Spanish Italian
17447 Users    

Operation and Characteristics of the Digital Pot Resistor Network

  Download PDF version of the Article

A digital potentiometer device has three main functional blocks. These are the serial interface, the system control and the resister network. Today we will be focusing on the resistor network.

The resistor network has 3 terminals or connection network can be used in one of 2 applications configurations. These are as a potentiometer or also thought of as a voltage divider. And a rheostat or also thought as a variable resistor. Now depending on how the device is packaged, not all the terminal may be available to the outside world. That is in a rheostat configuration, only the wiper and one of the terminal, either terminal A or terminal B, need be available. While in a potentiometer configuration, terminal A or terminal B could be tied to one of the power rails, either VDD or VSS. In here configuration it is useful to understand the implications of the voltage and temperature to the circuit characteristics. The digital potentiometer’s resistance is the resistance between terminal and terminal B. This is also called the RAB resistance. The resolution of the potentiometer devices determines the number of position or steps on the resister network that the wiper can connect to. Therefore, 6 bit resolutions gives you 64 positions and 8 bit resolution gives you 256 positions. The step resistance is the resistance between one position and the adjacent position on the resistor network. This value is the total resistance RAB divided by the number of resistor in RAB. We call the step resistance RS. Now the number of resistors is related to the resolution of the digital potentiometer. But the exact number depends on how the wiper connections are implemented. So where the variable N equals the number of bits of resolution, if the zero scale position connects to a terminal and the full scale position connects to a terminal, then there are one less than two to the N positions. If the zero scale position connects to a terminal and the full scale position connects to a resistor RS, then there are two to the N positions. The resistor RS can be on the zero scale position instead of the full scale position. Lastly if the zero scale position connects to a resistor RS, then the full scale position connects to a resistor RS. Then there are one more than two to the N positions.

In the MCP401X and MCP402X devices, the wipers can connect to either terminal A or terminal B. So there are 63 step resistors. This allows full scale and zero scale connections. In the MCP 41XXX and MCP 42XXX devices the wipers cannot connect terminal A. There are 256 step resistors.

The RSBAS and therefore RAB values are stable across voltage and temperature due to their silicon implementation. There is some minimal resistance due to the construction of the pad circuitry such as EST structures. Also since the RSBAS is uniform in their layout and die placement, they will track each other as the voltage and/or temperature change. In a voltage divider configuration, any drift of the RSBAS value will occur in the same direction magnitude, so the voltage at the wiper is not affected.

The wiper switch is like a big analog mux. Each date of this analog mux will have slightly different characteristics. This causes a difference of resistance between any two wiper switch positions. The characteristics of these switches is much more sensitive to voltage and temperature than the RSBAS resistance. So looking at the wiper resistance with respect to voltage, with any voltage range, the change in resistance will be linear relative to the change in device voltage. At some voltage level, as the voltage decreases, the resistance characteristics of the switches will become non linear and the resistance will increase exponentially. This is why the NCP 401X and MCP 402X devices show larger rheostat INL and DNL specifications at 2.7 volts than at 5.5 volts. This is related to the operational characteristics of the wiper switch devices at the lower voltage. All the wiper switches will start to increase non-linearly at about the same voltage.

So now we will look at some characterization graphs. The 2 graphs on the left, we are showing the change of the wiper resistance with respect to the operational voltage of the device and the wiper position. Choose any wiper settings and points typically work best and compare the 5.5 RW wiper resistance to the 2.7 volt wiper resistance. The resistance scale is on the left side of the graph. What also occurs at low voltage is that the wiper resistance becomes more dependant on the wiper position in the circuit. The wiper resistance increases as the voltage delta between the resistor network node and the voltage on the analog mux switches becomes small. So that the switch is not fully turned on. The wiper resistance curve would look different if terminal A was at VDD divided by 2 while terminal B was at VSS. In this case the higher value wiper codes would have the higher wiper resistance.

The top right graph shows the effects of voltage and temperature on the RAB resistance. Notice that the change in R is minimal. The bottom right graph shows the monotonic nature of the device.

The 2 graphs on the left show how the wiper characteristics for the 2.1 kohm device when in rheostat mode. In this mode the change or variation of the wiper resistance has a greater effect on the INL and DNL of the device compared to the potentiometer mode since the error is added right into the RAW or RBW resistance. In potentiometer mode, the variation of the RW resistance is not so important since the device is acting as a voltage divider. SO the error is determined by the RSBAS errors.

In the 2.1 kohm device the step resistance is very small, approximately 33 ohms. So the variation of the wiper resistance can be much larger than the step resistance values. That is why the INL error can be several LSPs. When looking at the device with a much higher RAB value, for example, 50 kohm, the step resistance is much larger, approximately 790 ohms. So even though the wiper resistance value has a large variation it is still much smaller than a single resistive step. This means that the errors of INL and DNL can remain less than one LSV. So the effect pf the wiper resistance variation on INL and DNL decreases as the total resistance RAV value increases.

Some devices such as the MCP 41XXX and MCP 42XXX offer a shut down feature that will disconnect some terminal pins from the internal resister network. To disconnect terminal A or terminal B an analog switch needs to be placed between the resister network and the device pad. This analog switch will have similar characteristic issues as the analog switch used in the wiper. So the variation of the switch on terminal A and/or terminal B add to the variation of the total RAB resistance. The variation has voltage and temperature implications as discussed on the previous slide. While the error that may occur in the system with respect to the operation of the device is important another characteristic is important as well. That is the operation is monotonic. Being monotonic ensures that the wiper code value increases the resistance increases as well. For a digital potentiometer, it is analogous to a ADDs no missing code. Figure 211 shown previously on slide 5 shows that the wiper setting is increased, the RBW resistance increases. There are no points where the resistance decreases for an increase in the wiper setting. This is an important characteristic especially in cases where the INL error becomes larger than 1 as shown in figure 29 on slide.

I just had a opportunity to

I just had a opportunity to go through your blog and found it very informative . I represent
a leading B2B directory www.jazdelectronics.com, web built on web 2.o with a strong

functionality's which are very unique , blogs related to your theme on submission in

future will get guaranteed traffic and visibilty, i would request you to visit our site to get an understanding of value proposition which we have to offer .. at the same time I would also request that you approve the blog and give us a link . My contact id is

jazddomains@gmail.com.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

More information about formatting options

CAPTCHA
This question is for testing whether you are a human visitor and to prevent automated spam submissions.
2 + 9 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.

Who's new

  • JM
  • samsilva77
  • araghube
  • stoll
  • mt
  • orionkw
  • pulper
  • mauriss
  • jbares
  • christiank79

Who's online

There are currently 0 users and 45 guests online.