Advertisement

Blog

Potentiometers: Mechanical & Electronic, Part 2

In Potentiometers: Mechanical & Electronic, Part 1, I described the use of pots generically and hinted at some of the more common configurations with circuits. I now hope to cover the comparative advantages and disadvantages of both mechanical and electronic potentiometers.

There are many types of mechanical potentiometers varying in the size, technology, power dissipation, resolution, and thermal stability, but they all follow the same principle. There is a fixed resistor between two pins and a mechanical wiper that moves along the resistor element providing the variation in the resistance between a resistor end and the point of contact of the wiper. Gearing the wiper mechanism (resulting in multi-turn adjustments) allows for greater resolution in the adjustment. However, the point of contact of the wiper is a weak spot in the design. Wear and dirt can result in the contact going open circuit or introducing additional resistance. Older readers will remember the crackling sounds that could be heard on radios when there was a dirty volume control. I would be remiss if I didn’t add the aside that contact cleaner aerosol is often an effective remedy for this problem.

Because of their physical size mechanical pots typically have relatively large wattage ratings, although of course the higher the wattage dissipated, the worse the accuracy of the device because of the self-heating. Adjusting most mechanical pots is done either with a small screwdriver or fingertip control, but when you get to higher power devices like the Ohmite 210 series. The wiper consists of a clamp that fits over a tubular resistor and adjustment can be cumbersome and even dangerous if there is high voltage or the resistor is hot. Also pay attention to the wattage rating — to see what I mean, take a look at the restriction the datasheet link above. It stipulates:

    The stated wattage rating applies only when the entire resistance is in the circuit. Setting the lug at an intermediate point reduces the wattage rating by approximately the same proportion. Example: If the lug is set at half resistance, the wattage is reduced by approx. one-half.

I suspect that this is may be true for all mechanical pots.

Figure 1

A 4-20mA signal conditioner using two pots to interactively adjust the zero and span of the output. You can see the pots at the top of the circuit board.

A 4-20mA signal conditioner using two pots to interactively adjust the zero and span of the output. You can see the pots at the top of the circuit board.

The advantage of a mechanical pot is that it is non-volatile (poor joke, I know), but really, its setting does not change with power cycling. Setup may be intricate, but anyone intuitively knows how to adjust it. On the other hand, mechanized adjustment as part of a calibration process can be expensive and difficult.

The electronic (or digital) pot is not direct replacement for the mechanical pot. There are times that one can substitute, but there are times that it just won’t work. The electronic pots are electronic devices- they need a power supply and the signal/voltage that you are conditioning must fit within the power supply restrictions (although there may be some that allow signal operation beyond the supplies). Obviously the maximum power dissipation is much lower than the mechanical parts. Also the operating temperature range is much reduced compared to their mechanical brethren. Beware the gotcha on the wiper current. Often it must be limited to a very low value as per the specifications — forewarned is forearmed! On the plus side, the resolution is fixed and any setting is repeatable and even more importantly it allows for automation during calibration, although there will almost certainly have to be additional firmware to achieve this.

Figure 2

A true RMS current to 4-20mA signal conditioner with programmable maximum/minimum limit detection. The limits are controlled through a hand-held programmer that would be connected to the RJ 45 connector on the left hand side. The settings are established on two digital pots with EEPROM capability and up/down adjustment. Ironically the 4-20mA loop is calibrated by mechanical pots.

A true RMS current to 4-20mA signal conditioner with programmable maximum/minimum limit detection. The limits are controlled through a hand-held programmer that would be connected to the RJ 45 connector on the left hand side. The settings are established on two digital pots with EEPROM capability and up/down adjustment. Ironically the 4-20mA loop is calibrated by mechanical pots.

Being monolithic it means that you can have multiple pots on the same chip (although there is a possibility of interaction or even breakdown between them) and you can have additional circuitry to provide boosted functionality. The addition of an EEPROM allows the pot to be non-volatile and drivers allow for different interface protocols. Application of electronic pots can be extended well beyond the trimming capability of mechanical pots and start overlapping Digital-to-Analog Converter territory, depending on the designer’s imagination.

The wide selection also precludes me from trying to provide some kind of generic table to cover all eventualities, but as a guide here is a list of some manufacturers of electronic pots: Maxim electronic pots, Intersil electronic pots, Analog Devices electronic pots, Microchip electronic pots, On Semiconductor electronic pots, and Texas Instruments electronic pots.

