Hey guys, been a while! I came across this schematic while getting ready to build a halloween costume for my buddy's kid. It is the lighting section of a Ghostbuster's 'proton pack'...someone was nice enough to lay it out.

I would like to change the incandescent bulbs featured to 'super-bright' LEDs...my question is, should I simply add a dropping resistor before each LED, and just go with this the way it's written, with the same value of base resistor? Since I assume the incandescent bulbs would draw a LOT more current, the transistor aspect should be fine as-is, right?

Also, I am assuming that about any NPN small-signal switching transistors should do this job well...right? (hafta order parts...)

Thanks!

 

The lamps use 100ma at 6 V. The 4401 transistors have a MINIMUM gain of 100 at 150ma. That makes them pretty good transistors for this job. You are close to the right answer but...

The question is, what current do the super bright LEDs need and how much voltage for the LEDs. "Super Bright" doesn't tell me these things.

 

Hi Mike,

The question is, what current do the super bright LEDs need and how much voltage for the LEDs?

Ditto. Do you have a part number or datasheet for the LEDs you plan to use?

You can stick with the 1N4401 transistors, but I prefer having a lower parts count myself. If each LED or LED string (if you opt for more than one LED per CD4015 output) consumes less than 500mA, then I'd suggest using a couple of ULN2003 (7 outputs each, 14 total) or ULN2803 (8 outputs each, 16 total if you want to expand). These will allow you to eliminate the base resistors. You'll still need the LED resistors of course.

To help everyone picture what the OP is trying to do, here a link to a video of the prop: http://www.youtube.com/watch?v=GrMQgS5d6ZI. Skip ahead to 1:50 to see the lights.

While some of these are certainly bright, I don't think you need to go too bright. The four red circular lights in the back are probably close to 100mcd-300mcd depending on the viewing angle. Much of the rest appear much brighter, perhaps from 500-2000mcd at a guess. I'd recommend about 1000mcd (or less) OR higher if you're going to add a diffuser. Much brighter than that and you'll be blinding anyone that looks directly at the lights, especially indoors or at night. You can also diffuse brigher LEDs by lightly sanding their tops.

 

Hi, thanks for the replies! I just snatched a few white LEDs, 5mm, from RS. 7000mcd, with a forward current of 25mA. For the cyclotron.
Now, these are 3.6v MAX, so I'm not sure if they're ideal...could run a 270R, 1/2 watt resistor and try 'em out! The 4 'cyclotron' windows have colored gels in them...they won't need much to illuminate, I don't think.

The above are only for the 4 lights of the 'cyclotron' thing...for the power cell indicator LEDs, I'd like to just run typical amber run of the mill ones...they will be bright enough! (assume 20mA). Probably will use 10 or so, and set it up just like in the video. I have TONS of transistors meeting that 4401-type specs, so if they will work ok, I don't mind using up the resistors with them (have tons of those, too!).
I am still sourcing the 4015 shift registers, ack....

 

Hey, I got it to work with LEDs

Only problem now is...I am using 10 LEDs with the shift registers (the 4015)...I need to make RESET High momentarily to return the count to zero and start over (the lights just cycle thru over and over).

I thought of making one more output section like the lights, but just using a transistor to trip the reset to H....will this work? Or will it "stick" high?? Worth a try, I guess - but with these CMOS chips, I am worried I'll fry something, he he.

If I DO take the last light output from say pin 4 of U2A, where should I take the input for the decade counter? I guess what I am asking is.....what makes the RESET high in the given schema????
Thanks, almost there!!!

 

U2B pin 2 is what causes the shift registers to be reset, and it also clocks the 4017.

Here is a datasheet for a CD4015B:
http://www.ti.com/lit/ds/symlink/cd4015b.pdf
Pin 4 is Q2A. If that is the last light you want to flash, then connect RESET to Pin 3, Q3A instead of pin 2, Q4B as shown in the schematic. Don't connect the RESET inputs to more than one OUTPUT, or you will have problems.

 

Thanks, SgtWookie, you're always very helpful!
I'd have tried to work this out better myself, but time is of the essence - appreciate your clarification!
I need to sit down and just mess around with these registers...doing what I'd call an "early termination' because of not needing all the registers. So, making the next pin in sequence momentarily high will reset the system, and have only 1 output connected to the reset line....got it!

 

Make certain that you disconnect pin 4 on U2 before you connect the reset/clock line to U2 pin 3. Otherwise, the two output lines will "fight" each other - to the death! The circuit won't work and the IC will quickly become damaged due to the high current flow.

Always make certain that you don't have two OUTPUT pins connected together.

Also, with 4000 series CMOS IC's, it is very important that all INPUT pins have a current path to either Vdd or GND; otherwise they can oscillate unpredictably. Unused OUTPUT pins should be left disconnected.

 

Thanks! I took it all the way out to 12 lights (the pins I'd referred to were on U2B)...it works great now, and seems to be stable. I tied U2B, pin 13 (which would have been 'the 12th light') to reset via another transistor used as a switch, and didn't use pin 2 on that chip. So it resets great!
Next up is the 4017, which looks a lot more basic! I think my friend's son will enjoy Halloween