Hi,
This is a continuation of this topic:
https://www.eevblog.com/forum/projects/the-tlv760xx-regulators-are-a-nice-piece-of-crap/However, I managed to solve the power up issue that I was having. However, this voltage regulator module was made to be controlled via an external SPI controller board, which is the CP2130 Evaluation Board (CP2130 EB Rev. 2.0). It also has to deal with a very leaky constant current load, that is fed via a SMPS transformer that leaks out a tiny residual current from the main to ground.
So, the first problem that I had was when disconnecting the module with the CP2130 Eval board both connected to it and to the computer. It seems that the chip select output was supplying current via the AD5551 to the 5V rail (the LED was semi-lit). The TLV76050 didn't liked it. By connecting the module again, the TLV76050 simply shut down, no LED lighting up, until I disconnected the cable. That didn't damaged the TLV76050, but it was suspicious behaviour. Anyway, the TLV76050 doesn't have reverse current protection, so that might explain it.
The second problem, the one that damaged the TLV76050 regulator, plus the AD5551 (with 7V from the damaged regulator), has arisen when doing the same with the leaky load attached and disconnecting the power to the said module, with USB and load still connected and powered. I suspected that the OPA2180 was affecting the AD5551 (that is the only explanation that also explains the behaviour seen in the past).
So, I've dessoldered the AD5551 and replaced the TLV76050 with a TPS70950. The TPS70950 has reverse current protection, so it should handle a disconnection while the 5V rail is powered via somewhere else. I've also added a 10K resistor in series with the output of the AD5551. That resistor was added because I've verified a 6V spike at the first op-amp input on connection, without the AD5551 soldered there. That could have damaged the regulator and the AD5551.
I'm still waiting for a new AD5551 chip, but the modified schematic is there. I've modified the module according to the attached schematic (replaced the regulator and added the 10K resistor R3). My question is: am I missing something? I plan to use the TLV76050 in the final project. It will use a digital isolator (ADuM1310) for the SPI, with its downstream side supplied by the the same regulator (therefore, no chip select pin powering up the rail, and no need to use the TPS70950, IMHO). What worries me is the possibility of a leaky load doing some damage, despite that 10K resistor.
Kind regards, Samuel Lourenço