Digital pots are made up of a series of resistors with analog switches that allow a particular tap setting to be selected. This allows a potentiometer configuration, but is restrictive since the current must flow through the resistor string and the wiper current (as mentioned above) is constrained by the limitations of the analog switch. I should mention that there is a way to make a true rheostat using a MOSFET with the gate being adjusted to regulate the on-resistance of the device. A more detailed discussion is beyond my self-imposed scope of this blog. However as far as I can tell, Advanced Linear Devices has extended this idea to a concept called EPAD (Electrically Programmable Analog Devices) and it claims it can be used as a trimmer substitute. You can find the data here and AN1108 covers the applications.

There are thousands of different potentiometers out there, but I know of one case where our subcontractor actually made their own high wattage potentiometer. You can see a description in my blog PCB Test Jigs & My China Connection. I hope you never have to resort to the same approach.

For a bit more information on the mechanical/digital pot comparison, see “Understanding and Applying Digital Potentiometers” from Analog Devices. Do you use pots? Have you ever replaced a mechanical pot with an electronic one?

37 comments on “Potentiometers: Mechanical & Electronic, Part 2

  1. antedeluvian
    August 18, 2014

    Beware the gotcha on the wiper current. Often it must be limited to a very low value as per the specifications — forewarned is forearmed!


    I forgot to mention that there is a series resistance associated with the wiper. It can be from tens of ohms to hundreds. That does not matter normally when used as a voltage divider, but in rheostat mode…

    I repeat  forewarned is forearmed!

  2. David Maciel Silva
    August 19, 2014

    Hello Aubrey,

    My experience with digital potentiometers was not the best of all … lol!

    But it helped a lot in developing an automatic calibrator, and not to be tweaking the voltage levels at each stage he was the best option.

    Spared some technical hours of calibration personnel would have to do everything manually.

    Basically the equipment generated a voltage reference and compared with an analog metering point equipment that was being calibrated …

    But to reach that point, it was not very easy.

  3. antedeluvian
    August 19, 2014

    I would be remiss if I didn't add the aside that contact cleaner aerosol is often an effective remedy for this problem


    The was featured as item #7 in my blog “Top 17 Helpful Hints for Constructing Electronic Systems“. There are some good hints in the comments as well.

  4. vasanjk
    August 19, 2014

    AK,

    It always is a pleasure to read your articles. My thoughts are like “Hey exactly my view''', “I too tried this once”, “Oh so simple yet I didnt figure it out” and so on.

    It is amazing that you reflect the most practical points with so less math, a rare combination and is a delight to read.

    I once replaced a mech pot with a digipot in an electronic pain killer device called TENS-Transcutaneous Electro Neuro Stimulator. You may recall my blog in MCC on this subject. The intensity was set through the POT by the patient just above the pain level and pain slowly fades away. It is all about how the patient feels and the analog pot provided her the much needed seamless adjustment. A digipot with a push button control did not provide such an experience and the fully digital model failed while the mech pot model is still being sold.

  5. Davidled
    August 19, 2014

    Changing volume or lighting is easily adjusted by digital potentiometer (DPT), compared with traditional potentiometer. DPT could provide a very flexible way depending on how user controls with microcontroller.

  6. antedeluvian
    August 19, 2014

    JK

    Good to hear from you even though you over-flatter me.

    I once replaced a mech pot with a digipot in an electronic pain killer device called TENS-Transcutaneous Electro Neuro Stimulator. You may recall my blog in MCC on this subject. 


    I do recall the blog, but not much of the nitty gritty.

    A digipot with a push button control did not provide such an experience and the fully digital model failed while the mech pot model is still being sold .

    Coincidentally I just bought myself a TENS machine. It certainly has no pots on it and I don't know how the adjustment is made internally. I can't see myself risking the investment to crack it open just yet- it is far too effective in treating some of my ailments.

  7. samicksha
    August 20, 2014

    I agree to your point here, Digital Potentiometer is more complex than mechanical potentiometer, although we have to admit that digital ones are more widely used for adjustments and calibration.

  8. RedDerek
    August 20, 2014

    @AK: There are thousands of different potentiometers out there, but I know of one case where our subcontractor actually made their own high wattage potentiometer. You can see a description in my blog PCB Test Jigs & My China Connection. I hope you never have to resort to the same approach.

    Reading this brought back to mind a project. I was out in the field doing some testing and our power regulator went out and parts were not easily obtained quickly. We needed to find some way to regulate the power to our dc load. We ended up finding a junk pile of heater coils that provided our solution. We just paralleled or series a set to set the correct power to our load. Fortunately the test was cancelled due to other equipment malfunctions.

