Recommended alternate(s) for the ADS1241

Hello David,

One possible option may be the ADS1256.  It offers a multiplexer, PGA and higher sampling rates (2SPS to 30kSPS); the output code is in two's complement and the full scale is +/-2*VREF/PGA

 LSB=+2*VREF/PGA(2^23-1); full scale is +/-2*VREF/PGA

The ADS1248 offers a mux and PGA with 24 bits of resolution with sampling rates up to 2000SPS but the common-mode input voltage range needs to be at least 0.1 V from the rails; so if the user needs to measure single-ended signals with respect to GND; it will need to use bipolar supplies.  The output code is in two's complement; the full-scale voltage is : +/-VREF/PGA

Best Regards,

Luis

Hi Luis,

Both suggestions look good and if I had to make a decision right now I'd go with the ADS1248.

But your caution about the input voltage range raises a flag. In this situation, the input range is between 0.4 and 0.5 Vref. Vref is from an external 2.5V source and is applied to both the ratiometric sensor and REFP0. AGND is common to the low side of the sensor and REFN0.

My requirement is not that I measure near AGND or Vref but that I be able to accurately measure the range listed above. Am I right in assuming that the ADS1248 would perform as expected for this?

Thanks,

Dave

Hi David,

The ADS1248 restriction on the common-mode input voltage range  requires both the INP and INN inputs to be at least 0.1V from the voltage supply rails (AVDD/AVSS).  As the differential voltage between the inputs or the PGA gain becomes greater, the common-mode voltage range is reduced.  If the user is measuring a single-ended signal with respect to GND, where AINN=GND and AVSS=GND, this will create a common-mode voltage range issue.  The device would need to be set up to run with bipolar supplies AVDD/AVSS =+/-2.5V to avoid the common-mode voltage range problem.  

The design note below explains the ADS1248 common-mode voltage range in more detail, providing examples.

http://e2e.ti.com/support/data_converters/precision_data_converters/w/design_notes/input-voltage-range-requirements-for-the-ads1248-and-ads1148-families.aspx

Thank you and Best Regards,

Luis

That's intersting. I'm not deep into analog at the extreems (the rails), but in comparing the ADS1256 and the ADS1248, wouldn't the ADS1256 have a similar limitation?

I noted that there's no specification for the "Common-mode input range" for the ADS1256, so would the same problem exist there?

Let me backup a second and reiterate what I had said earlier. In this case, AINn will be in the range of 0.4 to 0.6 of Vref where Vref is 2.5V. In all cases, AINn is well away from the 0.1V restriction. Would this issue be a problem?

What I'm looking for is the assurance that my (single ended) inputs will be correctly converted to the number where 0x000000 is AVSS and 0x7FFFFF is Vref. There is no need to measure near these limits.

Thanks,

Dave

Hello David,

The ADS1256 and ADS1241 have a different input stage than the ADS1248.  As you mentioned, the ADS1241 and ADS1256 specify the input range in terms of the absolute voltage with respect to the supplies.  the ADS1256 input stage is similar to the ADS1241 where a buffer can be enabled or disabled changing the allowed input voltage range.

ADS1256 and ADS1241

-  The ADS1241 input voltage range is similar to the ADS1256, if you disable the buffer, the inputs can go 0.1V below Ground (GND-0.1) and +0.1V above the supply (AVDD+0.1V).  If you enable the buffer,  the allowed input voltage range is GND+0.1V to AVDD-1.5V.

- On the ADS1256, if you disable the internal buffer, the inputs can go 0.1V below Ground (GND-0.1) and +0.1V above the supply (AVDD+0.1V).  If you enable the buffer on the ADS1256, The input voltage range is GND to AVDD-2V.

ADS1248

The ADS1248 is different, the input voltage range is defined in terms of the common-mode voltage. In addition, the ADS1248 can be set up with analog uni-polar supplies or bi-polar supplies; where the AVDD =2.7V to 5.25V and AVSS=GND (unipolar) or AVDD =+2.5V  and AVSS=-2.5V (bipolar supply). 

