Page 1 of 4 1 2 3 ... LastLast
Results 1 to 10 of 39

Thread: Smooth color mixing saber with NB v1.0

  1. #1

    Default Smooth color mixing saber with NB v1.0

    I'd been thinking about this idea for a long time now, how to build a color changing saber out of my first custom saber featuring a NB v1.0 (entry level electronics, well, compared to Hasbro, a NB is a formidable toy!).
    At first my saber had one color. Then I made 2 connectors for my Cree XP-E RGrB module (red, green, royal blue) to be able to switch on 2 Colors at a time. For 3 connectors I did not have the space, and moreover it was not "dedicated" enough. I wanted to have real color mixing, not just switching. And then...

    ... I made it at last: a self-developed "color mixing engine" hooked to a NB v1.0. This little piece of electronics can mix the 3 base colors to get (theoretically) any color out of it, and it's close to be true in fact. It's also not magic, basically you need only a few components and it can be crammed into the hilt along with the NB.

    What I like about it is the fact that it's totally in the spirit of Star Wars, a science-fiction world where all technical problems can be solved by a screwdriver and/or a well placed hit of a wench/hammer/whatever comes handy.

    Here is a pic of the hilt (mix of TCSS and parts from other vendors, actually I picked one of the hilts I have , I still did not decide which of them will get this combo) and the chassis:

    Basically the idea is simple: I have 3 potentiometers controlling the gate voltage of 3 p-channel MOSFET's in a low-side driver configuration, which turns them into variable resistances, which are capable of delivering power to the individual LED's. By turning the potmeters with a screwdriver, I can mix the 3 Color in any combination and hue (I was insired by the cut-out scene from the Return of the Jedi, where Luke made final adjustments to his new green lightsaber with a screwdriver before setting out to Jabba's Palace). I will in the next days make some sketches of the schematics so that it becomes more clear what I mean. I also devised 2 different ways to do it:
    - for setups needing more than the 2A spec'd for NB a bypass-technique
    - for standard setups staying below 2A an integrated one

    This is a pic of some of my favorite colors I mixed out:

    I also made a (maybe not so) short video to show how it works. Sorry for the coughing, last weekend the weather around here was rather Hoth and I caught a bad kind of flue.

  2. #2


    Wow. Gorgeous work. I love it when Sabersmiths take what exists and modify it to do awesome things.
    "Bacon and cheese>squeezed lemon." - FenderBender
    "But, ya know...rules are made to be broken." - Yoshi-Taka
    LDM's Recipe For Noobs. Totally Worth The Read.

  3. #3


    Thanks MandoJoe.

    Now let's have the promised tutorial.

    My first idea was to use potentiometers instead of serial resistances to adjust the resistance and thus the intensity of the individual colors. Simple and easy to do...BUT does not work quite as well in reality. On one hand common potentiometers are not designed to carry any power, else they get nasty big, on the other hand you need a high resistance in order to be able to fully switch off a LED, because these besties give out a decent light even with some tens of mAs.

    In this way I came to the idea to use a potentiometer to control the gate of a MOS Transistor. First I tried with n-Mos (well, courtesy of having some at homr), it worked well, but when the saber was swicthed off, instead of no light, sudddenly all LED's switched on fully. I realized that the NB (and if fact all other LED Drivers from Erv) have a p-MOS low side Driver to drive the LED's, which for simplicity's sake means it connects the GND to the LED, while it's always connected to the battery. With some sketching I figured out how this can be used to make a color mixing without sacrificing all the marvelous FX's a Plecter Lab product offers.

    I dubbed this add-on a "direct drive", due to the fact that the whole circuitry including the Color mixing is connected still between LED+ and LED-, see the schematics below:

    Basically the potentiometers can pull the gate of the p-MOS (I used IRLML6402TRPBF, because I had them as leftovers from another project) from LED+ (battery, in this case nom. 3.7V from a 18650 battery pack) to LED-. When pulled to LED+, the transistor turns off, shutting off the LED, when pulled to LED-, it turns on, turning on the particular LED connected to it's Drain.
    The FX's like shimmering and vibrations are still there, because the p-MOS Output stage of the NB will still connect the GND to the color mixing circuitry. If the blade is switched off, the potentiometer will pull up the gate of the transistor to LED+ no matter where you tweak it, because the GND connection gets removed, so the mixed color will "retract".

    I hope you will find this description useful if you ever decide to make a simple color changing saber. I already plan to have a NB v2.0, using the color mixing with a RGBW module, saving off the white LED for flash-on-clash.

