ADs1248 o/p to temperature conversion formula

Yash,


The ADS1248 output gives digital result that represents a ratio between an input voltage and a reference. The digital result is in a twos complement notation and can represent both positive and negative numbers.

Assuming you are using the ADS1248 internal reference of 2.048V when the PGA gain is 1, the input voltage is given as:

Input voltage = 2.048V * (Data output/2^23)

If you have that input voltage connected to the L35DZ, then the input voltage is related to temperature by this equation:

Temperature = (LM35 Output)/(10mV/°C)

From that, you should be able to calculate the temperature from the data conversion.

However, I do have one strong word of caution. The input range of the ADS1248 does not extend to ground when the ADS1248 has a unipolar supply (operating from 0 to +5V, or 0 to +3.3V). The input amplifier is limited in range to 100mV of the positive supply and negative supply. As an example if the supply is 0 to +5V, the input will be limited to a range of +100mV to +4.9V. This range is further limited when the device has a gain more than 1, where the input signal is amplified and the output of the amplifier is limited by the supplies.

If you want to use the ADS1248 for this application, you will need some sort of bipolar supply (± 2.5V), or use a different ADC which has a range that extends through the entire supply. The ADS1220 is a similar device which does have this capability and should be able to make the measurement.


Joseph Wu

Yash,


For the thermocouple measurement, using the VBIAS is fine. This should place the input measurement near the center of the input common-mode voltage range.

The comment about needing to put the input in the common-mode input range was intended for the measurement of the LM35DZ which was mentioned in the first post. The measurement for this device is typically referenced to ground, and if this is a unipolar supply, the negative input for that measurement (connected to AINN of the ADC) would be attached there and would be outside the common-mode input range of the ADC.


Joseph Wu

Yash,


1. The common-mode input range refers to the input of the PGA which is integrated into the device. You cannot use the ADC without the PGA. This would still be for the AINx analog inputs for the device. We refer to the entire device (including the PGA) as the ADC.

2. The ADC cannot measure ±10V signal. This is much too large for the ADC. If you want to measure a ±10V signal, it will require some sort of front end conditioning. It will either require a front end like a PGA280 to lower the signal, or some sort of instrumentation amplifier. Additionally, I wouldn't use the VBIAS as the negative input for anything other than a thermocouple. Setting the DC operating point was the intended use for the VBIAS line. The VBIAS isn't meant to be used to sink or source any significant current which may be present depending on the sensor. Also, this signal may be noisy which may also cause errors in the measurement. Using VBIAS for the thermocouple, this noise is common-mode and should not add direct error to the measurement.

3. I'm sorry, but I don't know the LM35DZ very well, and I don't know the terms lead sense and body sense. If you are really interested, I would post the question in the Temperature Sensors Forum.

4. I'm not sure I understand this question. Again, I would say that the input does not extend out to +10V in any way. The input signal must be within AVSS+0.1V to AVDD-0.1V with a PGA gain of 1. If the gain is greater than 1, then the input range will be further limited as the output of the PGA must also stay within AVSS+0.1V to AVDD-0.1V.


Joseph Wu

Yash,

This is the exact same question as this following post. Are you two working together?

I'll be a bit more detailed in my answer for this just in case if comes up another time. In this case, the input is outside common-mode input range of the PGA. I'm not sure about all of the setup you're describing, but I'll show a diagram to explain what I mean.

First, you describe using a voltage divider for the 0V to 10V input. Is this the way this is set up?

If this is the setup, then the input is outside the common-mode input range. Note the AINN input. If AINN is connected to ground and AVSS, then this input is outside the input range.

(AVSS+0.1V+(Vin)(Gain)/2) ≤ VCMI ≤ (AVDD-0.1V-(Vin)(Gain)/2)

Assuming AVSS=0V, Vin=0.909V and Gain=2, you get the following if you replace the values in the above equation.

1.009 ≤ VCMI ≤ 3.991

With the input at 0 to 0.909, the common-mode voltage is 0.909V/2 = 0.4545V so the common-mode input voltage is not inside the input range.

The common-mode input range is limited by the input and output range of the PGA. Let's look at a diagram of the PGA. It is shown below.

If you look at the PGA, the amplifiers will still be limited by the supplies going from AVSS to AVDD. Again, lets assume that the inputs are AINN=0V and AINP=0.909V and the PGA is set to a gain of 2.

This diagram shows two things that will be wrong with these input voltages. First, the AINN is too low. Even in a gain of 1, the input must be above 0.1V for the input range of A2. Second, the common-mode is too low because the output of A2 cannot go below the negative rail. The output A2 will be limited to maybe 100mV above AVSS.

So using the voltage divider is not going to work when the negative input is tied to ground (or really AVSS for this example).

There is still a small discrepancy for the second case where the input is driven by the DC calibrator. I'm not sure why this gives you a different result, but I suspect it is the same problem. It depends on the exact setup, but the it could be the calibrator has a different DC offset point, and that it still violates the common-mode range.

The common-mode input range is also addressed in a different post listed below. There is also a presentation that gives a couple of examples of how this is limited based on the input PGA limitations. You can find it as an attachment to one of the posts.

I'd read through this post, and the presentation. If you have any more questions, I would also include a diagram of your setup. It's important to note exactly what voltages are seen by AINN and AINP separately.

Joseph Wu