The common-mode voltage is defined as the average voltage of the inputs:

VCM = (AINP + AINN)/2  equation (1)

The differential voltage is defined as the difference between the inputs:

VINdifferential = AINP-AINN  equation (2)

The allowed common-mode voltage range is a function of the device supplies, the differential input signal and the PGA gain used; where the following common-mode range limits  must be met in order to guarantee proper operation:

AVSS + 0.1V + (PGAGAIN*VINdifferential) /2  <  VCM  <   AVDD- 0.1V - (PGAGAIN*VINdifferential) /2   (equation 3)

As the differential voltage increases and/or the PGA gain increases the common-mode voltage range is reduced.  Ideally, to maximize the range of the allowed PGA gain and or differential signal allowed; the common-mode voltage of the signal is biased close to the middle of the supplies. 

It sounds like you are using uni-polar supplies (AVDD=5V) and (AVSS=GND) and both the positive input (AINP) and negative input (AINN) are biased away from the supplies, close to the middle of the supplies (2.5V).  If this is the case, most likely, the inputs are in the allowed common-mode range depending on the PGA gain and differential signal.  The best thing to do to ensure this is true; is to take the case for the maximum differential signal expected, maximum PGA gain that will be used and solve for equations 1, 2 and 3 to ensure the common-mode range condition is met.  The design note I previously referred to provides examples. 

http://e2e.ti.com/support/data_converters/precision_data_converters/w/design_notes/input-voltage-range-requirements-for-the-ads1248-and-ads1148-families.aspx

----

The ADS1248 device will provide a result of 0x000000 when the differential voltage accross the inputs VINdifferential = 0V; it will produce a 0x7FFFFF when VINdifferential= +VREF/(PGAGain).

The ADS1256 device will provide a result of 0X000000 when the differential voltage accross the inputs VINdifferential = 0V; and 0x7FFFFF when VINdifferential=+2*VREF/(PGAGAIN).

If this is not clear, please forward a diagram so I can understand your set up showing the connections for AINP/AINN sensor connections, supply connections AVDD/AVSS and letting me know the maximum differential signal that is expected out of the sensor; and I coud go through it,  

Regards,

Luis 

Hi Luis,

On going back and comparing the ADS1256 and the ADS1248, I think I see where some of my confusion creeps in. For the ADS1256 (and similar) one side of the internal amplifier can be set to AGND, see the MUX control byte. It would appear that this can't be done for the ADS1248, see the MUX0 and MUX1 control bytes. This would imply to me that the ADS1248 can only be operated in single ended mode by externally grounding one of the inputs and this is where the common-mode parameter comes in.

I'd like to show you what the circuit looks like just for confirmation that I'm taking the right appriach, but how do I upload a file associated with this thread? It would be in the form of a pdf file.

Thanks,

Dave

Hi David,

I will review the circuit.  To attach/post the pdf file, you may click on the "insert file" button on the tool bar; a "Insert File" window will appear asking you to browse for the pdf file on the computer.  Please see below.

Thank you and Regards,

Luis

Hmmm, OK, I hadn't noticed that before!

Anyway, the file is a paper that I wrote for my client to explain how we could measure temperatures to a high degree of precision. You should ignore the reference to the ADS1243 in the next to the last paragraph. The first design went with the ADS1241 and worked very well. But they come from an 'old world' where samples are taken every two seconds and average in software for 5 minutes. They wouldn't buy my argument of the ADC doing some of the filtering for them! So, that's why I'm here today.

Anyway, the technique VT and VT1 in the diagram and computing from there. It relies on AGND corresponding to 0x000000 and VREF corresponding to 0x7FFFFF.

Any thoughts? WIll the ADS1256 do the job?

Thanks,

Dave

0675.RTD-Details.pdf

I can see the advantage of having the internal current sources. But in this case, I feel that the application doesn't warrant it. At this point, I'll pursue the ADS1256 and examine the details.

Thanks for all the info!

Dave