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.

AMC3301: AMC3301

Part Number: AMC3301

Hi,

We want to measure Line to line voltage using AMC3301. 

We have referred the application note of the device and so accordingly we have not connected the VIN_N pin to the ground, but we are getting distorted output differential voltage. Please refer below snaps.

Next, the simulation is done with Vin_N connected to the ground, but here the positive and negative peak values are different for Vin_diff (Input differential voltage) and Vout_diff (Output differential voltage).

So, please give suggestions to eliminate the above errors or if possible provide some reference for the same. 

  • Hi Smitha,

    As shown in the EVM user's guide picture that you posted, the high side ground in connected to the left side of R2, the INN filter resistor. This ensures that the input voltage is within the common-mode range of the inputs pins. To verify this is the issue, try looking at the absolute voltages on VINP and VINN, not the differential between them. 

    I would recommend moving R3' closer to R3, on the other side of the differential RC filter. 

    I see you have HGND connected to a GND node, but I do not see this connected anywhere else. Try removing this GND node and connecting HGND to the bottom of your sensing resistor, R7. 

    I cannot comment on what changed from your first to second set of waveforms, since I'm not positive where you connected the GND to VINN. Please provide a screenshot of the schematic used for the second set of waveforms. 

    If you attach your TINA file, I'll be happy to work through it with you.  

  • Hi,

    Please refer below snaps where the issue is about different positive and negative peak values of Vin_diff. This simulation is done without connecting ground to Low voltage side (3.3V).

    When we are connecting ground to low voltage side (3.3V), the error message is popping up (snap of which is attached below).

    Can you please give suggestions on how to resolve the issues (either with ground is connected to 3.3V or not).

  • Hi Smitha,

    Can you please attach your TINA file? 

    If you provide the TINA file I can help work on it with you. 

    I would recommend removing R3', splitting R7 in half, then connecting HGND in a center tap configuration. 

    Depending on how your power supplies are setup, you may need to have a separate GND node in order to connect the low side of the amplifier to a virtual ground. 

  • Hi Smitha,

    Can you please attach your TINA model to this thread? 

  • Hi Alex,

    Sorry, I missed to attached the file previously.

    The file is attached for the reference.

    VLL_sense_AMC3301.TSC

    The issue related to positive and negative peak values is resolved. But we are facing other issues:

    1). Related to connection of ground to low side supply (3.3V)
    2). Related to frequency

    1). When we are connecting ground to low voltage side (3.3V), the error message is popping up (snap of which is attached below). Otherwise, when the ground symbol is not connected the simulation is working fine. So, please give suggestions on this.

    2). We found that there is another issue w.r.t frequency. We are getting proper output when the simualtion is done with 50Hz, however we are unable to get proper output for our requirement (140KHz). As per device datasheet, output bandwidth is minimum 250Khz.

    Since we are about to freeze our design we would like to know whether this device is suitable for our application and is it only a model issue? Please confirm.

    Snap for 50Hz frequency;

    Snap for 140KHz frequency;

  • Hi Smitha,

    I have made some changes to your TINA file to change the VDD sources and implement the center tap connection like I talked about in previous posts. The benefit of the center tap connection is that R3' will not be required and there will not be an issue with the bias currents or gain error from using R3'. 

    The reason that you are seeing attenuation on your inputs is that the RC filter is not quite what you think it is, 10ohm and 8.2nF. You must also consider the resistance of the HV scale down resistors, and sensing resistors when calculating your R value. Your RC cutoff frequency was much lower that the expected 1MHz. I've updated the inputs as you can see in the file. 

    6505.VLL_sense_AMC3301_AS.TSC

  • Hi Alex,

    With the suggested changes in the circuit, we are again seeing issues related to the positive and negative peak values of Vin_diff (the peak values are different, highlighted in the attached snap).

    Please give suggestions on this.

  • Hi Smitha,

    Looks like the center-tap GND1 node was not actually connected and INN is tied to HGND. I connected the center-tap GND1 node and removed the INN to GND1 connection. You can also slightly increase the size of the sensing resistors to maximize the full-scale range. 

    8865.VLL_sense_AMC3301_AS.TSC

  • Hi Alex,

    Thanks for the response.

    Hope that the +5V ground reference used for the High voltage side is only for reference!

    Or alternately can we connect the midpoint between the split resistor to HGND?

    We understand that both things are the same and if not can you please explain the difference.

  • Hi Smitha,

    The 5V+ VDD1 net is not necessary, it's simply to help the simulation, a configuration that I copy/pasted from another TINA file. 

    The midpoint between split resistor is connected to the GND1 net and HGND is connected to GND1 net as well, they are the same. It is simply easier to connect them this way.