Hi,
I have a Siglent SHS1102, a portable oscilloscope that has two galvanically isolated inputs. Despite a few quirks with saving data, it's a remarkably nice instrument. It's only a few years now and suddenly it didn't turn on anymore. I checked the battery and that was at 0 V, meaning the protection probably switched it off. I bought a new battery from Siglent (worked very well, I was surprised, 2018 timestamp even) and the oscilloscope came back to life. For half a hour.
I had already checked whether there was any input to the battery before, but there wasn't. Since buying a new battery was a reasonable choice over taking the thing apart and potentially ruining it, I didn't follow up on this issue and hoped that it was just detecting whether a battery is present or not before charging. Since the second battery just ran dry and the instrument also didn't turn on without a battery I finally took it apart.
There are a bunch of switchmode regulators on the lower quarter of the main board. None of those was a dedicated LiPo-charger, though. I searched, until I finally found the LiPo-charger: U27, a Texas Instruments BQ24170:
http://www.ti.com/product/BQ24170Checking the voltages, I found an almost absent output voltage (300 mV or so) but also no input voltage to the IC! Aha!
I looked at the datasheet and traced the elements from the recommended schematic on the board. There are two transistors regulating the input voltage: Q1 and Q2. As it happened, these are right next to the input plug: U24 and U25. While measuring them, U25 had a nice diode drop across source and drain, but U24 was short circuited across S and D. I guess that is dead!
The markings on these things were a bit enigmatic but turned out to correspond to the suggested device from the datasheet:
TI CSD17313
http://www.ti.com/product/CSD17313Q2So I invested the 50 Cents at Mouser and got me a nice new one. Reworking the board wasn't too difficult despite the large ground planes and double sided (components on both sides!) board. 400 °C hot air and covering all the parts around with Kapton tape. I put the oscilloscope back together e voilâ, it came back to life and is now also nicely charging the LiPo.
I suspect that the Fet might have overheated, as one of the thermal pads (source) might not have been soldered well. It didn't have any solder on it on the broken IC, so maybe there was a wicking issue.
Attached are pictures of the device and the repaired FET (pink arrow). Maybe it helps somebody some day!