How many lights (LEDs)? Once you have a plan boards can be made very inexpensive. The trick is selecting the parts to run with.

Ron

One LEDfor the most part however I do have another type box I make that needs blue light

 

I just caught the name... Torpedo Chief?
<Off Topic> Spent my last 22 years before retiring working with Naval Reactors but prior to that worked with the MK46, MK48, MK48ADCAP and finally the early MK50 torpedoes. Reactors was much, much more fun. </Off Topic>

I enjoyed the image you posted, really quite good. Light houses have a certain lure to them. I grew up on the Atlantic coast (Long Island) and always had an attraction to light houses. I am sure the use of LEDs will offer some nice lighting effects to your work Chief.

Ron

I spent 22 years riding submarines. The last project that I got help from this sight became the worlds only anamatronic tank commander in 1/16th scale and it still is.

 

I can draw one for you tonight... if you're patient and wait a few hours

Thank you so much! That would be great!!

 

Does a lighthouse ever use two colors, such as red and white?

John

 

Howdy, Torpedochief! Long time no see! Panzer still working?

Since your requirement is to operate from 3v, any solution you use will require boosting that voltage higher, as typical white and blue LEDs will have a Vf (forward voltage) of 3.4v to 3.9v at their rated current. Less voltage than that, and the LEDs will either not light at all, or be very dim.

While staying with "traditional" linear circuits, such as a 555 timer, comparators or op-amps may seem attractive due to so many options being available, difficulties (and expense) occurs when changes in the operation/function is required; you may have to design another circuit board or modify many existing ones, etc. Using a microcontroller minimizes the need for most external components, as most functionality is accomplished via software.

Your application might be as simple as the microcontroller, three capacitors for smoothing power, a small power MOSFET, an inductor and a diode for the boost converter, and a small PCB to mount the components on that has power and output connections - and perhaps connections for switches to increase/decrease speed or increase/decrease maximum brightness.

I suggest that a linear ramping up and down of the light intensity is not what is desired. Lighthouses frequently use a rotating parabolic mirror along with a Fresnel lens to focus the beam of light. When the assembly is rotated, at a distance the light is perceived to go from very dim to suddenly flash brightly for a moment, then rapidly dim again.

John: the Fort Gratiot lighthouse at Port Huron, MI alternated white and green back in the 70's.

 

Hi SgtWookie, long time too.

The reason for my question is that needing two colors would complicate a 555-type of circuit. I suspect it could be wired to source one color and sink the other, but then getting the duty cycles right could be a headache. That is why I suggested the simple 10F series (actually any series) of microcontroller. I thought his needs for "small" were smaller than what it seems he has space for.

Regards, John

 

John,
Appreciate your concern about the dual (more?) colors. Someone (can't remember who offhand) had/has a website with a schematic of flashing red eyes - alternating red LEDs that ramp up and down in brightness; useful around Halloween to put in the bushes, scaring the bejeezus out of the little kids... however, that circuit would also require a boost of the 3v.
edit:
Ah, just remembered - it was on Bill Bowden's website:
http://www.bowdenshobbycircuits.info/fade_led.htm
Fading Red Eyes.

Alternating colors could be implemented with a microcontroller by using a couple of MOSFETs (or transistors, if you don't mind the extra parts count and the base current) to supply the ground to one LED or the other, while a 3rd MOSFET is used for the boost function.

 

If one wants a fairly small boost converter, I suggest the Sparkfun one: https://www.sparkfun.com/products/10968 There price is not much more than the cost of the parts.

The only thing odd about it is the +5V comes out both ends, but the ground/common doesn't. It is easy to incorporate on a regular PCB and take care of the ground at the same time.

Edit: If you are really size-restricted, as I was recently, I made a 2S LiFeP pack from 123A -size and used a regular linear regulator (LP2989). Batteries are going to be the biggest space hogs in almost any design.

John

 

Check this circuit which uses a single 1.5V cell battery and has 2 year battery life!

 

@torpedochief , as soon as you tell me which circuit you'd like me to draw its PCB for you, I'll get on it.
The first circuit is cheaper, but uses a 9V battery. The second one uses more components but uses a single 1.5V cell battery. Also, it doesn't flash gradually like the first one.

 

Howdy, Torpedochief! Long time no see! Panzer still working?

Since your requirement is to operate from 3v, any solution you use will require boosting that voltage higher, as typical white and blue LEDs will have a Vf (forward voltage) of 3.4v to 3.9v at their rated current. Less voltage than that, and the LEDs will either not light at all, or be very dim.

While staying with "traditional" linear circuits, such as a 555 timer, comparators or op-amps may seem attractive due to so many options being available, difficulties (and expense) occurs when changes in the operation/function is required; you may have to design another circuit board or modify many existing ones, etc. Using a microcontroller minimizes the need for most external components, as most functionality is accomplished via software.

Your application might be as simple as the microcontroller, three capacitors for smoothing power, a small power MOSFET, an inductor and a diode for the boost converter, and a small PCB to mount the components on that has power and output connections - and perhaps connections for switches to increase/decrease speed or increase/decrease maximum brightness.

