I think it best for 10mA, as I will drain the battery to quickly.
Yes these led's are Chinese and are very bright, when I tested it today with 9 volts and 470 ohm resistor.

 

you are going to find your currents between 17ma to 27ma. This is the problrm when pushing the limits with LEDs and power supplies. It probably won't matter in your application though.

Yes, it is a N-Channel MOSFET. Very generic.

I have a tutorial on LEDs in general.

LEDs, 555s, Flashers, and Light Chasers

Hi Wendy,
I've check the Led's and they are 3.1 volts so is your circuit OK to use?
Andrew

 

Board ready make below:

 

OK, here is what I came up with on the merge.

View attachment 88836

It is pretty preliminary, I would test it first before committing to it. The LEDs will not be at full brightness, but they would probably last all night.

I would consider adding a shut down for low battery voltage. I will be glad to add one, but I wanted to get this part done.

If you would like to see how I would do it with a constant current source I would be happy to draw that up too. It would suck the battery dry pretty quickly though, since there would be 18 LED legs, each drawing 20ma for a total of 0.36A.

While I show TLC 555, they are just CMOS 555 chips, a 7555 would work fine too. You could even merge both chips into a dual 555 called a 7556.

Just wondering if you have tested circuit?

 

Hi Wendy and A45,
The circuit in Post #36 is showing damaging very high LED voltages and currents when run in LTSpice simulation.

Please confirm that this circuit is being used.
The White LED's have a Vfwd of 3.2v at 20mA.

Eric

 

Hi Eric,
I was going to start making this circuit of Wendy's, but I was waiting on her to reply to see if it good to use?
So it look No good in LTspice simulation?
I see you add a few components? Why? so LtSpice can workout the current?
Andy

 

hi A45,
I have marked with blue text, the components for the sim.

What is the inductance and resistance value of the inductor on your PCB.?

E

 

220 mirco Henry's.
I would need to electrical measure the resistance of the inductor, as I don't have a clue!

 

hi Andy,
If you do measure the resistance, please post the value

I would suggest we wait for Wendy to reply.
E

 

Hi Eric,
I just measured the Inductor and it read exactly 1.0 ohm.
Andy

 

hi Andy,
Re-run the simulation using a different White LED model, I suspect the first model was not working as expected.
Now using the NSPW500 White LED Model.

The images show the current when using 22R LED series resistors, the current is very high.
and with a 470R resistor, the current checks OK.
The series resistor is labelled {rs} so that I can quickly change the resistance value for the Sim.

The interesting point is that with such a high value of resistor, another LED could be added into each LED chain and the resistor reduced in value.
This would allow you to reduce the number of LED chains.

If you could measure the individual voltage drop of say 6 LED's , when running at approx 15mA I could fine tune the {rs} value, let me know what you measure.

Eric

BTW: As this is Wendy's circuit I feel it would be proper to ask his advice regarding any changes.

 

Thanks Eric for your time, and like you said, let wait for Wendy comment.
If I need to do the led test, we will go from there.

 

Hi Andy,
I would suggest that you do a quick voltage and LED brightness test on say 6 LED's, this will help Wendy and myself to modify the circuit as required.
Eric

 

OK How do I limit current 10mA as that what I using, not 15mA as you stated.
I assuming you what me to test 6 individual led's? how as I'm very new to electronic?
I have a variable power supply so I can tune it to 3.1 voltage exactly for led voltage, the only problem is how to limit to 10mA, what size resistor?

For example
If the supply voltage VS = 9V, and you have a red LED (VL = 3.1V), requiring a current I = 10mA = 0.010A,
R = (9V - 3.1V) / 0.010A = 700

, so choose 820

(the nearest standard value which is greater).

 

hi,
Your formula is correct.
Rs= (Vsupply-Vled)/Iled
but you have miscalculated the value
Rs= 9v-3.1v = 5.9v, so at 0.01A thats is 590R , not 700R

You could use either a 560 R or a 620R, I would choose 620R with a 9V supply, that will give ~9.5mA.

E

 

Will do if I've a 620 if not it be 560 ohms.

 

OK I managed to make a few resistors to get to 617 ohms.
Test a individual led at 9 volts with 617 ohms in series, then measure voltage across led.
Led 1- 2.957 volts
Led 2. 2.954 volts
Led 3. 2.945 volts
Led 4. 2.964 volts
Led 5. 2.959 volts
Led 6. 2.950 volts

 

hi Andy,
So lets say a nominal 3Vled, I would say from that, Wendy will be able to modify the circuit to get the optimum performance.
Perhaps a PM to W, would be helpful.

E

BTW:
Have you seen this DC/DC converter.?
http://www.amazon.co.uk/dp/B00HV43UOG/ref=pe_385721_37038051_TE_3p_dp_1

I use them with 3Watt and 10Watt high power LED's
The Blue LED on that page is a good buy, self powered 4.5V thru 30V.

 

Yes I've 2 of these, one good for 2 Amps and another good for 3 Amps, very cheap why would you own an Workshop power supply station, these are very portable.
I don't see the Blue Leds you're speak off?

 

hi,
It is this one, they are ideal for 12V or 24V battery status indicators.
http://www.amazon.co.uk/dp/B008KY5ULW/ref=pe_385721_37038051_TE_3p_dp_1
E