoltage on the upper terminal. This application can be used as asimple digital-to-analog converter (DAC).
In many applications, the DCP's three-terminal voltage dividermode can be used as an inexpensive digital-to-analog converter(DAC) because it performs a digital-in, analog-out function.
Figure 5: The DCP can also be used as atwo-terminal variable resistance device. A simple example of atwo-terminal application uses a DCP to program the current throughan LED or light emitting diode. The DCP functions as a variableresistor and establishes the current through the LED.
The two basic applications illustrate the use of the digitallycontrolled potentiometer in a digital-to-analog voltage circuit andin a digital-to-analog current circuit.
A Xicor XDCP is a mixed-signal device that includes internalregisters and data paths in addition to a resistor array andswitches (Figure 6 ). On its digital side, an XDCP is a device thatcan be programmed using its instruction set.
Figure 6: An XDCP includes an interface and controlsection on the input side and the series resistor array and wiperon the output side. Between these two are data registers and datapaths. The WCR holds the digital value that determines wiperposition, while data registers DR0 to DR3 may hold alternativewiper positions or other data.
A digital value determines the position of the wiper(VW or RW ). That value is stored in avolatile register called the Wiper Counter Register (WCR). XDCPswith a two-wire or SPI bus also include four nonvolatile DataRegisters (DR0-3) that can store data or additional wiper settings.The Wiper Counter and Data Registers can be programmed from thebus, or data can be transferred between the registers through thedevice's instruction set. Normally, Data Register DR0 stores thewiper setting for the power-up condition.
XDCPs with a three-wire interface have a single internalnonvolatile register that stores the wiper setting for restorationduring power-up operations. The wiper position in a three-wiredevice is changed by incrementing and decrementing. The power-upvalue, which determines the starting point for the wiper, is loadedusing the input control signals.
Figure 7: XDCPs with either a two-wire (I²C)or SPI serial bus have a small but powerful instruction set thatcontrols specific device functions. The instructions control theflow of data, the increment/decrement feature, and some specializedcommands. For example, the Global Transfer command transfers databetween registers in XDCPs with multiple potentiometers, and theWrite in Progress (WIP) instruction monitors the completion ofnonvolatile writes.
XDCPs are frequently used in circuits with active devices suchas operational amplifiers and comparators, so Xicor offers thesedevices integrated into a single chip (Figure 8 ). Althoughintegrated in a single chip, these devices are not internallyconnected. The terminals of both the potentiometer and the op ampare brought out to accommodate all possible configurations.
Figure 8: Xicor integrates XDCPs and active devicessuch as operational amplifiers and comparators on a single chip.The devices in the chip are connected externally to accommodate allpossible configurations.
The op amp is powered or biased with the externally connected V+and V- analog voltage supplies. The WCR programs the location ofthe wiper of the potentiometer and the Analog Control Register(ACR) programs features of the operational amplifier.
When selecting a DCP for an application, the key deviceparameters are number of taps, end-to-end resistance, maximumvoltages on the potentiometer pins, wiper resistance and current,power rating, resolution, noise, linearity, and temperaturecoefficient.
Table 1: This table lists critical parameters for atypical XDCP device. Key device parameters are number of taps,end-to-end resistance, maximum voltages on the potentiometer pins,wiper resistance and current ratings, power rating, resolution,noise, linearity, and temperature coefficient.
XDCP taps vary from 16 to 1024, and reflect the resolution ofthe device or its ability to discern 1 of n. The end-to-endresistance (RH to RL ) is RTOTAL and comes in 1-k to1-Mvalues. The voltage VCC , 3- to 5-V, provides the voltagebiasing for the digital control and memory section, and V+ and V-provide the voltage biasing for the analog section. The voltagesVTERMINAL are the maximum voltages that can be appliedto the potentiometer pins during operation.
Wiper resistance, nominally 40-, modelsthe channel resistance of the MOS switches used to connect thewiper terminal to a node in the resistor array. The wiper currentspec, 1-mA to 3-mA, limits the maximum amount of DC allowed throughthe wiper switches.
Absolute linearity describes the actual versus expected value ofthe potentiometer when used as a voltage divider. It is accuratewithin one least significant bit (LSB) or minimum increment (MI).Relative linearity describes the tap-to-tap accuracy and isguaranteed to be 0.2 of an LSB or MI.
As with virtually any analog device, XDCPs have theirlimitations. And, as with virtually any analog device, theselimitations can be minimized by proper circuit applicationstechniques.
If a designer needs to vary a parameter in an analog circuit ina mixed digital/analog system, the typical choices are a digitallycontrolled potentiometer, mechanical potentiometer, or a morecomplex circuit using a digital-to-analog converter (DAC) and anEEPROM. Here is a summary of DCP advantages when compared tomechanical potentiometers and circuits using DACs. First, considerDCPs vs. circuits using DACs:
- The DAC provides circuit level variability (DC voltage-only)while the XDCP provides component level variability(resistance-universal)
- A circuit using a DAC requires several components to accomplishwhat a DCP can do with a single componentthat's important forapplications where space is at a premium
- A DCP costs much less than the total cost of the componentsneeded to do the same job when using a DACthat's importantfor cost-sensitive applications.
Now consider a DCP vs. a mechanical potentiometer. A DCP and amechanical potentiometer share some characteristics, but differ inothers. Both can be set and will retain their settings even whenpower is turned off and back on again. Both will hold theirsettings for long periods of time. In other cases, however, theyare not equal.
- A DCP can be controlled by a microcontroller or microprocessorwhile a mechanical potentiometer must be set manuallya DCPdoesn't require the services of a technician every time anadjustment is needed. The result is a lower cost of testing andmanufacturing.
- A mechanical potentiometer may be modified by changes intemperature and by mechanical shock while a DCP has no moving partsand is much less susceptible to changes caused by environmentalconditionsand if temperature extremes should cause somedrift, the host system can detect it and make any necessaryadjustments immediately.
- The reliability of the electronic potentiometer is orders ofmagnitude better than its mechanical counterpart. The operation ofthe electronic potentiometer is orders of magnitude faster.
Clearly, analog engineers who design circuits for use inmixed-signal applications must include digitally controlledpotentiometers in their design tool boxes.