I suggest that a linear ramping up and down of the light intensity is not what is desired. Lighthouses frequently use a rotating parabolic mirror along with a Fresnel lens to focus the beam of light. When the assembly is rotated, at a distance the light is perceived to go from very dim to suddenly flash brightly for a moment, then rapidly dim again.

John: the Fort Gratiot lighthouse at Port Huron, MI alternated white and green back in the 70's.

Sarge!!!!! Hope you were still around!! As,always guys on here top notch folks. Yep panzer doin great. Books are published,soon to be mini series.

 

Howdy, Torpedochief! Long time no see! Panzer still working?

Since your requirement is to operate from 3v, any solution you use will require boosting that voltage higher, as typical white and blue LEDs will have a Vf (forward voltage) of 3.4v to 3.9v at their rated current. Less voltage than that, and the LEDs will either not light at all, or be very dim.

While staying with "traditional" linear circuits, such as a 555 timer, comparators or op-amps may seem attractive due to so many options being available, difficulties (and expense) occurs when changes in the operation/function is required; you may have to design another circuit board or modify many existing ones, etc. Using a microcontroller minimizes the need for most external components, as most functionality is accomplished via software.

Your application might be as simple as the microcontroller, three capacitors for smoothing power, a small power MOSFET, an inductor and a diode for the boost converter, and a small PCB to mount the components on that has power and output connections - and perhaps connections for switches to increase/decrease speed or increase/decrease maximum brightness.

I suggest that a linear ramping up and down of the light intensity is not what is desired. Lighthouses frequently use a rotating parabolic mirror along with a Fresnel lens to focus the beam of light. When the assembly is rotated, at a distance the light is perceived to go from very dim to suddenly flash brightly for a moment, then rapidly dim again.

John: the Fort Gratiot lighthouse at Port Huron, MI alternated white and green back in the 70's.

 

Ok, here's a preliminary of the PCB, I made it as compact as possible... see if you can fit it in most of your models. You need to buy the connector for the battery, and use extension wire from the PCB to where you are planning to physically place the LED.

Parts list is (one of each):

• resistor 470 ohms, 1/4 watt
  
• resistor 33 Kohms, 1/4 watt
  
• transistor BC547C, or 2N2222, or 2n3906
• electrolytic capacitor 100 µF at 16V (or at 10V, if you like)
• 555 IC
• 1N914 diode
• LED

I'll try to modify the circuit later on for it to flash two alternating LEDs, maybe then you could use two colors.

Let me know if there's anything else you need.

 

Hey everyone...

I'm trying to make that circuit work with two alternating LEDs instead of just one. So I figured I'd use the same circuit, but with a totem-pole configuration... aaaaannndddd it didn't work... at all...
I do understand how a totem pole works, but it's obvious that I don't fully understand the innards of a 555.

Here's the circuit, in case any of you would like to tweak it and see if you can bring it to life, and BTW, the circuit works fine if D1 is removed.

 

Hey everyone...

I'm trying to make that circuit work with two alternating LEDs instead of just one. So I figured I'd use the same circuit, but with a totem-pole configuration... aaaaannndddd it didn't work... at all...
I do understand how a totem pole works, but it's obvious that I don't fully understand the innards of a 555.

Here's the circuit, in case any of you would like to tweak it and see if you can bring it to life, and BTW, the circuit works fine if D1 is removed.

Your output circuit is loading the charging circuit too much. I changed the 33k resistor to3.3k and it now oscillates. Of course, the frequency is now way too high...

I then changed your buffer transistors to Darlington's and was able to go back to 33K. Note that the LED fading will likely look different with the buffer transistors added (and even more so with the Darlington configuration). A Sziklai pairs would be better than Darlington pairs.

p.s I hate seeing large electrolytic caps in 555 circuits!

 

Your output circuit is loading the charging circuit too much. I changed the 33k resistor to3.3k and it now oscillates. Of course, the frequency is now way too high...

I then changed your buffer transistors to Darlington's and was able to go back to 33K. Note that the LED fading will likely look different with the buffer transistors added (and even more so with the Darlington configuration). A Sziklai pairs would be better than Darlington pairs.

p.s I hate seeing large electrolytic caps in 555 circuits!

Nice!
Would you mind posting the asc file?

 

Nice!
Would you mind posting the asc file?

Here tis'.

 

Here tis'.

I assume that the darlington pairs could be replaced by single darlingtons such as MPSA63 and MPSA64 ?

 

D2 is working almost fine, oscillating between 2 mA and 9mA, but D1 is oscillating between 7 mA and 14.4 mA ...
I don't think this circuit is the right configuration... I'm thinking about using a voltage follower at the C1-Thr node, and take it from there...

 

Ok... using a voltage follower things work perfectly with one LED... now, how do I do the reverse? that is, how do I make a circuit whose output would be the reverse of this one? If output one is 6V, then output two should be 0V, and then vice versa...