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.

THS4521: high offset voltage drift at temperature

Part Number: THS4521

Hi Team,

Customer is  having THS4521HD failures even after replaced the part. The replacement part still failed at 150C due to the high offset voltage drift. This failure delayed the delivery of theirPCB boards and production.

They're seeing around 7mV drift from 25C to 150C on failed parts. As in datasheet, the typical offset voltage at 175C is +/-0.13mV.

  • Does TI test each this part at temperature?
  • What is the offset voltage drift from 25C to 175C from your test?
  • Is the offset voltage drift repeatable?
  • Do you have any suggestions that we can screen this part before solder the part on board?

Also, may i confirm if THS4521EVM rated to high temperature?

Best Regards,


  • Hi Ernest,

    For the drift specs in the datasheet the limits were found through in lab characterization and simulation. This means this spec is not tested at final test. In-lab characterization tested the drift for the device from -55C to 125C, unfortunately there is no data taken to 175C. As for the repeatability of Vos drift, a device's drift performance should follow the same general trend excluding small changes in this spec as the device is broken in. As for a potential screening solution, it is hard to tie a device's drift performance to any measurable spec, even voltage offset itself at any given temperature does not relate to the Vos drift performance so there is no clear way other than testing Vos across temperatures. The THS4521EVM is rated to the operating range of the standard THS4521, which is -55C to 85C. Some of the components are rated higher than 85C but some are right at 85C which is where we recommend the limit should.

    Best Regards,


  • Hellow Ernest, 

    You are talking about the HT version which is done by a completely different group than the development group using the core die and different packaging. 

    1. Offset drift is rarely tested in outgoing package test, and certainly not in that >125C range. 

    2. Typical drift is a marketing number, you need to use the max number created by that high temp group way back in 2011. That 50uV/C max over 25C to 175C is a 7.5mV so actually your measured 7mV is not unexpected. Odd data sheet, apparently only specified for 3.3V supply for this 5.5V max die, apparently scared of internal self heating?

    Also these front page numbers almost tell me their test part was broken, both of the circled items are way off from the core die, 

    3. TI will not screen this, you could with your own test jig using appropriate high temp materials and components. If you use a socket, but sure to configre the device in very high gain to avoid oscillations. 

  • Hi Michael and Ignacio,

    Thanks for your response.

    The part that customer use is THS4521HD, which is rated at -55C to 175C. They have use this part in the same product for a few years. They didn’t had this high offset voltage drift issue till last a couple of months. Till now they have 19 THS4521HD failed within last a couple of months due to the high offset voltage drift.

    • Is the production line for THS4521HD changed?
    • Do you know how long this part (THS4521HD) will remain active and in production?
    • What does output voltage Low (TYP = 0.3V @ 175C) means on datasheet? Does it means the lowest amplifier differential output voltage is 0.3V at 175C? Why does the part that is rated at -55C to 210C has the TYP output voltage low of 0.09V, which is much lower than the part rated at 175C? Is this output voltage low at single end output pin (Vout+ OR Vout-)? Or is this output voltage low is the differential output voltage of the amplifier?

  • Hi Ernest,

    We were able to look into this for you and there does not appear to be any changes in the production line for this device. As Michael mentioned, the value you are seeing is not completely outside the range and expectation for this device. Due to process variation, there is a possibility that some device lots will perform more along the drift max limits highlighted in the datasheet. There are also no plans to end the production of these devices. The output voltage spec is the absolute voltage of the output pin and is not a differential spec. The reason the typical spec at the higher temperature is lower than at 175C is because it is taking the mean of the output voltage low across the entire temperature range. However, the spec at 175C was taken at that temperature which is why it appears higher than the mean crosses its temperature.  

    Best Regards,