Is the VSS Pin (pin 103) connected internally to GND_Shield?

[Stefan Schmidt]

Dear Nordic Support Team,

 In the datasheet of the nrf9160 (nRF9160_PS_v2.2.pdf) the VSS pin appears in the section 6.4.5.1 GPIO Electrical Specification, but I cannot find the information if it is internally connected to GND_Shield.

Best regards

Stefan

[Øyvind]

Hello Stefan, 

No sure what you are referring to. The GND_Shield pins of the nRF9160 are connected to Ground, similarly the VSS is connected to Ground. This is described in chapter 10 Hardware and layout. 

Kind regards,
Øyvind

[Stefan Schmidt]

Hi Øyvind,

thanks for your reply. I know that they have to be connected them to ground, my question is: Are they connected in the nrf9160?

I tried to read VCC_GPIO on a custom board and was wondering why the readings are so far away from the 1.778 V I was measuring with an external Multimeter. The readings are outside the 4% DC error, and always lower than 1.779 V. I increased the aquisition time to 40 µs, but this did not solve the issue.

When the modem was active the VCC_GPIO read with the ADC was even lower, so I thought that ground bouncing might be the issue and that it would be better to have the VSS pin routed to the star ground point separately from the GND_Shield pins (which I suspected to be the ground pins for the modem).

But now I have set up this sample zephyr/samples/drivers/adc on a nRF9160-dk and this is what I read for the 1.8 V VCC_GPIO:

- adc@e000, channel 1: 7621 = 1674 mV

I use this devicetree setup for this channel:

    channel@1 {
        reg = <1>;
        zephyr,gain = "ADC_GAIN_1_6";
        zephyr,reference = "ADC_REF_INTERNAL";
        zephyr,acquisition-time = <ADC_ACQ_TIME_DEFAULT>;
        zephyr,input-positive = <NRF_SAADC_VDD>;
        zephyr,resolution = <14>;
        zephyr,oversampling = <8>;
    };

Do you have an idea why the reading is so wrong?

[Øyvind]

Hello, and thanks for providing your project. I see the same result and inquired internally. This does look like a known issue, which there is currently no solution to. 

Kind regards,
Øyvind

[Stefan Schmidt]

Hi, thank you for your reply. Please let me know when you get feedback.

Is this issue somehow related to the fact that a common ground plane is used? Instead of a ground plane for the modem supply and a separate one for the VDD_GPIO?

Best regards

Stefan

[Stefan Schmidt]

Ping. Not sure, if my editing triggered an update on your side.

[Øyvind]

Thanks, I see the update. Have forwarded to our R&D team. 

[Øyvind]

Hi Stefan, my apologies for the late reply. The R&D team have been looking into this, and they have the following questions in order to understand the situation:

1. When you use the DK, are you using single ended mode? (I have no idea how to read the zephyr config below, that’s far above the SAADC config)

2. If you’re using single ended, it would be interesting to switch to differential, and connect the negative side to the exact ground that you use for the multimeeter where you measure the VDD voltage. Ground bound will affect the measurement, but 80 mV is signficiant ground bounce.

Kind regards,
Øyvind