• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 81 subscribers
  • Views 719 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Variable resistance to balance a bridge

Jake G.
Prodigy 190 points

Hi,

Currently in my project we balance a bridge of strain gauges(350 ohms) by using a 10ohm potentiometer.  The main problem with this is

1. Our system will be embedded so for calibration on the fly is impossible since we have to access pot with a screwdrivere.

2. The smallest digital potentiometer I could find was 1kohm which is too high and would lower the range of voltage across the bridge to an unacceptable point.

I was wondering if there are any solutions using the MSP430 that I could digital change a small resistance at the top of my bridges to keep my bridge balanced as the strain gauges start to change in small ways.  I could accept a total resistance of 100 ohms added but would prefer to stay below that value.

  • 0
    Prodigy 50 points

    What kind of granularity do you need?  With a 1k digital pot with 256 taps you get aprox 4ohms per count, That would give you ~25 settings below 100ohms.  If your need better, it may be possible to parallel a quad 1k pot, though I don't really know if the properties of a digital pot allow you to do this.

  • 0 in reply to dilbertclone
    Expert 2080 points

    If you have a DAC you should probably think about using a MOSFET as a variable resistor, It should be even possible to drive the FET with sort of PWM (better?).

     

    With the correct circuit and the right FET the output should be nice, but this is analogical, and analogical is not user friendly:D

  • 0 in reply to dilbertclone
    Prodigy 190 points

    dilbertclone said:

    What kind of granularity do you need?  With a 1k digital pot with 256 taps you get aprox 4ohms per count, That would give you ~25 settings below 100ohms.  If your need better, it may be possible to parallel a quad 1k pot, though I don't really know if the properties of a digital pot allow you to do this.

    Thanks for the response. Not sure if you are familiar with balancing a bridge of strain gauges, but I will try to explain myself a bit better.

    The pin coming out of the wiper is the V+, one side goes to the left half of the bridge, and the other side goes to the right side of the bridge.  The set up essentially adds the little extra resistance to the side that is lacking, due to error (hence I only use a 10 ohm pot). Each gauge has a resting resistance of 350 ohms, and usually they are within 1 to 2 ohms of that value at rest.  So having a 4 ohm per count is not going to cut it we need about 1 ohm per count or even under that to successfully balance the bridge.

    Also adding a 1k resistance to the top of my bridge would essentially lower my excitation voltage across the bridge quite severely, which in turn kills my resolution.

    I hope this clears it up a bit more,

    Jake G.

     

  • 0
    Guru 227245 points

    digitally controlled potis usually have such a high resistance because they use FETs to switch a resistor chain. These FETs have a low internal ON-resistance but it will be acceptable only if each resistor of the chain will be above a certain level. In your case, a chain of , say, 256 resistors with 0.4 Ohm each would be well in the range of the on-resistance of the switching FET on it, introducing a large nonlinearity due to the on-resistance as well as the temperature coefficient of the FET, and the differences between the FETs. And yoiu cannot ask much for a nonlinear or low-quality product, so nobody bothers to build one :)

    If you just need a calibration resistance, not a 'trustworthy controllable', you can simply use a normal FET and feed it with the output of an D/A converter. Some of the MSPs have an internal one with up to 12 bit resolution. Around the offset voltage of the FET gate, the FET behaves like a voltage-controlled resistor. The gate voltage to output resistance is nonlinear, but who cares if you can find a voltage where it fits. It works only if the actual current through the FET is small compared to the maximum current, and it has a relatively large temperature coefficient (and only works for DC signals) , but maybe it is the best you can get without introducing things like OpAmps.

    When selecting the FET for this purpose, the diagram in question is the Drain-to-source current over Drain-to-source voltage for different gate voltages.

**Attention** This is a public forum