  9. vasanjk
    August 21, 2014

    AK

    Am working on a medical device where they use a digipot chip. There are three sensors that are mux'ed with analog switches and fed to a single gain amplifier. The gain amplifier has the digipot in its feedback path and behaves as  non-inv amp. Works nice for different fixed gains for each channel. User can enter a calibration value for each channel and this value goes into the gain equation as an input.

    Works well. Gain can be altered on the fly without much hassles.

  10. David Maciel Silva
    August 21, 2014

    But really we have today, access to information with greater speed that makes things easier.

    I am not of that time in my country had to wait for the manufacturer to send the data sheet by mail ….
    But I experienced the “explosion” of the Internet, and as it became easier to develop …

    In the article below we have a good working demonstration of a digital potentiometer.

    http://arduino.cc/en/Tutorial/SPIDigitalPot

  11. Victor Lorenzo
    August 21, 2014

    I'm currently using the Arduino UNO R3 board for several small projects with low requirements in CPU and I/O performance. It is very usefull and we can find plenty of documentation and almost-ready-to-use sample code on the Inet. But we must check everything twise before copy-paste into our application.

    @Maciel, please correct me if I'm wrong.

    In the sample project (http://arduino.cc/en/Tutorial/SPIDigitalPot) the circuit is configured as a voltage divider using the 10kOhm part, with 5V applied between the A-B terminals, being A at 5V and B at 0V (GND).

    According to the datasheet, for the 10kOhm part the current at any terminal (IA , IB , IW ) must be kept bellow 11mA . This max value drops to 2.5mA for the 50 and 100kOhm parts.

    No LED part reference is given so we can assume the LED's forward voltage drop ranging from 1,8V to 3,3V. As the current limiting resistor is 220Ohm, and assumming a common red LED with a 1,8V forward voltage drop, the maximum LED's current (IW ) would be (5V – 1,8V)/220 = 14,5mA . This value is grater than the absolute maximun rating for any of the IC pot terminals, for any variant.

    In case we consider the wiper resistance as specified in the datasheet (50~to~100 Ohm), the IW current would drop to (10~to~11,8mA ), which is still too close to the absolute maximum rating or slightly above.

    Perhaps the IC doesn't get overheated as the sample code keeps the wiper at values that produce an overcurrent for only a fraction of the time and the IC can widthstand current pulses of up to 20mA.

     

    p.d. An updated datasheet version (newer than the one referenced in the project's page) is at: http://www.analog.com/static/imported-files/data_sheets/AD5204_5206.pdf.

  12. David Maciel Silva
    August 22, 2014

    You're right, sometimes we come across in technical details that are not observed in the project design.

    I analyzed the data sheet on the Ad5206 the maximum allowable current is 20mA, as you said correctly.

    However, considering that it is possible to control the variation of current through a reference resistance, maximum and minimum acceptable agrees can be used?

  13. geek
    August 22, 2014

    “It is amazing that you reflect the most practical points with so less math, a rare combination and is a delight to read.”

    @vasanjk: I agree with you on this. I find Aubrey's blogs to be the right mix of technical details and analysis. The good thing is that there's limited technical aspect which makes the blog interesting for both technical readers and non-technical ones.

  14. samicksha
    August 23, 2014

    I guess one of the challenge with digital potentiometer is to decide whether to use three terminal configuration or two terminal potentiometer configuration.

  15. vasanjk
    August 30, 2014

    AK

    I went to a design centre recently and saw something quite innovative. They had a 99 tap digital pot and achieved a 199 tap variation. They actually did something similar to PWM based DAC technique. The SPI code sends command for consecutive tap positions alternately. The output is fed to a RC network which filters the variation.

  16. antedeluvian
    August 30, 2014

    JK

    The SPI code sends command for consecutive tap positions alternately. The output is fed to a RC network which filters the variation.

    Thanks far passing this along. Several thoughts occur to me:

    1. This is so simple, yet so smart- a logical extension of dithering the outputs of a DAC or oversampling an ADC.

    2. Why didn't I think of this?

    3. I know exactly where I could've used it.

    I love the information sharing that happens on Planet Analog.

  17. vasanjk
    August 30, 2014

    AK,

    “Why didn't I think of this?”

    It is now someone's turn to be amazing. Probably they got into AK's shoes.

    In fact, I too felt the same way when I came to know an enterprising youngster who is just out of college did this.

    Bob Pease and other greats would be pleased.

    Another thing – All great things seem to be always simple.

  18. geek
    August 30, 2014

    “I guess one of the challenge with digital potentiometer is to decide whether to use three terminal configuration or two terminal potentiometer configuration.”

    @samichka: Doesn't that depend on the nature of the application? I assume that some applications might require a three terminal configuration while some may require a two terminal one.

