Redefining High Resolution and Low Noise in Delta-Sigma ADC Applications

Here are the questions and answers from the webinar:

Is it possible to use a thermocouple in combination with a load cell?

Yes, there are sufficient input pins to allow both of these sensors to connect to a single ADS1262 or ADS1263. Each of these sensors requires only two input pins. Additionally, two input pins could connect to the load cell’s excitation voltage and be used for the ADC’s reference voltage (for ratiometric measurements). Therefore, only 6 out of a total of 11 input pins are used up in this configuration, leaving room for an additional sensor or two!

Are these series of ADC’s available on an EVM?

Yes, you can find both of these EVMs on TI.com! Here are the links…

• ADS1262EVM-PDK
• ADS1263EVM-PDK

Do you know what the resolution would be at 38 kSPS?

At the highest PGA gain setting, the ADS1262 and ADS1263 are capable of resolving signals down to 3.3 µV_RMS at 38.4 kSPS.  When used with a PGA gain of 1 V/V, the ADS1262 and ADS1263 allow for a wider input signal range, but input-referred resolution becomes 103.5 µV_RMS (at 38.4 kSPS).

Can you explain why the status register is sent with the data stream?

There is an optional STATUS byte that is sent prior to the 32-bit data output. This byte is transmitted with the data because it indicates if any errors were detected during the conversion result that may have led to an erroneous conversion result.

What's the best way to use the device with bipolar input signal such as +/-5V?

The ADS1262/3 can measure differential signals up to +/- V_REF. With an external 5-V reference, you could measure this +/-5V signal directly (as long as you meet the input-common mode range requirements). Well almost…if you don’t need to buffer this signal, then you can bypass the PGA and measure the +/-5V signal directly. However, when the PGA is enabled, the largest signal you can measure is +/-4.7 V because of the PGA’s output swing limitations. So, if you enable the PGA, or if you are using a smaller reference voltage (such as the internal 2.5V reference), than an external resistor divider would be needed to attenuate the input signal.

Is there a reason why most high-resolution sigma-deltas can only use a 5V analog supply? (I noticed it on many devices but never really understood why it’s so difficult to get them with a 3.3V AVDD supply rail, for example.)

There are many high-resolution sigma-delta ADC’s that can operate off of only a 3.3V analog supply rail (for example the ADS1220, ADS1232, ADS1248, and ADS1292). However, 5-V is a typically analog supply voltage because it allows for a larger input signal range. This offers additional dynamic range and added headroom to support larger gains prior to performing the analog to digital conversion, while also maintaining compatibility with most reference voltages and logic level s.

A good informative presentation. It was good to see the discussion about the sigma delta basics as it helps understand how the chip works and achieves 27bit accuracy at the lower sample rates. It also made me appreciate how sigma delta uses a small chip area allowing extra functions to be added. My application needs higher sample rates, so thinking about sigma delta has helped me to understand how the effective bit resolutions in the data sheet change with output sample rate. The technique of swapping the inputs using the multiplexor to cancel out the DC offsets is useful to know. Thanks for answering my question about how best to use the chip with +/-5V signals.

Thanks,

Martin

Great presentation. Inspired me to try out the Roadtest.

Recording is now available in the "Webinar Recording" tab of the event. Please let me know if you have any questions!

Dave

verheyenrudy

They are good. Gutted I missed this one.