Yes, something like 10k should do the job.
It will be the current going through R10, though small, is still enough to give the dim glow.

Yes, thank you AlbertHall, I will try a 10k but I was thinking of something like a 1k or even less.

 

This also applies a small voltage to Q1's base, turning it slightly on.

At the time that the MOSFET is off, Q1 is turned fully on, so the extra base current through R10 is neither here nor there.

 

Here's my attempt at simulating that circuit sans the feedback resistor, along with a version using a p-fet that I couldn't make work properly.
I'm not sure what range values for R8 are.

​

 

Yes, thank you AlbertHall, I will try a 10k but I was thinking of something like a 1k or even less.

I'm sure 1k will work as well but it will draw extra current. The resistor just needs to be low enough to stop the glow.

 

I like the idea of using a comparator much better, btw. I think it would be easier to design the circuit so that the led strip would snap on and off with some hysteresis included.

 

I like the idea of using a comparator much better, btw. I think it would be easier to design the circuit so that the led strip would snap on and off with some hysteresis included.

Yes, this was discussed in post #28 above, but we are just expanding the idea of using transistors and a MOSFET.

 

One of the good circuit was from ScottWang, back in post #46

Minimal standby consumption and the hysteresis resistance is placed so it does not give extra current consumption during standby.

 

Yes, something like 10k should do the job.
It will be the current going through R10, though small, is still enough to give the dim glow.

Thank you AlbertHall, I inserted a 10k resistor in parallel with the striplight as you suggested and is no longer glows dimly in daylight and the striplight snaps on at twilight. I have noticed that the voltage level at the junction of R2, R10, Q2 and the striplight with respect to ground is now 11.2v (at the red arrow) and the circuit seem to be working well.


I would certainly at some time like to have a go at ScottWang's in post #46 above, it looks good.

 

I am very interested in constructing ScottWang's circuit that was posted above:

ScottWang's post

but can someone please help me with some of the component values?

I think the LDR will be about 60k ohms when I want the striplight to switch on, but I have been trying to work out the values of the resistors highlight in red in the attached picture but cannot work out what values they should be..
Of course there should be a small amount of hysteresis but I do not a great amount.
Is anyone able to suggest some values I can use?
Also can I use a BC557 transistor as opposed to a 2N3906 as I have loads of BC557's?
Many thanks

 

You can try R2=47K, R3=47K first and adjust the VR1.

 

Here is my take on the simplest way to make this thing.

The CD4093 has built-in hysteresis, the output snaps on-off cleanly.
The output of the NAND is a clean 0 - 12 Volts, perfect to drive any N-ch mosfet.
The inputs of the NAND draw almost zero current.
Its just really dumb and simple.

I have used the lowly 4093 for years as a super simple analog-digital fix-all glue solution for so many problems. You can make timers, oscillators, comparators, level shifters, discriminators etc, all with a few caps, diodes and resistors.

 

Here is my take on the simplest way to make this thing.

The CD4093 has built-in hysteresis, the output snaps on-off cleanly.
The output of the NAND is a clean 0 - 12 Volts, perfect to drive any N-ch mosfet.
The inputs of the NAND draw almost zero current.
Its just really dumb and simple.

View attachment 141261

But you need to swap the pot and the CDS to get the light to come on in the dark.

 

But you need to swap the pot and the CDS to get the light to come on in the dark.

Yes- good catch. Reverse the Pot and CDS for dark-active.

-or you could invert the output with one of the 3 unused gates in the package!
- Add one cap and one resistor and you could make it a dark-activated flasher.

 

You can try R2=47K, R3=47K first and adjust the VR1.

Thanks ScottWang I shall try this.

Here is my take on the simplest way to make this thing.

The CD4093 has built-in hysteresis, the output snaps on-off cleanly.
The output of the NAND is a clean 0 - 12 Volts, perfect to drive any N-ch mosfet.
The inputs of the NAND draw almost zero current.
Its just really dumb and simple.

I have used the lowly 4093 for years as a super simple analog-digital fix-all glue solution for so many problems. You can make timers, oscillators, comparators, level shifters, discriminators etc, all with a few caps, diodes and resistors.

View attachment 141261

Thank you Sensacell, this is a very good and simple circuit, although the LDR and variable resistor need to be reversed. I think it should be:


where the strip light switches on when the it gets dark.
EDIT: I wrote this as AlbertHall and Sensacell were posting above.

 

I have used the lowly 4093 for years as a super simple analog-digital fix-all glue solution for so many problems.

Me too.

 

Thank you AlbertHall, I inserted a 10k resistor in parallel with the striplight as you suggested and is no longer glows dimly in daylight and the striplight snaps on at twilight. I have noticed that the voltage level at the junction of R2, R10, Q2 and the striplight with respect to ground is now 11.2v (at the red arrow) and the circuit seem to be working well.

View attachment 125112
I would certainly at some time like to have a go at ScottWang's in post #46 above, it looks good.

That circuit looks like it should work OK without damping the load.

Is the MOSFET leaky, or does the 2222 have impaired gain?

The only path is via the 100k when the MOSFET is off - you must be very critical about residual glow.

 

That circuit looks like it should work OK without damping the load.
Is the MOSFET leaky, or does the 2222 have impaired gain?
The only path is via the 100k when the MOSFET is off - you must be very critical about residual glow.

Hi Ian.
The circuit I built back in April I must admit works extremely (with a capital 'E') well. I have wired it up to about a 1.5 metre strip light and at twilight the strip light snaps on and off as the sun comes up.
I just wanted to try out ScottWang's circuit above. I also like Sensacell's circuit....I must have a go and build both and see if they work as good as mine.

 

With Sensacell's circuit, the addition of just a few passive components would allow you to use the other Schmitt gates in the IC for flashing the LEDs or dimming them under PWM control.

 

With Sensacell's circuit, the addition of just a few passive components would allow you to use the other Schmitt gates in the IC for flashing the LEDs or dimming them under PWM control.

Yes, thank you for that suggestion, I like your thinking and it's got me really thinking now!

 

Here's a simple PWM add-on :-

The output could connect to one of the inputs to IC1a in Sensacell's circuit.

Edit: ..... but, if connected like that, the logic would need inverting. Also, if PWM is used, the '4093 will struggle to drive the MOSFET gate capacitance, so a simple push-pull current amplifier would be advisable. I'll post a revised circuit.