• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 47 subscribers
  • Views 1052 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.

AMC1311-Q1: Another AMC1311 single ended output question

Alex G
Prodigy 80 points
Part Number: AMC1311-Q1
Other Parts Discussed in Thread: AMC1311, AMC1300B-Q1, ISO224, AMC1311EVM

Hello, I was reading the discussion in this question but am still unclear on something: https://e2e.ti.com/support/amplifiers/f/14/t/908199

For an analog input voltage that can swing negative, it makes sense to me why we have a VCM_ADC. But the AMC1311 is a unipolar device, or in other words VIN can only range from 0-2V, and VOUTP will range from 1.44V to 2.44V, and VOUTN will range from 1.44V to 0.44V, just like in Fig 22 of the datasheet.

For a single-ended ADC with 3.3V reference, VCM_ADC = 1.65V. If we use the recommended circuit in Fig 50 of the datasheet with VCM_ADC = 1.65V, then when VIN = 0V (the minimum input voltage, because again the AMC1311 is unipolar), VOUT = 1.65V and we lose half of the ADC input range. Is this the necessary tradeoff to use the full input range of the AMC1311 or am I missing something?

  • 0
    Prodigy 80 points

    Perhaps another way to phrase the question is: in high voltage DC sensing applications, we want say 0-400V input to map to 0-3.3V output, and it's not clear to me how to achieve this with the AMC1311 VCM_ADC not tied to GND (through a resistor).

  • 0 in reply to Alex G
    TI__Guru 55322 points

    Hi Alex,

    So essentially, you would take your 0-400V HV input and attenuate it by a factor of 200 (2M and 10k sense resistor for example) using a resistor divider to match the 0-2V input of the AMC1311. 

    The internal gain of the device is fixed internally at 1 and has an output common-mode voltage of 1.44V as you've mentioned and as shown in figure 22. 

    You then want to take this and apply it to a 0-3.3V input Single-ended ADC. The VCM_ADC does indeed need to be tied through a resistor to GND in order to do this. I've attached a TINA simulation that shows this. 

    The reason the VCM_ADC is often tied to the mid-point of the ADC input for these example diagrams is that the input voltage sensed by the isolated amplifier could be bipolar, such as the AMC1300B-Q1 or and ISO224.AMC1311 Differential to Single-Ended.TSC

  • 0 in reply to Alexander Smith
    Prodigy 80 points

    Thanks for the fast reply Alex Smith, and for confirming my understanding!

    It seems, then, that for unipolar devices such as the AMC1311, the recommended circuit in Fig 50 of the datasheet with VCM_ADC is in fact not ideal (unless VCM_ADC = GND2, which would be confusing because why not just write GND2, as VCM_ADC usually implies the midpoint of the ADC input range).

    Also, for AMC1311EVM, VCM_ADC is tied to a low, but not 0V, reference. This would also "waste" approximately 10% of the input range of the ADC. Why design it like this?

    edit: one more thing - in this post, a TI employee says "Using only VOUTp (or VOUTn) utilizes half of the AMC1311’s output voltage range, which reduces your Signal to Noise Ratio." - but as established in this question, the AMC1311 is unipolar and so MUST utilize either half the output range of the AMC1311, or half the input range of the downstream single-ended ADC, correct?

  • 0 in reply to Alex G
    TI__Guru 55322 points

    Happy to help Alex! 

    It really depends on the ADC you are connecting to. The diagram is simply an example way to transfer the signal from the output of the ISO AMP to the input of the ADC. Since there are so many different ways to do this, a simple example is shown and customers can design their own opamp circuit that fits their needs. 

    For the AMC1311EVM, the direct differential outputs can be sampled at J2, or the single-ended output can be accessed at J6. The VCM_ADC is set by the buffer U5, where an external voltage can be applied to TP4 or the resistor divider can be used as you mentioned, providing a voltage slightly above GND. While this may be a waste of FSR for some applications, it is important to consider that some ADC front-ends require headroom from the supply rails in order to operate without being damaged. i.e. an ADC is supplied with 5V AVDD and 0V AGND. The ADC inputs may require 100mV of headroom from the supply rails in order to operate without potential damage. Similar to the figure above, this is not ideal for some applications, but it was designed like this to service as many different needs as possible and switching out a resistor is simple if customers are inclined to modify the EVM to their own specifications. 

     The input of the AMC1311 is unipolar, but the output is bipolar. Referencing figure 22, you can see that the outputs go from 1.44V to 2.5V for VOUTP or 0.5V for VOUTN. If only using VOUTP or VOUTN, you are not taking advantage of the other half of the output swing to increase the accuracy of the measurement. However, if doing a differential to single-ended conversion, this resolution is not lost, and can be level-shifted as needed depending on the ADC input type.

    If you are interested in additional examples, please see these cookbook circuits: 

  • 0 in reply to Alexander Smith
    Prodigy 80 points

    Thank you for the answers and cookbook links Alex!

  • 0 in reply to Alex G
    TI__Guru 55322 points

    My pleasure. Let us know if you have additional questions!