• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 49 subscribers
  • Views 321 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

AMC1311: Spice Model

AMC1311: Gain and Offset Error

Michael David
Intellectual 460 points
Part Number: AMC1311

Hello,

Since this an isolated amplifier, I am not sure how the two errors reflect on the output side. Would I just consider the gain error as the input offset is reflected to the output side?  

  • 0
    TI__Mastermind 47625 points

    Hi Michael,

    I often compute errors with respect to the full-scale range on the input side, otherwise the offset terms must be multiplied by the gain. Of course, since the AMC1311 has a gain of 1, that is not much of an issue. I don't entirely understand your second sentence, give the excel calculator a look and let me know if that clears up your question. 

    Here is an excel sheet that can help you estimate typical and worst case errors for the AMC1311: https://e2e.ti.com/support/amplifiers/f/14/p/815789/3019110#3019110

    Be sure to select "single-ended" under the configuration tab in order to select the AMC1311 under "device".

  • 0 in reply to Alexander Smith
    Intellectual 460 points

    Hello Alex,

    Thanks for the excel, it does make my job a little easier but I think it created more questions. LOL. It did not answer my original question. Let me try this again by first seeing if I understand how this device works.

     

    From my understanding the input gets reflected to the output as the total difference between the deferential signal centered at the common mode output voltage.

    Vout_iso_P = Vcm + Vin/2

    Vout_iso_N = Vcm - Vin/2

    The common mode voltage will always be a fixed value set by the device, in this case 1.44V. The device has unity gain so whatever voltage comes in will come out the same just as a difference centered at the common mode. The difference will be between the P and the N outputs (Vin = Vout_iso_P - Vout_iso_N). Is this correct?

     

    Now on to my error question. If we only consider the gain and offset error as the main contributor to any variation in the output differential signals difference. The offset error is added to the input signal and reflected to the output by multiplying it by the gain error. This will only effect the outputs differential signal difference. Is this correct?

     

    Here is another set of question related to the common mode voltage variation.

    1. Is does the common mode voltage variation reflect error change in the P and N signals?
    2. Does it add directly to the difference output or are they not correlated?
    3. If the common mode voltage was to increase will it increase the differential signals difference?  

  • 0 in reply to Michael David
    TI__Mastermind 47625 points

    Hi Michael,

    As with most things, the more you explore the more questions you'll have :)

    Correct. This FAQ may help, it shows some simulation waveforms that confirm your observations: https://e2e.ti.com/support/amplifiers/f/14/t/889581

    I am not so sure about your second conclusion: "The offset error is added to the input signal and reflected to the output by multiplying it by the gain error."

    The transfer function for any analog device can be shown as y = mx+b

    Offset error is defined as what is on the output when the inputs are shorted together. Ideally, it would be 0V, but there is inevitably a small amount of differential voltage present. This is your 'b' term. Since this term is specified at x = 0, gain error is not accounted for - hence me not being sure about your statement. 

    Gain error is the difference in between the slope of the input signal and the slope of the output signal. So, if you have a 0.2V to 1.2V input signal, but the differential output produces 0.1V to 1.3V, that would be a gain error, your 'm' term.

    1. Output common mode? No, any variance in output common-mode voltage would be an offset and not affect the gain terms. 

    2. Not correlated. I would expect 0Vin to produce a 0V differential output regardless of if the common-mode output voltage is 1.39V (for example) or 1.44V. 

    3. No. Same as #2.

    Just FYI, high level comparison of isolated amplifiers and isolated modulators: 

    https://www.electronicdesign.com/power-management/whitepaper/21130089/introduction-to-isolated-amplifiers-and-modulators

  • 0 in reply to Alexander Smith
    Intellectual 460 points

    Hello Alex,

    This clears up my questions. On another note, does TI have a similar error calculator tool for current sense amplifier such as the AMC1303?. I do have a few questions about the error calculator tool you referred me tool.

    1.  Under the User Inputs section, is the resistor drift related to the input voltage divider resistors? This would be the R1, R2, R3, and RF. Also there is not input for RF.

    2. Under the specification section, there are line items that is not clear what they represent.

    a. Gain Error Rdivider - is this the external voltage divider resistor gain error determined by the resistor drift input? Looks like its divided by the ISO AMP input impedance. Is this error added to the ISO AMP gain error?

    b. GE Rdivider Drift - if is a user input then how does the internal resistance drifts?

    c. Offset Error Rdivider and drift - would I assume this is related to the external resistor tolerances (R1, R2, R3, and RF)?

  • 0 in reply to Michael David
    TI__Mastermind 47625 points

    Hi Michael,

    Glad to hear the explanations helped. It's in the works. The calculator you have can be used for current calculations, but it requires tweaking the user-inputs and ohms law, i.e set R1=R2=0ohm then manipulate R3 and HV+ to where you need it to be. 

    The trouble with the AMC1303 is that it is a modulator and the performance of modulators are ultimately determined by the digital filter.This is in the works as well. 

    1. Yes. Adding Rf is on the to do list as well. 

    2a. This is accounting for the fact that the sense resistor, R3 is in parallel with the differential input resistance. This can be an issue when using a device with low input impedance for voltage sensing. The differential input resistance is composed of a differential opamp as well as a few other things that I can't get into on the forum. When using the "with R3'" configuration, R3' adds to the gain resistor Rg, influencing the gain of the opamp.  

    2b. Internal resistance drift and internal resistance variation are not yet accounted for in this calculator. It will be included in the next version. 

    2c. This is actually looking at the variation of the bias current exiting the input pins and it's amperage variation over temperature. This is again an issue with our lower input impedance parts. Not so much of an issue for the AMC1311 where the bias current is in the nA range. 

    This cookbook circuit goes into further detail on the R3' configuration: https://www.ti.com/lit/an/sbaa350/sbaa350.pdf?ts=1596050575133&ref_url=https%253A%252F%252Fwww.google.com%252F