Not if it solve your problems...use a 8 or 14 pin variant. With internal OSC. By the way have you tried connect the 555 directly to the relay..

Yes I've connected the relay directly to the 555. It works. So, while I wait for the new LEDs to arrive, I'm altering the schematic a little....

 

This schematic:

should work just fine.

R1 should be at the smallest, 100 Ohms per volt of Vcc; so if your Vcc is 6v, then R1 should be 600 Ohms minimum. If R1 is too low in resistance, the timer might get "stuck" trying to discharge the capacitor; the output pin 3 will be low all the time.

If you connect the diode upside-down, pin 3 will be high most of the time.

Got this working.

Thanks!

 

The project was finished a few days ago, and it does what it shall do. Everything fits in a waterproof jar. This is an early prototype. The schematic on a second version is starting to take shape.

Anyway... heres a link to a video... for those who are interested....

 

Recapping your design...

You are using two 9V batteries.

You are using qty 24 LEDs, white (Vf 3.2V) at 20ma each

Your base freq is 0.67Hz, but I'm still not clear as to duty cycle. 92.5% (actually it would be 7.5%, as that is the LED on time). period at 1493ms ≈ 1500 ms, LED on time = 112.5ms, LED off time 1378.5ms

Is this correct?

I am going to draw up a version of what I think you are doing. It will be totally different, as I truly hate relays. It will be a long life version, read that to be 7555 CMOS ICs. I'll aim for minimum parts, but I'm also thinking of constant current regulators because 9V batteries are not that stable in terms of voltage output.

 

Recapping your design...

You are using two 9V batteries.

Yes. 18v total.

You are using qty 24 LEDs, white (Vf 3.2V) at 20ma each

Yes, correct.

Your base freq is 0.67Hz, but I'm still not clear as to duty cycle. 92.5% (actually it would be 7.5%, as that is the LED on time). period at 1493ms ≈ 1500 ms, LED on time = 112.5ms, LED off time 1378.5ms

Is this correct?

Well... I'm using the schematic and values in post #42, so I'll guess that's correct.

I am going to draw up a version of what I think you are doing. It will be totally different, as I truly hate relays. It will be a long life version, read that to be 7555 CMOS ICs. I'll aim for minimum parts, but I'm also thinking of constant current regulators because 9V batteries are not that stable in terms of voltage output.

About the relay... It's gone. As I wait for a new load of LEDs, I've time to draw a proper schematic, that is battery-life friendly. I was thinking of using some of the 25 open end LEDs flasher design.

 

You should test it in water first. To see if the effect is as you wanted. I have a feeling the light will not have so long range as you want in in daylight.

 

I'll take the risk. If it fails the task, I've learned a bit more about the 555, some calculations, enhanced my pcb design, etching and soldering skills...

All in all it's been a fun project.

The next project forming in the head, besides doing some fine-tuning to this, is a 60-something LED light....

 

I'll take the risk.

Will you take the time to tell us, how the water test went

 

Yes, I will. I can maybe include a video, if I remember to bring the camera. Actually, I will tell my girl to remind me to bring it. If I don't bring it, we can blame her. Mohahahaha.

 

Yes, I will. I can maybe include a video, if I remember to bring the camera. Actually, I will tell my girl to remind me to bring it. If I don't bring it, we can blame her. Mohahahaha.

Or some very salty LEDs.

 

OK, first attempt. It occurred to me that a conventional 555 could handle the current, which is 120ma. I also drew a 7555, which can handle the voltage (but not the current).

I have a couple of 555 resources you might be interested in.

http://forum.allaboutcircuits.com/picture.php?albumid=41&pictureid=308

http://forum.allaboutcircuits.com/picture.php?albumid=41&pictureid=1233

 

The last drawing would be nice to make, but is there a substitute for the 1N4733A? I don't think I have that Zener...

The weather forecast on Sunday is rain...

 

Well, if you are using two 9v batteries in series, then instead of using a Zener diode you could just power the 555 timer off the lower battery.

 

Nice thought, I didn't occur to me, but I think it would work. Any zener would work, 3.3V to 8V.

I don't really like the designs much though. Any battery drop at all they will dim dramatically. A 9V battery tends to drop to 7V fairly fast, so I'm told. Not sure I believe it totally, but it is a bothersome concept.

I'm going to post a fairly radical design I think will last much longer. It's not good for the LEDs, but I suspect something else bad will happen to this before the LEDs start wearing out.

 

The last drawing would be nice to make, but is there a substitute for the 1N4733A? I don't think I have that Zener...

The weather forecast on Sunday is rain...

Jens then it comes to zener diodes they are quite interchangeable with equivalent types. Just get the wattage as correct as possible. Somewhat bigger wattage do not hurt. With smaller wattage you need to check if it will be OK in your design (P=U*I). As an example. Bill have used a 1N4733A in his circuit. This is a 1 W 5.1 volt zener. Elfa ( Jens and my component pusher) do not have 1 watt zeners in hole mounted package. But they have a good assortment in 1.3 watt models. And those will work just fine. So just use a BZX 85/C5V1

 

Well, if you are using two 9v batteries in series, then instead of using a Zener diode you could just power the 555 timer off the lower battery.

I thought of powering the 555 from the two batteries. But I think I've read somewhere that it doesen't like more than 15v. Thats where the 7812 regulator comes in.

OK, read you post one more time. Could I use both batteries to power the LEDs (18v), and only one to power the 555 (9v)?

 

.So just use a BZX 85/C5V1

Think I have the BZX 85...

 

I don't really like the designs much though. Any battery drop at all they will dim dramatically. A 9V battery tends to drop to 7V fairly fast, so I'm told. Not sure I believe it totally, but it is a bothersome concept.

Can we use some NPN transistors, one for each LED, like we did on the 25 open end...?

 

Can we use some NPN transistors, one for each LED, like we did on the 25 open end...?

Why this transistor fetishism. You only need one switch to control all the LEDs. In Bills design a P-channel MOSFET are used as a switch. I think a IRF 9510PBF could do fine here. You could perhaps replace this with a NPN transistor for each LED cluster. In this case the voltage drop over the transistor is not that important. Just recalculate the R series for the LEDs

 

It is not the driver that is the problem. It is the sum of the LEDs (which is 4 X 3.2V = 12.8V). The 9V batteries are supposed to drop to 7V over time, so instead of 18V, you now have 14V for power supply. 4V change is pretty large, and causes a current shift from 20ma to 5ma. I know of ways around this, I'll get back with you.