Possible bad batch of RTL8201's in circulation

[5U4GB]

For anyone working with  RTL8201 PHY's, a recent  post on the Konnected.io blog indicates that there may be a bad batch of these in circulation:

They are pretty confident now that one component, the Ethernet PHY chip on the board, had a bunch of defective or somehow damaged product in circulation, and we may have received defective components on more than one occasion from different suppliers.

[...]

It seems that those that were built with the newer stock of the RealTek chip are working just fine. To prove this, MacroFab’s test engineers removed some chips from tested and known working boards, and transplanted them on boards that were failing – and they worked! 100% of the time!

[CaptDon]

One of the 'Big' manufacturers of instrument grade opamps had an internal quality control issue. They informed us that they would replace 1 for 1 any chips we sent back within the failed code date range but they WOULD NOT be responsible for our labor costs. We had a few hundred circuit boards out in the field using those date codes with around 6 'suspect' chips per board. What a freaking night mare!!!!

[5U4GB]

An update on this, turns out the problem was a missing pullup resistor which didn't matter for previous revs of the RTL8201 up to date code 'M' but suddenly became an issue with the latest revs, date code 'N' and onwards.  Apparently this also affected other devices using the RTL8201.  Full details here, in the entry for 10 May.