You only know the variation betweens samples of the LDRs thay you have. You need to get samples of the LDR at the lower and upper limits of their specification. From Eric's calculation your samples must be closer together than the limits of the specification.
Les

 

You only know the variation betweens samples of the LDRs thay you have. You need to get samples of the LDR at the lower and upper limits of their specification. From Eric's calculation your samples must be closer together than the limits of the specification.
Les

I understand. I will be continuing to test and track when unit is powered on at dusk. Over time, I can make slight adjustments to R7 as may be preferred.

 

hi john,
As it will be outside in all weathers, I would suggest a simple temperature stability test on one or two units.
Say 0C and 40C, use your fridge for ~5C.
E

 

hi john,
As it will be outside in all weathers, I would suggest a simple temperature stability test on one or two units.
Say 0C and 40C, use your fridge for ~5C.
E

Do you mean stability overall? Or, just the dark sensor circuit specifically?

 

hi,
I would consider both tests, it will give you confidence in knowing it will work at the High and Low temperatures.
For sure, some of your clients will ask the question.
E

 

hi,
I would consider both tests, it will give you confidence in knowing it will work at the High and Low temperatures.
For sure, some of your clients will ask the question.
E

Yes. Sounds good

 

FYI, I added a fourth light to the test group last evening. They all came on within a ten minute period from 5:58pm to 6:08pm, which are ideal light levels at dusk for this application. All chatter is gone now of course. I'm ready to order updated boards, the 470k and 15k SMT resistors and capacitors. I already have all other parts. If anyone is interested in reviewing/checking my board before I order them (for a fee of course), please PM me. Thx again all !

 

Can someone check my traces on my PCB to make sure I did not miss anything. I have attached a .pdf with the schematic and PCB image. I want to order boards, but need a second set of eyes before I do.... don't ask me how I know
Thx
John

 

Do you mean stability overall? Or, just the dark sensor circuit specifically?

LDR's are temperature sensitive and the resistance will vary with temperature.
The datasheet doesn't show a graph of temp vs resistance. It's possible that the performance of each units will vary.
More expensive LDRs minimize temperature effects.

 

BTW, I have added R9 (4.7k) to the base of the 2N2222

 

LDR's are temperature sensitive and the resistance will vary with temperature.
The datasheet doesn't show a graph of temp vs resistance. It's possible that the performance of each units will vary.
More expensive LDRs minimize temperature effects.

Dark sensor specifically. Can you tell me approx how much resistance could be expected due to temp swings between 32F and 90F? 1k, 50 k ???

 

Dark sensor specifically. Can you tell me approx how much the resistance could vary due to temp swings? 1k, 50 k ???

I can't, but the manufacture should be able to tell you.

 

You mentioned an "LED flash" at startup.
Did you ensure the pin5 of ATiny is not pulled up internally? If it is, disable the internal pull-up resistor, then use a pull down resistor to pull-down LDD-700 pin3 so it is disabled at power on.

 

You mentioned an "LED flash" at startup.
Did you ensure the pin5 of ATiny is not pulled up internally? If it is, disable the internal pull-up resistor, then use a pull down resistor to pull-down LDD-700 pin3 so it is disabled at power on.

Yes, R10 (1k) resistor has been added to my board. Thx

 

Yes, R10 (1k) resistor has been added to my board. Thx
View attachment 361658

Then you should re-test without the 1-sec delay. It probably doesn't need the delay now.

 

Then you should re-test without the 1-sec delay. It probably doesn't need the delay now.

But be sure the internal pull-up is disabled. Typically, they are disabled at power-up, then automatically enabled afterwards.
So even disabled, they will behave as "disconnected" until the micro-controller initializes. Disabling it and using an external resistor ensures LDD-300 dim pin will be pulled down.

 

Then you should re-test without the 1-sec delay. It probably doesn't need the delay now.

I had neglected to add the 1k resistor that was shown on the circuit that was provided to me on this thread (pins 5,6 & 7). It was on my schematic, but not on my board. It is actually R5 (not R10). I have also connected pin 6 to pin 2 on the LM393 (not shown on this PCB)

 

hi john,
Did you do those operating temperature tests I recommended in post #105?.
Without knowing the result of those simple checks, it would not be wise to finalise the PCB track design.
E

 

hi john,
Did you do those operating temperature tests I recommended in post #105?.
Without knowing the result of those simple checks, it would not be wise to finalise the PCB track design.
E

I haven't yet. I will do that today. These temperature tests are for the overall function of the light and not necessarily the dark sensor, right? Or, should I plan the test at dusk also like I've been doing?
Nevermind. I revisited post #105
Thx

 

hi john,
Just ordered 10 off LDR's of that type, due for delivery this weekend.
I will do some Temperature and Lux tests, let you know the results.
E