My green LEDs worked OK, and were brilliant with 3V. The base design said you could even add more LEDs than 3. They were OK at 1½V, but nothing special. I suspect you'll like em. BTW, what are going to use them for?

 

Settled for a Tant at 35V for the cap.

That'll be fine, but for longest life, charge it up slowly to it's working voltage using a 20k Ohm resistor until it's leakage current is less than 1mA.

Tantalum capacitors tend to fail rather spectacularly when they're not conditioned properly.

 

Bill,
To funny. At first I couldn't figure out who BTW was. I went through all the post and could not find anyone with that login. Then I read your post again and it dawned on me....it stands for "by the way". Or at least I think it does. OK it's late here in California. Anyways I want to make some signage that draws peoples attention. I figured if it blinked at them they just might read it.

 

I wired up the circuit today using 2 AA batteries. The leds come on but no flash. I get no lights using 1 AA battery. Bill, in your post you said you were suprized that your bread board came up using one battery. I double checked the circuit for wiring error and it checks out OK. Could this be due to the 1.0uF tant cap not behaving properly? Thanks in advanve for your time.

 

It could be, if you measure pin 2 on the 556 what do you see?

 

I have some metalized poly film 1uF/50V capacitors that have no leakage current and they are small with 0.2" lead spacing. They are about 0.7cm high.

1uF (marked 105) ceramic capacitors are smaller and also have no leakage. But their value is not accurate and it wanders all over the place.

 

For this application ±50% is OK, or even worse. Because Radio Shack stocks them I got a bunch of tantalum capacitors for 19¢ each from Tanners. It's not that I like RS, but I'm sticking to the standard started in the AAC experiments book.

I used some electrolytics in my design with no problem, but I haven't quite got the JT working. I tried substituting a choke for the inductor, don't know if that's the problem or not, but I working it. I'll put the original inductor type to see if that was it.

Basically I shelved the project for Christmas, then 3 computers died that I have to fix.

 

Good news....I'm up and flashing. It was the tant cap. Bill you were right in that a standard LED is plenty bright enough and the flash rate is good. Next test is for battery life. Time will tell and I'll be sure to post with that information.
O.K. with all that said what would happen if I wanted 6 LED's flashing instead of 3? I tell you people in our business are never through engineering anything

 

No problems, though it will shorten battery life some. Did you stick 6 LEDs in series or what? BTW, "D" cells will last pretty long.

 

I think our OP wants to know how to modify the circuit to run 6 LEDs.

No matter how you slice it, twice the power will be required from the batteries. Since the power is drawn at short peaks just to charge the inductor, the battery life will be less than half as it would be with 3 LEDs due to the battery internal resistance.

 

The basic joule thief said something about supporting more LEDs in series, which is why they are so brilliant at 3V. It would be possible to add 3 more, but as I understand it there is less wiggle room for battery drop. The flip side is it won't actually take more current (I think).

 

The basic joule thief said something about supporting more LEDs in series, which is why they are so brilliant at 3V.

Yes, it will - but there will be less power available for each LED.

It would be possible to add 3 more, but as I understand it there is less wiggle room for battery drop.

Yep.

The flip side is it won't actually take more current (I think).

If you want the same brilliance out of the additional LEDs, you will have to spend porportionately more time charging the inductor. If the inductor was borderline (near saturation) to begin with, you'll need a larger inductor. This means that it will require more current to generate the flyback voltage - which is basically what has been created; a flyback step-up converter.

 

About 30 years ago I had an LM3909 flashing a dim old LED for about 2 years on an alkaline D size cell.
Your modern green LEDs might be much brighter and will not be too dim if they are connected in parallel to share the current. The forward voltage of the LEDs must be 2.0V or less or the LEDs will slowly dim and the battery cell won't last long.

 

The LM3909 has been obsolete for a decade now, and are unavailable unless one wishes to pay ridiculous sums for them.

Paralleling LEDs doesn't work well unless they're carefully matched, which most don't have the patience for. It's certainly difficult to do in a production environment, even if you have automated test equipment.

The easy way is to use resistors to limit/balance current, but that's not really very economical in a battery-powered device where you're trying to eke out every last bit of juice from the supply.

Using multiple LEDs in series is certainly much easier; you don't have to worry about the individual Vf's.

But our OP wants to double the number of LEDs in the circuit - and I'm sure that he thinks it could be done using the same amount of power.

There is no "free lunch".

If you want double the light output, it will take double the power input. If your design is not good, it will take more than double the power input to double the light output.

 

Greetings from not so sunny California (rain today). The good news is it was working. Last night it was showing signs of dieing....this morning it's dead. This is with 3 LEDs in series and using 2 AA batteries for 3V. Any ideas on why there is such a quick drain on the batteries? P.S. go easy on me I'm a girl not a boy

 

Somewhat similar to a joule thief, isn't it? That sucker flashes at 2.2Khz. To the OP, do you want want it that fast, or something slower?

I've been working on a 1Hz flasher that also uses a CMOS 555, with similar design goals. Still have to breadboard it though.

http://forum.allaboutcircuits.com/showpost.php?p=105259&postcount=38

"Joule Thief". Now THAT deserves a place in the annals of classic electronicsdom. Brilliant!

(You'd better copyright that soon!)

eric

 

"Joule Thief". Now THAT deserves a place in the annals of classic electronicsdom. Brilliant!

(You'd better copyright that soon!)

Too late!
http://www.emanator.demon.co.uk/bigclive/joule.htm
Brief excerpt:

In the November 1999 issue of EPE (Everyday Practical Electronics), a small and intriguing circuit was published in the Ingenuity Unlimited section by Z. Kaparnik.

Big Clive called his version the "Joule Thief".

 

Greetings from not so sunny California (rain today). The good news is it was working. Last night it was showing signs of dieing....this morning it's dead. This is with 3 LEDs in series and using 2 AA batteries for 3V. Any ideas on why there is such a quick drain on the batteries? P.S. go easy on me I'm a girl not a boy

I'll have to finish mine to get some readings, but it should have worked (least according the base article we based it on). The add on stuff added about 2µA total. I'll finish mine up and get back with you.

Exactly how long did it last? Did it get wet or damp?

Since you're in California and I'm in Texas I can say this, are you single? I am.

 

Bill...The first 48 hours things were looking good in that the LED's were still quite bright. Then they started to dim although the flash rate stayed pretty constant. This morning they were dead to the world. Total time of experiment approx. 72 hours. I know I said it is raining in California, but no it didn't get wet or damp, it's on my kitchen table. I'm going to replace the 1uF Tant cap with an electrolitic, as I think it is a problem child. When I put in new AA batteries it takes awhile before it starts to flash ( about 3 hours ). And yes I'm single and loving it.

 

Brand new name-brand AA alkaline cells have a rated capacity of 2500mAh but only if the battery discharges too low for the circuit. The capacity is about 1000mAh when the battery voltage drops from 3V down to 2V.
1000/24= 41.6mA per hour for a total of 24 hours.

The circuit steps up the voltage. Then the battery has much more current than the LEDs use.