Many thanks for Tom for the hours of investment in this data.
I support a number of businesses (in USA) with various Synology (as well as 2 2415's for myself), and this issue has become vexing.
Traditionally we have replaced ONLY Q2, and for nearly all units this had been sufficient.
However, one stubborn unit came back as soon as the weather got cool (this client's network office is in a building basement and it gets quite chilly).. so when the unit was warm, it works ok, but when it gets below ~40F (~5C), the unit will not power on.
So we performed the 3.9k-->10K switch.
I'm including photos here, for ease of reference, for others looking for this in the future.
Note: this is actually faster and easier to do than doing the other adjustment, and as I can read from the schematics, there is very low risk in voltage problems..
As I read it, it is safer changing the power supply from 3.9k -> 10k, because it effectively drops the total possible voltage load to the sensor circuit (and doesn't harm other feeds).
Of course with all of this, it's quite easy to cause a calamity if the user damages/shorts while making the mod, so please don't attempt this if you are not dead 100% certain you have proper tools and know what you're doing.
I would begin by being certain your Power supply has been fully discharged. fortunately for this unit, it had been running, and mains were disconnected, effectively bleeding off remaining voltage (though this may not always be the situation)..
Anhang anzeigen 76300
Anhang anzeigen 76299
We only replaced the 3.9k with a 10.0 k resistor. and of course left the 3.9k "parked" on the board, in case someone finds it in 2044 and then we will have them baffled...
Anhang anzeigen 76298
Anhang anzeigen 76297
So, many thanks, Tom. This has helped us tremendously.
Tom, another question: do you think perhaps performing ONLY the 3.9-10k switch is possible to work, without replacement of Q2? I wonder if Q2 ever fails, or if it's simply working but at lower level than previous.. Next time we get one of these, I intend to try the resistor (3.9-10k) first, then see if that solves it...