Sir fritzdecat . . . . .
I think that with a circuit talk thru, that you might then FULLY understand that circuit.
I'm using Bertus- es- es- es # 10 posts picture
Look at the extreme outer wiring loop that starts with the battery + to go up to the top and then turn right to go across to the corner and then down to the the emitter of Q1 (*****) then it continues thru the transistor and comes out the collector and then passes down to R1.
Now R1 has a computed resistance value, such that its value will not let enough power pass on thru the LED , such as to ever overload it.
Now continue with the loop on thru R1 and LED and then down and turn to the left and then finally go up, until ending at the battery - terminal.
Now we need to acknowledge the above cryptic marking of (*****), as that was being at the circuit tracing instant, where there needed to be ~0.6---0.8 negative vdc present at the base of Q1 for your wiring loop to be able to conduct / pass thru Q1.
Should you want to meter that, you use Q1emitter for - meter lead referencing and + meter lead to he Q1 base,now, if you are having an auto ranging meter it reads out correctly. . . . . or . . . . . if having a simpler meter, just get an upscale reading on the meter and the leads will then be correlating their color coding with the circuitry polarity.
Now here is how you get that base voltage level set up.
You have the R2 wiper initially adjusted such that it has the whole 100K being across its extreme ends. Then trace the path from the battery negative to the bottom of R2 and then up and thru its 100,000 ohms and then further up and turn right until you reach the base of Q1.
Look at the other possible circuit path of battery + into LDR and out of it down to the R2-Q1 base common junction.
Now, here is how to adjust R2.
You said / desired that at night / dark onset that the units LED will be LIT and then at SOME particular level of sunrise illumination, the LED will go off.
So here is how you simulate that , first you need to make it DARK by covering up that LDR sensor.
Now you need to see if that full 100K total resistance alone is letting enough voltage thru to already get that ~0.6vdc to the base to have the circuit loops conduction thru Q1, then lighting up your LED.
If no LED light then you start slowly changing R2 to a lower resistance, until that Q1 base conduction threshold has the LED lit. (++++).
Then you simulate sunrise onset by gradually exposing LDR sensor to progressively receive more light upon it.
Now when being covered up from receiving any light, that LDR resistace could be up in megohms of resistance. But as you start gradually exposing it to light, its resistance plummets downward, potentially, eventually reaching as low as 5000 ohms. At one incoming light level threshold, the LDR lowered resistance will have bled off / away all of your base drive that you had set feeding it and the LED will then go off.
Now referring back to an earlier place in time as was marked by the (++++).
You had THEN set that pot just where the LED came on for you, now should you have gone on adjusting a bit further, you would have had the LDR needing a brighter sunrise/ very slightly later on after the sunrise in order to then overcome that different R2 setting.
Modifying the schema by adding a fixed 1K resistor between the battery negative line and the bottom of the R2 pot would electrically limit / prohibit the mentioned Q1 junctions overload or a possible burnt spot on the pots extreme limits resistance element.
Thaaaaaaaasit . . . . .
A carrying forward of the same post #10 schema . . . .
73's de Edd . . . . .
Sometimes, I wonder, just how much deeper the ocean would be, if it was being without all of its sponges.
.