SBOK097B April   2025  – June 2025 TMP9R01-SEP

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1 Introduction
  5. 2 Sample Identification
  6. 3 Irradiation Facility and Setup
  7. 4 Die Micro-section
  8. 5 Test Set-Up
  9. 6 SEL Results
    1. 6.1 TMP9R01-SEP SEL Data
  10. 7 SET Results
    1. 7.1 Summary of SET Results
      1. 7.1.1 Local Temperature Reads
        1. 7.1.1.1 Analog Alert Captures
    2. 7.2 Cross Section Events and Event Rate Calculations
  11. 8 Detailed Results Per Run
    1. 8.1 Detailed Run Raw Data
  12. 9 Summary
  13. 10Glossary
  14. 11Revision History

7 SET Results

Transient characterization data was captured while exposing the units to different levels of radiation. The cross-section plots were captured to provide information of the trend of how the device performs under radiation. During the exposure time, all I2C registers were continuously read. A device reset was taken before every read. The data collected in the registers and the alert pin signal was used to understand the complete functionally of the device and support the SET results presented in this report. Data captured in this report is labeled with reset or without resets.

TMP9R01 device supports reset using the two-wire general-call address 00h (0000 0000b). The TMP9R01 device acknowledges the general-call address and responds to the second byte. If the second byte is 06h (0000 0110b), the TMP9R01 device executes a software reset. This software reset restores the power-on reset state to all TMP9R01 registers and aborts any conversion in progress. The TMP9R01 device takes no action in response to other values in the second byte.

Reset Software ProcedureFigure 7-1 Reset Software Procedure

Data was read and captured every 50ms. With resets, the max time for the device to recover temperature is 50ms and no temperature value was seen to be stuck longer than 50ms during the exposure. The max time for the TMP461 to read all registers is shown in the following equation:

Where 36 is the number of clock cycles per communication cycle.

Equation 1. 36×clock period×# of register reads

Without resetting the device, there is a large chance the device performs poorly. TMP9R01 is a digital I2C device and transient/upset events were defined in the following way:

  • A remote or local temperature output with an error larger than 1.5C was recorded as a transient. Temperature delta was between the temperature readings during exposure and before. Data was read through I2C SCL and SDA lines and recorded through excel. Local and remote temperature is read through two bytes: high and low byte. The high byte represents the whole portion of the temperature result while the low byte represents the fractional portion of the temperature results. The amount of changed bits and location impact the overall accuracy of the device. The changes in the configuration registers can also affect the accuracy of the temperature result. The bits within the following registers impact temperature accuracy as shown in Table 7-1:
Table 7-1 Registers Impacting Accuracy Events

Pointer (HEX)

Register Description

Potential Accuracy Errors

 Event Type

R 00h

Local temperature Register (high byte)

Reading Incorrect Measurement

SET

R 15h

Local temperature Register (low byte)

Reading Incorrect Measurement

SET

R 01h

Remote Temperature Register (high byte)

 Reading Incorrect Measurement

SET

R 10h

Remote Temperature Register (low byte)

 Reading Incorrect Measurement

SET

R 03h W 09h

Configuration Register

Changing the range from [-40C, 127C] to [-60C, 191C]

SEFI

RW 11h

Remote Temperature Offset Register (high byte)

An unwanted temperature offset is set

SET

RW 12h

Remote Temperature Offset Register (low byte)

 An unwanted temperature offset was set

SET

RW 16h

Channel Enable Register

 Disabled the temperature conversion of remote and local temperature sensors

leading to inaccurate readings

SEFI

RW 23h

η-Factor Correction Register

An unwanted n-Factor calibration was set

SET
12-bit Q4 FormatFigure 7-2 12-bit Q4 Format
  • A SEFI was recorded when the alert/therm signal was tripped. The device was programmed with default settings to where the alert stays off. If the alert is tripped, the number of assurances is recorded. Operation of the ALERT (pin 7) and THERM (pin 4) interrupts is shown in Figure 7-3. Operation of the THERM (pin 4) and THERM2 (pin 7) interrupts is shown in Figure 7-4.
Alert and Therm Interrupt OperationFigure 7-3 Alert and Therm Interrupt Operation
Therm and Therm2 Interrupt OperationFigure 7-4 Therm and Therm2 Interrupt Operation
  • Any bit that was changed within any of the I2C registers outside of the temperature reading registers was recorded as an upset. These events test if any of the configuration settings was changed during exposure. All events are SEU events. No MBU events were present.

The bits within the following registers impact the alert/therm signal as shown in Table 7-2.

Table 7-2 Registers Impacting Alert/Therm Events
Pointer (HEX)Register DescriptionPotential Alert/Therm Errors

Event Type

R 02h

Status Register

Mismatch in Status of the Alert Signal

SEU

R 03h W 09h

Configuration Register

Unwanted mask of the alert signal and changing the mode from alert to therm

SEFI

R 05h W 0Bh

Local Temperature High Limit Register

Incorrect limit setting

SEFI

R 06h W 0Ch

Local Temperature Low Limit Register

Incorrect limit settingSEFI

R 07h W 0Dh

Remote Temperature High Limit Register (high byte)

Incorrect limit settingSEFI

RW 13h

Remote Temperature High Limit Register (low byte)

Incorrect limit settingSEFI

R 08h W 0Eh

Remote Temperature Low Limit Register (high byte)

Incorrect limit settingSEFI

RW 14h

Remote Temperature Low Limit Register (low byte)

Incorrect limit settingSEFI

RW 19h

Remote Temperature THERM Limit Register

Incorrect limit settingSEFI

RW 20h

Local Temperature THERM Limit Register

Incorrect limit settingSEFI

RW 21h

THERM Hysteresis Register

Incorrect hysteresis settingSEFI

RW 22h

Consecutive ALERT Register

Unwanted number of out-of-limit temperature measurements required for ALERT to be asserted

SEFI

A device reset was taken before every temperature read. All data captured in the report was taken with a temperature reset. Data was read and captured every 50ms. With resets, the max time for the device to recover temperature is 50ms and no temperature value was seen to be stuck longer than 50ms during the exposure. Without resetting the device, there is a large chance the device performs poorly. We have tested the results with and without resets and found that with resets the device preforms better.

The following table captures SET data with TMP9R01 in extended mode. Extended mode extends the temperature range to reach the range of -64C to 191C adding more bits to the temperature registers. The data below highlights the benefits of reseting the device before every temperature read.

Table 7-3 Data Comparing SET Results with Resets and No Resets
LetEFF

(MeV)

# Events (Reset)

# Events (No Reset)

8.7

24

22

48

26

324

56.36

33

643