• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 48 subscribers
  • Views 2016 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.

LM2904: Simple Analog Translator

Tiffany Greyson
Prodigy 80 points
Part Number: LM2904
Other Parts Discussed in Thread: TM4C1294NCPDT, , CD74HC4067, SN74HC00

I am trying to measure a 0-5V analog signal with a 0-3.3V tolerant analog input pin.  I'm planning on using an op-amp as a buffer then adding a voltage divider, but is there a simpler solution?  I need 20 signals, so that's a lot of external components with a cap and resistor for every opamp.  Any ideas?

  • Former Member
    0 Former Member
    Hi Tiffany,

    As far as a simpler solution...why not just use a voltage divider? Do you need more current than the DAC alone can provide? I would think the MCU cannot take/does not need a lot of current for a voltage sense.

    You mention a lot of external components including a cap, but I'm not seeing one in your schematic. If the cap is used on the supply pin of the amp, then one way to lower the total number of caps needed is by using quad-channel op amps. Then you'll only need one cap for every four op amps. Is this where the cap you mentioned is being placed?

    Also, do you have 20 output pins on your DAC and 20 input pins on the MCU all transmitting simultaneously? Just trying to get a better idea of what's going on here.

    Regards,
    Daniel
  • 0 in reply to Former Member
    Prodigy 80 points
    Thanks Daniel, I wasn't very clear in my application. I have 20 DACs providing a 0-5V analog out then I'm feeding those 20 outputs back to the TIVA for monitoring.

    The TIVA spec says the "ADC input leakage current" is 2uA. Does that mean the analog input pin will never pull more than 2uA at 3.3V? If that is so, maybe I can probably get away with a large series resistor then a voltage divider just before the TIVA ADC pin. I'm using the TM4C1294NCPDT. Does "leakage current" mean "maximum analog pin input current"?
  • Former Member
    0 Former Member in reply to Tiffany Greyson

    Hi Tiffany,

    Unfortunately, my experience with MCUs is limited as I don't work with these products.  I'm going to get someone with some more experience to try to help provide you with some more solid answers.  Please know that this might take a few days due to the Christmas holiday coming up.  In the meantime, I can offer some of my suggestions.

    To answer your question, I seriously doubt that "leakage current" and "maximum analog input current" mean the same thing.  Taking a look at the figure on page 1866 of the datasheet (also embedded below), you can see that there is some ESD clamping at the pin to help protect against unexpected ESD events.  The protection consists of diodes that leak current, even in idle state.  Additionally, I would think 2uA is a pretty small amount of current even for an MCU, though we'll need to double check on that.  For these reasons, I'm pretty sure this is not the specification we are looking for.

    I would say that the accuracy of the analog input into your TIVA might well depend on the proportion of current that is stolen away from the resistor divider by the MCU's ADC pin.  If the MCU is stealing a large amount of current from the bottom resistor, then your voltage divider will not divide the voltage as accurately.  There will be some offset.  If the input into the ADC is allowed to settle before the ADC samples (depends on the size of the C_ADC capacitor in the figure), then this may not matter.  Finally, some of this might also depend on the speed and current driving capability of your DAC.

    Hopefully this was not too much information or too overwhelming.  Just trying to put some relevant ideas out there.  I'm thinking you're not even going to need the op amp, if the DAC can provide enough current and the MCU is not overwhelmed.  I will work to get someone with more experience to lend you a hand in the meantime.

    Let me know if you have any more questions.

    Regards,

    Daniel

  • 0 in reply to Former Member
    TI__Guru**** 166216 points
    Tiffany,

    If settling time (speed) is not important then even a 2 resistor divider and no op amp would work. With that, there will be a compromise between resistor value loading the inputs and the real leakage input current of the Tiva. 2uA is maximum leakage which I think (not sure) includes the upper temperature of the Tiva. If true then lower temperature would be much less leakage. Ask about leakage in a new thread for Tiva's part number.

    Another option is adding a 5V mux to switch input signals and use just one input on Tiva so that just one op amp is needed.
    For example CD74HC4067 16:1 mux then just one op amp that is better than LM2904.

    As far as LM2904 goes you will need a 7V supply to make sure the op amp inputs can accept up to 5V input and so the output will reach 5V.

    Finally, tell me more about your input signals.
  • 0 in reply to Ron Michallick
    Guru 140200 points
    Hi Tiffany,

    I agree with Ron. 2µA should be the maximum input leakage current at maximum operation temperature. This makes absolutely sense.

    The SN74HC00, as an example for a CMOS chip containing very robust ESD input protection scheme, has a maximum input leakage current of 1µA at 85°C. At 25°C its maximum input leakage current is 100nA, while the typical leakage current is only 0.1nA. So, I think there's much headroom in your application for a passive voltage divider at the input of ADC, provided your application does not move much away from 25°C ambient temperature.

    To also fullfill the dynamic requirements you must slow down the ADC conversion frequency when using a passive voltage divider, and this the more the higher Rs is chosen. See section 15.3.2.5 of datasheet of TM4C1294NCPDT. An example: An effective source resistance of 500R allows an ADC conversion frequency of 1000Ksps. With an effective source resistance of 9500R, on the other hand, the maximum allowed ADC conversion frequency slows down to 571Ksps. If you add a filter cap to the voltage divider to form a low pass filter, the ADC conversion frequency must even furtherly be decreased.

    So, summarizing, a passive voltage divider without OPAmp buffer should be able to do the job.

    Kai