    ------------------------- beware, this 2nd part of the tutorial is only for those who want to Experiment ----------

    What comes now is part tried out, part speculations. This "direct drive" can be slightly modified to be able to get more power to your LED's than the absolute max Ratings of the NB of 2As. I tried out this configuration, dubbed "Bypass drive", because the power drawn from the battery(s) bypasses the NB, avoiding any damage to it.

    Difference to the direct drive is that the LED's are now between the Battery+/LED+ and Battery-/GND, the NB "only" controls the gates.
    Cons: one big problem with this is that the FX's are dimmed, i.e. the NB will not be able to fully overpower the potentiometer if turned on fully to LED-, the blade becomes static.
    Pros: Bypasses the NB, you can draw more than 2As from the battery to power your LED's. With a careful selection of the Vgs of the p-MOS, you can even compensate the above described Cons. I tried it with the above mentioned parts (under direct drive), and if the LED's were partly switched on, I could still see the shimmering if the values in the config file were tuned to extreme values, but with more light the blade turned static.

    The main reason I described this method - albeits the Cons - is that - and here begins the speculations, as I never tried this but theory says it must work - this Bypass method can be used to supply your LED's with more than 3.7V, like using 2x18650 for 7.4V or 4xAA's for ~6V, giving you more power for the extreme powerful LED's.

    I dubbed it "power Bypass" and included 2 batteries to drive the LED's. This method could be quite complicated and needs a careful selection of the transistor you use. Due to the fact that the source of the transistor is at say 7.4V and the gate at max 3.7V, you need a Transistor with Vgs parameter below -3.7V in order to turn off the LEDs when the saber is turned off.
    Last edited by Obi1; 05-28-2014 at 10:46 PM.

  4. #4


    Many thanks for the diagrams, it's a very interesting solution.

    Regarding the bypass method and the problem with missing flickering, perhaps it's a problem caused by lack of proper load on the NB output.
    I don't have a board to test right now, but I guess a possible test would be to wire one of the dies through the standard NB LED output to provide load and wire the other two dies through bypass.
    What do you think ?

  5. #5


    Buongiorno Don Se Wion! You brought me to an idea! It has to do with the load all right, point being the working point of the transistors in the NB on one side and in the driving ones on the other side. With the direct drive method the mixing Transistors all all fully open and therefore the current to the LEDs is regulated by the power stage of the NB in the linear mode of the driving transistor. In case the NB acts "only" as a control signal, it depends on the Ugs (voltage between gate and source) characteristics of the mixing transistors how far the NB can "modulate" the current to the individual LEDs.
    I use p-MOS Transistors designed for low voltage digital signals, and has an Ugs= (-1V) (around). This means that it turns fully on very early and is already in the saturation stage at voltages where the NB tries to module it. That explains also why the modulation of the shimmering and co. works when the LED intensity is still low: the Color mixing Transistors are themselves in the linear Region like the NB Transistor.
    Solution might be to match the driving transistor in the NB to that of the color mixing transistors, so that they are both approximately in the same working region.

    What do you think, might it be the explanation?

  6. #6


    That's probably right.
    Perhaps Erv can help us with the model of MOS transistor used on the NB to try another test.
    I'm not that good at recognizing SMD devices

    Knowing which MOS is used could virtually enable us to get the signal from the gate of the NB transistor to drive the additional ones, in the same fashion used to take the 3,3V signal from the capacitor on the left of the microcontroller used to drive an LED that turns off in deep sleep mode.
    Last edited by Don Se Wion; 06-01-2014 at 03:33 AM.

  7. #7


    Me neither, and some of the parts on the NB do not have the markings any more (for sure on purpose). Going by the package does not help, alas I use the same package, but different transistors (Micro3 package). Next time I build a color mixing for NB I will anyway change the package type, these Micro3 devices are damn hard to solder in a hobby-home environment, and then I might ask Erv about some of the Basic characteristics of the MOS he uses (simple Output characteristics curve would suffice).
    Another idea: although I never used a PEX, it's basically the same, just without the gate controlling potentiometers. They could also be used, although that would take up an aweful lot of space - welcome to cram-fu land.

  8. #8


    One thing I forgot to ask.
    What are the specs for the pots ?
    Something in the range of 100K Ohm or less ?

  9. #9


    I used 1MOhm potis, mainly because my first idea was the bypass one and there the potis are connected all the time between GND and battery, drawing a constant current, being 3 of them in parallel would give ~3uA of quiescent current flowing. With the direct drive method you can use any, size matters the most.

  10. #10


    Many thanks again.
    I actually have a saber in store for a friend that wants a purple blade, and we all know that the proper shade of purple is absolutely subjective.
    I guess I'll try your solution to be able fine tuning the shade and be sure he gets what he wants, and perhaps to provide full colors, we will see.
    It will take time, but I'll let you know the results


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts