Light detector switch?

[MrAl]

Hi again,

You should really breadboard this circuit to make sure it works
the way you want a circuit like this to work. There is not
much hysteresis even with the diode and resistor so the
turn on and turn off points wont be too far apart and
the circuit is a little sensitive to input voltage variation
meaning the turn on and turn off points will change somewhat
as the battery drains. Before i added regulation components
i thought you should try this circuit first.

Also, i was thinking maybe even 3.6v wont be that good for
driving white or blue LED's directly unless you're satisfied with the
light output even after the cells drain down to 3.3v .
Four cells would drive almost any LED although of course
this would take more cells than originally planned.
There's always back to two cells and switching boost circuit
of course. If the light output isnt as to your liking
you can always change a resistor in the boost circuit if you're
using one of those simple two transistor boost circuits to
restore the light output to max brightness even if the
batteries run down.

Viren:
Nice pcb! Just one note though...
I dont see the resistor R3 showing up on the board that well?
Is this resistor a 'stand on end' mount perhaps?

Ernie:
I hope you like the way it works. I'd like to see a pic of
your finished circuit yes, and maybe when it's lit up too :-)


Take care guys,
Al

 

[ViReN]

Hello Mr Al, Thanks /ubbthreads/images/graemlins/smile.gif
Yep.. its "stand on end" on the PCB.. This is the circuit that i am yet to optimize... I have build it on machine.. yet to build it on a board... I did this because... this will actually guide me to make it on board... Once it works up... i will make an AutoCAD Layout (which will be far more space efficient).... Also... I cant wait to get the SMD's.... there is a lot to do on that front too... may be ... this circuit can be built using the SMD's ... with a Much less space requirement...

-ViReN

 

[Ernie]

Just to let you know MrAl I breadboarded this circuit tonight and it works very well from what I can see.It's bright right now with a fresh set of nimh.I haven't had a chance to see exactly what light level will switch it yet though.

Ernie

 

[MrAl]

Hi there,

Viren:
Oh ok, i thought so :-)

Ernie:
That's great! If you would like to vary the turn off
point you can change R1 (from about 1.5k up to 10k).
The higher values will cause the circuit to turn off with
less light.
Sorry there is no adjustment for turn on light level,
but it's just a simple circuit...although if you dont mind
a different turn off level you can instead adjust R1 for
the turn on light level.

In other words, when you adjust R1 you shoot for one or
the other: turn on or turn off, and settle for whatever
the other level turns out to be.

You'll also want to test when the battery runs down to
see how much the light turn on and turn off levels
have changed to make sure it's acceptable.

Take care,
Al

 

[Ernie]

Hi Everyone,

After letting it go for a couple days and tinkering with different resistors for my light conditions I got it to go on and off just perfect for where I'm going to put it.The value for me ended up being 10k.

The only drawback is it's not very bright for long.I'm not sure how to fix that part.I thought I could use some sort of step-up circuit connected to the load points on this circuit but I guess that's not correct.

Maybe by tomorrow I can post some pictures of it and the little aluminum housing I made for it.

Ernie

 

[MrAl]

Hi Ernie,

Did you check the sat voltage of the output transistor?
Are the batteries fully charged up before the test?

Take care,
Al

 

[Ernie]

Hi MrAl,

I think I did it the correct way.I measured between the collector and emitter on q1 and got a reading of 1.5v.Is this the right points to check?

Thanks
Ernie

 

[MrAl]

Hello Ernie,

Take another look at the schematic. Look at the very
schematic in this thread because i had to update it to
reflect the correct transistor numbering. I had
both transistors labeled "Q2" so there was no Q1 :-)
Now there is a Q1 and a Q2 on the updated schem.

Looking at the updated schem, the correct transistor is
Q2. Check the voltage between it's collector and
it's emitter as you did before.

Take care,
Al

 

[Ernie]

MrAl,

This time I have a different reading.I have .180 across the collector and emitter of Q2 while it's on.I charged the batteries to full before taking the reading also.The last time I said 1.5v but that was when it was in the off mode
and the batteries were low too.

What should the reading be on that?If it's to low then do I need to adjust R3 to a lower resistance?Or change to a different transistor like the FET kind?

Thanks again MrAl

Ernie

 

[MrAl]

Hello again,

That's great news Ernie!
A reading of 0.180 volts is very good, even better then
expected.
When it's in the off mode it's going to depend on
the dc resistance of the boost circuit (when it's
got very low input voltage) and this is not really super
critical. Although the hysteresis depends on having
a decent 'high' voltage when it's off there isnt that
much hysteresis anyway.
As long as you're getting that sat voltage as less then
0.4 volts it will be ok. The efficiency wont be that bad
either due to the low voltage drop.

Since the voltage is certainly low enough you shouldnt
require an FET, but if you're still getting some we could
always try another circuit too if you like.

Probably one of the best circuits you could build to
detect light levels would use a comparator ic chip
like the LP339. You need a few resistors and a
voltage reference diode too, but the circuit comes out
super accurate and temperature and voltage stable...
better than you could ever need. If you're interested
let me know and i'll draw up a circuit as soon as
possible.
Come to think of it, an LM10 op amp would do the trick
and that has an internal voltage ref diode too.

Have you used ic's before?

Take care,
Al

 

[Ernie]

Hi MrAl,

I tried to find that LP339 locally but had no luck.I saw it at digikey though and since I have a couple other things I need to buy from them I will order a couple.

Although the light detector part of this circuit seems perfect to me.It's been going on and off for a few days
at pretty much the same light levels and I haven't charged the batteries at all.

Is there any change I can make it so that it will drive the LED at a more constant level as it drains so it doesn't dim as much?

Also MrAl just to make sure I didn't make a mistake breadboarding the circuit.The load is supposed to be just an LED connected right and not an LED driver connected there?I'm pretty sure it's just LED but then I got to thinking hey maybe I didn't follow correctly what I should do.

Thanks
Ernie

 

[MrAl]

Hello again Ernie,

If you are running without the boost circuit you may
wish to go with four cells in series instead of three.
You'll have to change the resistor for the extra
voltage however.

With the boost circuit it should work with 2 cells.

Take care,
Al