• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 3 answers
  • Subscribers 46 subscribers
  • Views 3629 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1118: Is there any reason to use the ADS1118 instead of some of the newer ADC ICs?

Kyle Sulek
Prodigy 210 points
Part Number: ADS1118
Other Parts Discussed in Thread: ADS1220, ADS1247, ADS1246, ADS1248, ADS1262, ADS1263, ADS122U04, ADS124S06, ADS124S08, ADS114S08

I am looking for a high-resolution ADC with a PGA for measuring thermocouples. I have seen the ADS1118 suggested may times. However, after looking around, TI appears to have many newer, higher-resolution alternatives. Here is a list I have found:

Part Name ADC Bits SPS Channels
ADS1220  24 2k 4
ADS1246 / ADS1247 / ADS1248  24 2k 2/4/8
ADS1262 / ADS1263 32 38k 10
ADS122U04  24 2k 4
ADS124S06 / ADS124S08 24 4k 6

Right now, I am thinking of using the ADS1262 or the ADS124S08. Is there any reason that I should use the ADS1118 instead?

  • 0
    TI__Guru* 98195 points
    Kyle,


    The choice of ADC will ultimately depend on what features you may need for the measurement, the channel count, the desired data rate, and what resolution that you require. Of the devices you mention, the ADS1118 is a 16-bit device, while the others are 24-bit devices (or higher in the case of the ADS1262).

    The ADS1118 is a fine device, but if you need more precision, the ADS1220 might be better suited to your application as a similar device. Both have two channels, so you would be able to measure the thermocouple and do a cold-junction measurement.

    If you also need RTD or bridge measurements and you need a higher channel count, I would probably recommend the ADS124S08. It's basically a next generation of the ADS1248 devices.

    What sort of application in mind? If you explain what you are trying to measure, we should be able to help make a recommendation.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 210 points
    Thank you Joseph,

    I have a system with 16 thermocouples. For eight of the channels, I would like at least 0.5° (C) resolution between 0° and 200° C. On the other eight channels, I would like 1° C resolution between 0° and 1000° C. I need to sample all 16 channels 2-4 times per second. I have a differential multiplexer that I can use if necessary but it would be preferable to connect all the probes to the ADC. I can use multiple ADC chips if necessary.

    I was using a MAX31856 but this chip does not allow for grounded thermocouple probes.
  • 0 in reply to Kyle Sulek
    TI__Guru* 98195 points
    Kyle,


    I would think that a few ADS114S08 or ADS124S08 devices would work for you. However, you'll need a faster data rate to get so many channels. If you have 16 channels and are sampling at 2 times per channel, you'll need a minimum data rate of 32 SPS.

    However, I'm not sure about your problem with grounded thermocouple probes. If grounded means where the thermocouple is welded to the outer sheath and the sheath is conductive, then this might be a problem. If the thermocouple is used in an environment with ground loop voltages, the ADC may not work properly.

    When making meausrements with the ADC, the thermocouple needs to have some DC bias that is typically near the mid-scale input range of the device. That is, if you use the ADC with a +5V supply, the negative thermocouple lead would be attached to +2.5V so that the measurement would be in the middle of the PGA input range.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 210 points
    The thermocouple probes I am using are welded to the outer sheath and the sheath is conductive.

    In this configuration, would I be able to safely remove the biasing resistors to prevent a ground loop?
  • 0 in reply to Kyle Sulek
    TI__Guru* 98195 points
    Kyle,


    There are a few ways to go depending on what device you use.

    If you're using the ADS1118, then you can just tie the negative thermocouple lead to ground. The ADS1118 doesn't use a real amplifier to increase the gain on a measurement. It uses input capacitor scaling to perform gain, so the input range is much larger. For devices like the ADS1220, you can do the same, bypassing the PGA, but only getting up to a gain of 4. By attaching the thermocouple to ground, you can make sure that the ground loop voltage is 0 or significantly reduced.

    If you use the ADS124S08 or ADS114S08, you could also attach the thermocouple to ground and bypass the PGA, but you'd only be able to make measurements in a gain of 1. This isn't a great solution because the resolution would be poor.

    The best thing to do would be to attach the thermocouple to ground and use a bipolar supply. Set AVDD to +2.5V and AVSS to -2.5V so that ground is midway between the two supplies. This way, the input would be within the input range of the PGA, gain can be used at any setting, and you still have the thermocouple set to ground where loop voltage is minimized.

    All of these possibilities require the thermocouple to be connected to ground. Of course, there will be some limitations on the connections. If the thermocouple is attached to a high voltage, then you'll need to consider protecting the ADC with series resistance, and make sure that the there is a low-impedance path to ground.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 210 points

    My understanding is that the "equivalent circuit" for a grounded thermocouple should look like this? Is this correct? Otherwise, I don't see the need for a bipolar supply since there are only positive voltages being measured.

  • 0 in reply to Kyle Sulek
    TI__Guru* 98195 points
    Kyle,


    A grounded thermocouple means that there is an electrical connection from the thermocouple tip, but it's not necessarily attached to ground. It's not grounded unless you explicitly connect it to ground. We attach the thermocouple to ground to try to eliminate the ground loop where ever the thermocouple is attached.

    For the ADS124S08/ADS114S08, the bipolar supply is required so that the thermocouple measurement (with a common-mode dc input near ground) is set up at the mid-supply point. This ensures that input measurement within the input range of the PGA. Because this is a thermocouple measurement, you'll likely need a high gain and use the PGA. The PGA is set up similar to the front end of an instrumentation amplifier, and has input range limitations. There's a description of PGA input voltage requirements on page 34 of the ADS124S08 datasheet.


    Joseph Wu
  • 0 in reply to Joseph Wu
    Prodigy 210 points

    Thank you. I now understand that the bipolar supply is necessary because of the PGA input limitations.

    In a previous post, you had recommended the following:

    If you're using the ADS1118, then you can just tie the negative thermocouple lead to ground.

    Is this picture correct? I think I may be misunderstanding your suggestion. I cannot removed the ground connection at the end of the thermocouple due to mechanic design restraints.

  • 0 in reply to Kyle Sulek
    TI__Guru* 98195 points
    Kyle,


    I understand now. Yes, that's the correct connection. The thermocouple is grounded to the chassis of whatever you are measuring and the other ground would be the ground of the measurement circuit with the ADC.


    Joseph Wu