Thanks. It has a mask ROM, so it can't be slowly forgetting its program, like an old EPROM or flash memory could.
According to section 2.10 of the datasheet, a reset is generated internally, but the TEST pin (pin 9) can be used for an external reset if the supply rises slowly, see https://pdf1.alldatasheet.com/datasheet-pdf/view/1284216/TI/TMS1000NC.html
However, in the FM4 schematic you posted, pin 9 is connected to the switch matrix via a diode...
True up to a point. Its at the top end of high 24 hours after charging finishes.
(I had a mains timer switch that started having an over contrasted LCD display such that you couldn't read it correctly. It used a single 1.2 volt cell for backup and that was also (bizarrely) used for limiting the voltage to the chip almost like a Zener. The voltage was around 1.8 volts when plugged in and overdriving the display. The cell had failed even though it provided a backup supply)
That sounds like they are using it as a low voltage Zener of some kind.
I really don't know what to suggest next tbh. The chip would have be the last suspect normally but eventually you reach a point where you just have to try it if only to prove it good... which I know is easier said than done. Beyond that you are looking to see what changes when the battery is fitted vs when removed and why that makes a difference. Absolute levels of the rails and perhaps rail sequencing... those 100uF caps C11 and C12. Once you have looked at all those things you have to try the chip.
I really don't know what to suggest next tbh. The chip would have be the last suspect normally but eventually you reach a point where you just have to try it if only to prove it good... which I know is easier said than done. Beyond that you are looking to see what changes when the battery is fitted vs when removed and why that makes a difference. Absolute levels of the rails and perhaps rail sequencing... those 100uF caps C11 and C12. Once you have looked at all those things you have to try the chip.
Is it possible that there is a menu in the tuner that lets you select a station that comes on at powerup?
Interesting. That's a CMOS input opamp. Can't say I've come across those before but it would be a 'new chip' back in the day when the Quad was released.