Add a voltage divider to pin 4.
SGView attachment 215132

Excelent, thank you.

 

Your circuit might be destroying the LM3914 because you are shorting its pin 7 reference voltage OUTPUT to ground.
You should never short an output to ground.
Pin 7 and pin 8 make a voltage regulator (like an LM317 adjustable voltage regulator) as shown in the datasheet that you can use to provide a regulated voltage to your light sensor, and the current drawn from pin 7 regulates the LED currents.

 

Your circuit might be destroying the LM3914 because you are shorting its pin 7 reference voltage OUTPUT to ground.

Correct, the schematic that the TS posted cannot be accurate. He shows pins 6,7 and 8 grounded, there would be no reference voltage for the comparators. Right?
SG

 

Correct, the schematic that the TS posted cannot be accurate. He shows pins 6,7 and 8 grounded, there would be no reference voltage for the comparators. Right?
SG

That's right. With pin 6 at 0V then the comparators do not work properly.

 

Add a voltage divider to pin 4.

A voltage divider is better than the diodes. The diodes have temperature stability problems. But it was a fast way to try something.

 

@Audioguru again Thank you for observing that. On my board it was not connected, because i make 1000 permutations there and i rarely and when absolutly necesary update the graphic page. So... the problem was strictly in the image left unmentained. Thank you for correcting that because it would had remained like that for the future generations. I have corrected the graphic page now. I also check the board wires with the graphic page and is fully updated now. Thank you for your attention.
From curiosity i linked very quickly pin6&7 to gnd and only pin1 and 10 was lit, definitely showing an error.

 

Your LM3914 schematic is still missing an important supply bypass capacitor that all electronic circuits need to avoid oscillation.
It is also missing a resistor to ground from pin 7 to set the LED currents which will be about 1mA with only the 12k ohms of the
LM3914 resistor ladder and reference bias current.

 

@sghioto I've tried your sensing circuit with the led as a light sensor and is kind of working. Meaning, it is shimmering/trembling all the leds at once, in standby. When I add light from a superbright led directly over the light sensor (the inversed led) it is dimming all the leds at once. It must be calibrated further and it proves is possible, but i will not get more into it because it was simply a curiosity.
The voltage divider you suggested, I try it and it works VERY well. Definetly an improvement over the diodes, because I can fine tune it now. Diodes are very good for debuging the problem and is a lesson learned with them. I used a single 10K potentiometer as a voltage divider and is performing perfectly.
Ive made a new graphical page as well with the new updates in this awesome research projekt:
https://www.deviantart.com/q12a/art/dBenzin100-ziud-20-circuit-3914-and-LDR-852598799

 

@Audioguru again
I looked after this capacitor in the pdf:

and i find:
" - (Due to possible oscillations at low voltages a nominal bypass capacitor consisting of a 2.2 μF solid tantalum connected from the pulsating LED supply to pin 2 of the LM3914 is recommended.) This ability to operate with low or fluctuating voltages also allows the display driver to interface with logic circuitry, opto-coupled solid-state relays, and low-current incandescent lamps.
"
and another reference to it:
" - The 2.2 μF tantalum or 10 μF aluminum electrolytic capacitor is needed if leads to the LED supply are 6" or
longer.
"
I think is more a specific problem they address by putting that capacitor there, thats why is with dotted lines; but I get your point, for more safety and avoiding future errors from the circuit or as you put it " to avoid oscillation ". I never though of any circuit will randomly oscilate. Is interesting that you point out it is "for all electronic circuits" - lesson learned. I definitely cross over this capacitor in the past but the hell of me if I ever understood its role perfectly. But you bring a bit of light into it and is a step forward (for me). Thank you, Ill integrate it.
My tendency is to super minimize everything in the scope of understanding and learning and focusing. If not, I will be fighting with other complicated errors from other components and never get to the point of what i want to get. That is my reasoning. I will be more careful in the future to put back what i took off in the testing stage. :] I hope but i always forget and get happy when i see it works so i leave the other stuff aside.
I have included C1 into my artpage and put a value of 2-10uF.

But there is also R2 that is lowering All leds current at once. For some reason, in my beginning testing, I could not make leds dimmer from pin 7, but i will give it another try soon.

 

Pin 7 is an adjustable regulated voltage. The current from pin 7 sets the current in each LED to be a regulated current that is about 10 times the pin 7 current.
Your circuit has one 100k resistor limiting the current of all 10 LEDs which share that current. For 2V red LEDs, the current in the 100k resistor is (5V - 2V)/100k= 0.03mA which is 0.003mA in each LED when they are all turned on. Very very dim.

 

Very very dim.

Indeed they were barely open all the time. You have to get over them perpendicularly to see them lit. From an angle they are almost visible. I dim them like that for testing purposes, for non-interfearence, but mostly because they were attacking my eyes and see dark spots. So i dim the life out of them to be able to work and just see the effects they make. When I will put them on the final circuit, I will put a lighter resistence, just to leave them at their peak intensity, or... depends on the further tests, how good they will look. Its not always necessary to be at full brightness. A good designer will balance and use its eye to feel what is good. But thanks for that math you did there, it's always good to know. I will play with the regulator on pin 7 also.
- I noticed some weird behaviour in my circuit today, when replacing the pot with the equivalent voltage divider resistors, and after some hours of looking like an idiot on the thing that didnt work as expected, i realized that 1 or probably more, but for certain 1, hole in the breadboard was not making contact and leave me in twilight zone completly. After figuring that out, everything worked as expected. But what a "shit" experience, pardon my french. :] Most probably I abused that hole with some thick wire/pin sometime in its life. I have to take it apart and look closely where the enlargement i suspect is occurring. Or probably just unclean contact. Who knows who had eaten the marmalade over that hole.

 

mister @ronsimpson , you are definetly my hero and deserve the glory for this circuit fix. You will remain forever in the books of heroes, and in my artpage if you desire. I can integrate your name in there because you really did a great job and make me happy. And this is the prize I can give you. If you want it of course. Be well.

 

Years ago I gave up with the intermittent contacts and stray capacitance between the rows of contacts and wires all over the place on a solderless breadboard. I soldered together all my prototypes on stripboard ever since. Most of my projects were custom-made with only one needed then my stripboard prototype looked good enough to be sold as the final product.

 

@Audioguru again , yes I did it with solder all the way until the solderles appeared. I buy 1 time a single big board all copper plated and etched with needles and knife and cutter, i dont even remember with what anymore, but it was a bit of work, doing it manually, I must certain used a ruler to make straight lines. I made a testing squares board. These days you even can buy one already made. Im happy I was with the wave back then and it was a testing board untold by anyone but in everybody minds. Pretty cool, yes?
I have a image with it: https://www.deviantart.com/q12a/art/dBenzin97-ziua-11-test-board-for-circuits-833379799
I made a separate forum thread in continuation of this one that is dealing with a more complicated problem I face. Well,3more.
See it here: https://forum.allaboutcircuits.com/threads/3-important-problems-to-resolve.172235/