There is a power supply BAT1 that provides +5 V power (the schematic shows 9 V, but it's providing 5 V):
After a certain period of time it shuts down, leaving its output in a strange tri-state (or high-Z) state:
- When there is no load, BAT1's output voltage drops to 2.4 V
- When there is a load, BAT1's output is pulled down towards 0 V (but it never reaches 0 V, it stops at around 1 V)
The supply detects if the load has been detached, and can only be restarted after a full detach-and-reattach.
I wish to modify this circuit to prevent this from happening, and let the BAT1's output float at around 2.4 V, just as if there is nothing attached to it (when it's shutting down).
Now my problem is R2: as it conducts, it pulls BAT1's output towards 0 V (to around 1 V).
I wish to eliminate the voltage drop on R2 when BAT1 is in tri-state. M1 was almost good for this purpose: just that unlucky budy-diode conducts even when M1 is closed, hence the voltage drop on R2.
I tried to simulate everything with an extra Zener diode with a modified schematic:
With SW1 closed, this is the exact same schematic as the previous, here are the plots of the test points:
Now, with SW1 opened:
It seems it increases tst2 along with tst1 up until D2 Zener's breakdown voltage (2.4 V), but it also distorts the Vout turning on only around 6 V.
The original circuit works fine, I just wish to eliminate the voltage drop on the R2 when BAT1 tries to measure the load with its tri-state output. How am I supposed to do that?