To tell if your DS18B20 is a B7 die, and therefore may have corruption problems:

1. Read the 18B20's ROM code. Look at only the serial number portion of the ROM code:

0xCC00000SSSSSSS28

where CC is the CRC
00000 is the customer code
SSSSSSS is the serial number
28 is the family code

2. The serial number will tell the device's die revision as follows:

If the number is equal to or less than 0x054BBDF  the device is a B6

If the number is in the range  0x54BBDF - 0x0662B4F the device is a B7

If the number  is equal to or greater than 0x0662B50 the device is a C2

One more thing, the DS18S20 has the die revision id printed on it. It is the last two letters of the second line of text. "B7" is the revision you want to avoid.

--Brian D. Hindman,
Dallas Semiconductor MAXIM.

05/04/04 - New Revision of chip die coming.

There is indications from Dallas Semiconductor that a new revision of the troubled DSx20 chips (DS18B20, DS1820S, and DS1822 ).is in the works. Its now estimated that a new revison for the troubled chip will be available mid august, early september time frame. Unknown at this time if it will incorporate a new front end (ala DS2408) or not. See the previous news item marked 04/08/04 for details of the troubles. We assume its the C2 version as mentioned above...

04/08/04 -  It appears there is a problem in the B7 die used in several popular temperature chips. (DS18B20, DS1820S, and DS1822 ). Due to efforts at Dallas there are now software work arounds for helping to solve this problem. The time for a hardware fix is unknown at this time.

Dallas semiconductor is now requiring a wavier as of April 8th 2004 to get some parts. (Though it effects parts made earilier).  It appears there is a problem in the B7 die used in several popular temperature chips. (DS18B20, DS1820S, and DS1822 ). Due to efforts at Dallas there are now software work arounds for helping to solve this problem.   Check your sensors, developers may wish to implement the software work arounds to their software.evelopers  Read the full application note about this issue

 

At the request of Dallas Semiconductor, we have removed the document AN247.pdf. titled "DS18x20 EEPROM Corruption Issue".  Dallas has decided this should have been a "internal confidental document" and would rather not have it available on the web. The document explains the issue and documents workarounds. If you have further questions or need to locate this document, you can contact Dallas at thermal.support@dalsemi.com to ask for it. It is also available from their distributors.

 

See the Wavier (43,260 tif file): DS1820 wavier.tif