If you received a MAX31855 Thermocouple breakout (https://www.adafruit.com/products/269) within the last few months and you may have a incorrectly trimmed chip as per MAXIM, the chip maker. We’ve only had 3 reports of this so far, however we sent this to all customers today.
** The affected chips are Lot #”FM385253AK” date code “201345” **
Please check your MAX31855 thermocouple breakout for “1345” on the second line or “253AK” written on the third line of the chip
** If you have a “253AG” chip that one is OK, we hand verified that the 253AG’s are not affected! **
Please email email@example.com to get a new breakout if you have a “253AK” breakout. The only ones we have in stock at this time are 253AG chips. If we are out of stock, don’t worry – we’ll have more in stock soon and we’ll be able to send you one shortly.
Please read below if you would like more details we received from MAXIM.
Product #269 is manufactured here at the Adafruit factory using known-sourced chips. We only purchase this chip direct from MAXIM (the original designer and manufacturer of the chip sensor.) This chip is available in a few different ‘flavors’, each one is for a different kind of thermocouple. Since we only stock K-type thermocouples in the Adafruit shop, we only order the K type MAX31855. We also have a QA process to make sure that the correct chip is received from the factory – in this case the full part # is MAX31855KASA+
About 2 weeks ago we received a report that two of the themocouple amplifiers they purchased from us seemed to be ‘miscalibrated’, specially at higher temperatures (1000*F). We made some suggestions and also requested a return. The customer observed that the amplifiers seemed to be tuned for J-type thermocouples. We checked our inventory and stock but could only verify that we did in fact have the correct sensor chips MAX31855KASA+ in stock.
Then about a week ago another customer contacted us, while we were waiting for the return, and mentioned that they were also experiencing issues with the raw chip and that they contacted MAXIM and were told that there is a bad batch of chips in circulation.
We emailed back and forth with MAXIM for a few days and the representative informed us they made a bad batch of chips back in late 2013 and discovered this issue in March 27, 2014. Before that discovery, we were shipped a reel of 2500 bad chips on 2/26/2014.
MAXIM did not notify customers in March 27, 2014.
We pulled and destroyed all of our current stock of incorrect 253AK breakouts and chips immediately upon learning this.
At this time we have some stock of correct 253AG chips that we hand-tested to verify that the temperature is reported right.
Here is the report from MAXIM on the root cause and corrective action
Affected lots were trimmed as 87-31855+JX0 instead of 87-31855+KX0.
MAX31855 is a thermocouple-to-digital converter, and is trimmed depending on the type of thermocouple application (K-type, J-, T-, R-, etc); so customers with mistrimmed parts will have incorrect temperature sensing and incorrect digital output codes.
1. uFlex Automated program loading does not set a value for SetupID variable. This variable is expected to contain the part number. The DT03 program reads this variable to determine which trim routine should be executed. Manual program loading has no issue and sets SetupID to the correct part number.
2. DT03 test program sets the part number to J-Type (87-31855+JX0) if SetupID is blank. Defaulting the program to a specific part number is not a good practice.
1. Camstar/XML update to set SetupID=PARTTYPE in auto-loading (Done).
2. New test program submission by primary Test Engineer to remove the part that defaults the program (Ongoing).
3. Review data logs and identify list of miss-trimmed lots. A total of 5 lots were identified miss-trimmed (Done).
4. Temporarily turn off auto-loading of DT03 program in Camstar (Done).
Please email firstname.lastname@example.org to get a new breakout if you have a “253AK” breakout. The only ones we have in stock at this time are 253AG chips.