  19. antedeluvian
    August 30, 2014

    JK

    They had a 99 tap digital pot and achieved a 199 tap variation. They actually did something similar to PWM based DAC technique. The SPI code sends command for consecutive tap positions alternately. The output is fed to a RC network which filters the variation

    The thought occurred to me but I forgot to mention in my last posting- if you are PWM'ing then you could probably increase the number of steps between taps to triple or even quadruple the steps or more. It may consume a lot of CPU cycles and will not have a fats response time, but it is a idea worth reflecting on.

  20. vasanjk
    August 30, 2014

    AK

    As an extension, would it be possible to convert a 12 bit DAC into one with a higher resolution with the Sam technique?

    I understand it could also consume more CPU cycles but may be faster than a PWM DAC.

  21. antedeluvian
    August 31, 2014

    JK

    See the article by Dave Van Ess (and the subsequent onles listed at the bottom) “Squeeze 10-Bit Performance From an 8-Bit DAC

  22. vasanjk
    August 31, 2014

    AK

    Thanks for the info on the article.

    This article is dated 2008 and the idea is old after all. We have so much information available, it is just that we do not look for it.

     

  23. chirshadblog
    August 31, 2014

    @vasanjix: Well there are so many in the archive. Its just a matter of searching. How about doing a flashback of all the past articles or atleast the most commented ones ? That might help a bit. Steve what do you think ? 

  24. chirshadblog
    August 31, 2014

    @anteduluvian: Thank you for sharing it mate. Very helpful. Reading it and once its done surely I will have some questions to be clarified. 

  25. vasanjk
    August 31, 2014

    @chirshadblog

    It may be a good idea to look back and learn more (and unlearn some unwanted things). This website and some others contain treasures which come in handy at times.

     

  26. dassa.an
    August 31, 2014

    @chris: Yes a good idea indeed but how to select which ones are the best ? Is the No of Comments can be justified the quality of the article ? 

  27. dassa.an
    August 31, 2014

    @Well yes it was very informative. I also benefited a lot out of that. 

  28. dassa.an
    August 31, 2014

    Mate what about having all those which are similar in topic and prepare a summer kind of a thing ? That will help and benefit a lot because those who read the old articles will surely get whats been written in the summary. Also they will not want to read again.

  29. vasanjk
    August 31, 2014

    @Dassa

    No of comments are by no means a measure of the quality. Sometimes comments dwell on the subject discussed by a leading comment and digress from the blog's subject itself.

    The blog must be selected based on the subject it deals with and the treatment it gives to the subject. It need not be filled with equations, but, must have a real  practical impact as this and other AK blogs do.

  30. dassa.an
    August 31, 2014

    @vasan: O also saw some old articles which were very informative. Yes you ahev to dig deep to look and it's a bit tiring but still its worth. I got stuck on some formula and I was looking for an solution for months. I digged the old article here and found it. Im happy now. 

  31. dassa.an
    August 31, 2014

    @vasan: Yes its possible to convert.

  32. antedeluvian
    August 31, 2014

    dassa.an

     O also saw some old articles which were very informative. Yes you ahev to dig deep to look and it's a bit tiring but still its worth. I got stuck on some formula and I was looking for an solution for months. I digged the old article here and found it. Im happy now. 

    I agree with JK (vasanjk) that the number of responses is not indicative of the quality of the blog. I have seen many excellent blogs with very few responses.

    I also agree with you that it can be hard to search for different topics since the Planet Analog search engine doesn't seem very effective. Google is often better. There is one other approach you can use if you want to work through a blogger's output chronologically. To the right of any blog you will see “More Blogs from…”. At the bottom of that list you will see “More from …”. If you click on the link you should see the complete blog output of that blogger on Planet Analog from February 2013 when it was transformed to its current format.

     

  33. PCR
    August 31, 2014

    dassa.an it's a great thought then all will be in one easy access point. 

  34. Davidled
    September 4, 2014

    I am curious why engineer need to consider for terminal configuration. Digital potentiometers are built by a chip that receives the requested data to change the value through the connected pin.

  35. samicksha
    September 11, 2014

    Yes two terminal or three terminal configuration depends on nature of application but again we need to this plan as per our demand.

  36. antedeluvian
    May 14, 2015

    As a result of lack of exposure, I never mentioned membrane potentiometers. I just came across this interesting article “The Membrane Potentiometer” and I thought I should add it to my blog.

  37. Steve Taranovich
    May 14, 2015

    Thanks for adding Membrane Potentiometers—they were new to me too Aubrey

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.