HPF effect on SD24 module input (MSP430i2041)

E L55,

There is no HPF/BPF integrated into the SD24 at all. The only filter applied is the digital decimation filter, which acts as an LPF for noise above half the sampling frequency. Additionally, this is a continuous time sigma-delta converter, meaning that there is no discrete switched capacitor input stage at all.

Your problem is not one I've seen before in my work with the SD24. Can you give an example of the attenuation you are seeing?

Do you see the same attenuation when you apply a DC supply to the input directly?

Hi Tyler, 

Thanks for your prompt reply.

We did the following experiment - Replaced the transducer with AC signal generator and applied sine wave at various frequencies.

Using oscilloscope, we have measured the output amplitude on AIN0P-AIN0N input with and WITHOUT MSP430i2041 installed :

You shall check ADC inputs using 2-channel scope and compare to ADC readings. Also please show how you initialize ADC and it's clock.

Hi, 

I followed the demo app per Resource Explorer:

static SD24_initConverterAdvancedParam SD24_Params = {
		SD24_CONVERTER_0,
		SD24_SINGLE_MODE,
		SD24_NOT_GROUPED,
		SD24_INPUT_CH_ANALOG,
		SD24_DATA_FORMAT_BINARY,
		SD24_FOURTH_SAMPLE_INTERRUPT,
		SD24_OVERSAMPLE_256,
		SD24_GAIN_1
};

...

void main(void)
{
    WDT_hold(WDT_BASE);

    // Internal ref
    SD24_init(SD24_BASE, SD24_REF_EXTERNAL);
    SD24_initConverterAdvanced(SD24_BASE, &SD24_Params);
    // Delay ~200us for 1.2V ref to settle
    __delay_cycles(3200);
...

Please confirm that you use external reference. It could be leakage that leads to clipping. I would test using stronger "virtual ground" resistors: instead of using 2x 66MOhm, I would test with 2x 10KOhm and signal from generator. If 2x 10K gives expected results then you have starting point to investigate further.

Hi,

We'e tried both internal and external ref. the same phenomenon occurred both ways.

>We'e tried both internal and external ref. the same phenomenon occurred both ways.
Did you try phenomena using lower resistance of virtual ground resistors? - 10KOhm instead of 66MOhm?

E L55,

Sorry I was unable to respond earlier. I have been out of the office for Thanksgiving holiday.

Are you still seeing this attenuation from the device? Have you tested by directly attaching your source directly to the input of the device?

Another thing to note is that you should not be exceeding 80% of your FSR on the inputs to the SD24. The FSR is defined as your VREF/GAIN. Assuming you are using the internal 1.158V reference and a GAIN = 1, you should not exceed +/- 928mV. Operating outside of that range, the performance listed in the datasheet may not be reached.