LTC2984H (high temperature +125C)

Hi! 

LTC2984H is expected to be used for working with thermocouples at a constant temperature of +125C (or +85C).

The continuous duration of the device operation is not longer than 8 hours. As a result, the following questions arose:

1) What characteristics does an abrupt increase of the input leakage currein affect at +125C? (accuracy/time conversion)

2) What is the ADC accuracy at higher temperatures?

3) Are there any changes of Uref parameters at higher temperatures?

4) Are there any changes in the time conversion at higher temperatures?

5) Does a high temperature affect the EEPROM memory?

6) What is the time duration of the LTC2984H operation at +125C (+85C)?

7) Is there any alternative for LTC2984H?

The LTC2984 documentation contains its characteristics at +25C. Is it possible to obtain LTC2984 characteristics for the temperature of +125C (+85C).

Parents
  • 0
    •  Analog Employees 
    on May 29, 2018 10:47 PM over 2 years ago

    Hi @dariaa - apologies for the delayed response. 

    1) Leakage current will affect sensors and filters with high series resistance. My article goes into more details but you may see degraded accuracy with large series resistances at hot - keeping any filter or sensor resistances low will help minimize this issue: 

    http://www.analog.com/media/en/technical-documentation/lt-journal-article/ltjournal-v27n1-00-df-ltc298x-logancummings.pd… 

    2) Offset and noise will increase slightly with temperature - you can see the typical characteristics on pg. 9 of the LTC2984 datasheet

    3) Vref is trimmed at 25C - as with all references there is some temperature coefficient - a typical curve of Vrefout vs. Temp is shown on pg 9. Maximum error can be calculated by using the maximum tempco of 15ppm/C giving a maximum 1500ppm error of Vref at +125C. The curve is typically a bow however, and the actual deviation is often less than this worst case. For additional info see the LT6654 datasheet for a very similar reference characteristic. 

    4) The conversion time will depend on the internal oscillator frequency and this does change slightly over temperature. In no case however will the conversion time maximums on pg 4 of the datasheet be exceeded. 

    5) We state the EEPROM retention in Note 17 on pg 6 of the datasheet - 10-year data retention guaranteed up to 1000 program cycles. This does not change with temperature.

    6) Conversion time will depend on the type of conversion being performed as well as any additional mux delay added in your design (see page 64 of the datasheet). Max times not including mux delay are listed on page 4 for two and three-cycle conversions and these apply over the full temp range (as indicated by the red dot next to the specification).

    7) We believe the Temp-to-Bits family to be unique with it's combination of multiple simultaneous ADCs for ratiometric measurements and simultaneous cold junction compensation, sensor linearization for both built-in and custom sensors, high input impedance and inputs that support grounded sensors without ballasting or bias voltages.

    The closest alternative to the LTC2984 is the LTC2986-1 - a ten channel version of the part with EEPROM. Parts without EEPROM are the LTC2983 and LTC2986.

    Hope this helps, 

    -Logan

Reply
  • 0
    •  Analog Employees 
    on May 29, 2018 10:47 PM over 2 years ago

    Hi @dariaa - apologies for the delayed response. 

    1) Leakage current will affect sensors and filters with high series resistance. My article goes into more details but you may see degraded accuracy with large series resistances at hot - keeping any filter or sensor resistances low will help minimize this issue: 

    http://www.analog.com/media/en/technical-documentation/lt-journal-article/ltjournal-v27n1-00-df-ltc298x-logancummings.pd… 

    2) Offset and noise will increase slightly with temperature - you can see the typical characteristics on pg. 9 of the LTC2984 datasheet

    3) Vref is trimmed at 25C - as with all references there is some temperature coefficient - a typical curve of Vrefout vs. Temp is shown on pg 9. Maximum error can be calculated by using the maximum tempco of 15ppm/C giving a maximum 1500ppm error of Vref at +125C. The curve is typically a bow however, and the actual deviation is often less than this worst case. For additional info see the LT6654 datasheet for a very similar reference characteristic. 

    4) The conversion time will depend on the internal oscillator frequency and this does change slightly over temperature. In no case however will the conversion time maximums on pg 4 of the datasheet be exceeded. 

    5) We state the EEPROM retention in Note 17 on pg 6 of the datasheet - 10-year data retention guaranteed up to 1000 program cycles. This does not change with temperature.

    6) Conversion time will depend on the type of conversion being performed as well as any additional mux delay added in your design (see page 64 of the datasheet). Max times not including mux delay are listed on page 4 for two and three-cycle conversions and these apply over the full temp range (as indicated by the red dot next to the specification).

    7) We believe the Temp-to-Bits family to be unique with it's combination of multiple simultaneous ADCs for ratiometric measurements and simultaneous cold junction compensation, sensor linearization for both built-in and custom sensors, high input impedance and inputs that support grounded sensors without ballasting or bias voltages.

    The closest alternative to the LTC2984 is the LTC2986-1 - a ten channel version of the part with EEPROM. Parts without EEPROM are the LTC2983 and LTC2986.

    Hope this helps, 

    -Logan

Children
No Data