...
I have a major client where TI DC-DC switching converter chips have been replaced with an alternative design from another manufacturer. I was a big fan of the TI bq battery charger chips, but now have used alternative brands for those and the fuel gauge chips.
...
Are you allowed to share those brand names? We're looking into replace BQ40Z50. This is due to a huge increase of price (400% and then unoptanium), as well as TI's crappy attitude towards medium scale production support (firmware update algorithm is proprietary, which means you either have to use EV2400 with cumbersome and slow TI Battery Studio or have them pre-programmed by TI. Great show, especially if the only source for those chips is some Asian scalper).
In the above mentioned design, the one that suffered from the betrayal by Mouser, we also had to replace several TI components by other brands. Among them the buck-converter, a component most interesting when it comes to EMC.
I also had to downgrade the MCU to something smaller, but available. As a result I will quickly rewrite the RTOS-based firmware to something less resource hungry. Boy, am I looking forward to applying all that 1990's firmware know-how again.
At this time I fully expect schematic/layout changes for every single production run.
STNS01 (lower max charging current though) and Maxim have some good chips. ST is a basket case for their micros, but their charger chips were available last time I looked. Not quite as great as say the terrific TI BQ24250, but good enough for some recent applications. The Maxim has an excellent fuel gauge chip I have been using as a substitute... the MAX17048G+T10. I had never used this before and I am happy with it. Not too expensive and very easy to interface. I also did a design using a Cypress CYBLE module and used an embedded fuel gauge algorithm which worked perfectly, so no fuel gauge chip was required.
There are some newer BQ chips released, but I have little trust and confidence that TI will supply these in the longer term based on their past performance. I am nervous about TI. But I have noticed TI BGA chips are generally more plentiful the VFQN, TSSOPs etc, so BGA might be viable providing you put in test points for each net if possible.
TI's buck and boost converters
are were very cost effective. And, thanks to the Webench, they just worked (I still use the datasheet though to customise designs more to my requirements and to avoid needless odd-ball resistor values.) That's why we use them. I generally add low cost chip ferrites at the input and output of their converters and have never had a problem with EMC.
Analogue device have some amazing new patented high current DC-DC converter chips that are ultra quiet.
https://www.analog.com/en/parametricsearch/11497#/