Flash on POINT of clash

[lectricpharaoh]

Here's a possible idea. Those sites that sell robotics stuff on the web often have various sensors you can purchase, such as infrared proximity sensors. I'm not sure if the polycarbonate blades are transparent to IR or not, but it's easy enough to test (try to use your TV remote through a blade, and see if it works). Assuming they are, you could mount some of these in the blade, at various points. If one of them detected something in close proximity to a blade segment simultaneously with triggering of the standard hilt-based clash sensor, you'd flash the appropriate segment. In fact, there is this one here, which seems nice. It's cheap, it's small, and the analog output would allow you to make the flash brightness of any arbitrary segment proportionate to the proximity of a nearby object.

Of course, this is far easier said than done. I can tell you from my extremely modest experience with programming Arduino that even just programming to account for the "bounce" in ordinary momentary switches (which a clash sensor is one kind of) is quite tricky.

The thing is, you don't need to worry about debouncing in this application. In fact, contact bounce would just add to the flicker effect, and could actually be a benefit in this case. That, or the bounce frequency would flicker the LED too quickly for the eye to register the flicker, and all that would be perceived is an increase in brightness (and possibly a color shift, if the clash LEDs were a different color).

...so how would lighting a tiny section (such as in an MR-style blade) of LEDs (likely to be low-powered, aka 20mA or heck, let's say 40mA) make any kind of impact at all on the eyes... ... especially when the REMAINING LEDs (not in the impact zone) are still supposed to be fully on, at full brightness for the "main blade"??

I'm new to the whole electronics thing, but isn't it true you could heavily overdrive LEDs if the duration was kept short enough? If they were flickered at very short intervals, you might be able to coax a fair bit more brightness out of them, no?

[pointoforigin]

Ahh I'm liking this idea. I'm not thinking of several IR sensors in the blade though, surely that'd bring us back to the blocking light problem? But! What if like with the laser idea I had before you could mount them around the edge of the blade holder? Wouldn't have to worry about going through the polyc blade then. This assumes that these sensors have a range of about a meter, which they may not.

Matching that up with a single traditional clash sensor is genius. Again, the programming to make it all work is waaay beyond me at the moment. Maybe in a year or so lol!

I think the problem with overdriving LED's for clash only would be that you wouldn't see much of a difference. Surely you'd need it to be overdriven at least 2x the original to make it obvious at that kind of transient, and that would probably blow the LED. With a string, that's not a repair I'd like to make!

[Matt Thorn]

Here's a possible idea. Those sites that sell robotics stuff on the web often have various sensors you can purchase, such as infrared proximity sensors. I'm not sure if the polycarbonate blades are transparent to IR or not, but it's easy enough to test (try to use your TV remote through a blade, and see if it works). Assuming they are, you could mount some of these in the blade, at various points. If one of them detected something in close proximity to a blade segment simultaneously with triggering of the standard hilt-based clash sensor, you'd flash the appropriate segment. In fact, there is this one here, which seems nice. It's cheap, it's small, and the analog output would allow you to make the flash brightness of any arbitrary segment proportionate to the proximity of a nearby object.

Wow. I give you points for originality. But even without knowing anything about proximity sensors, I foresee numerous problems. Even assuming that you could adjust the sensors with such accuracy that they would react only to something within, say, 2 centimeters, that would mean you would have to space them no more than 3cm or so apart, to avoid gaps. Extending the range would allow you to space them farther apart (and therefore use fewer), but that would result in a lot more false positives, since anything within, say, 10cm would trigger a reaction.

The thing is, you don't need to worry about debouncing in this application. In fact, contact bounce would just add to the flicker effect, and could actually be a benefit in this case. That, or the bounce frequency would flicker the LED too quickly for the eye to register the flicker, and all that would be perceived is an increase in brightness (and possibly a color shift, if the clash LEDs were a different color).

Do you have experience programming microcontrollers? Because either you know a lot more than I do and know something I don't, or you are just guessing based on intuition. If it's the latter, I can tell you that it's a lot more complicated than that. Just explaining why bounce cannot be ignored would take a lot of space and time.

I'm new to the whole electronics thing, but isn't it true you could heavily overdrive LEDs if the duration was kept short enough? If they were flickered at very short intervals, you might be able to coax a fair bit more brightness out of them, no?

We tend to overdrive out sabers as it is. Show our set-ups to a real light technician and they would be be appalled at how we're shortening the life of our LEDs. But those people are dealing with LEDs that are intended to stay on for hours at a time, or even 24 hours a day, whereas as we rarely run ours for longer than 30 minutes a day, and just want them to be as bright as possible. So to get a flash that is really noticeable, you would have to have your LED(s) running pretty dim when not in clash mode.

pointoforigin, I don't even know where to begin with your follow-up. This conversation has yet to rise beyond the "wouldn't it be cool if..." level. We might as well be speculating about how to make real lightsabers.

[pointoforigin]

There's a giant thread about that over at FX anyways

*sigh* When I have the time and money I'll definitely start looking into the practicality of the idea, but for now, it is just that. A giant what if.

[xl97]

Do you have experience programming microcontrollers? Because either you know a lot more than I do and know something I don't, or you are just guessing based on intuition. If it's the latter, I can tell you that it's a lot more complicated than that. Just explaining why bounce cannot be ignored would take a lot of space and time.

this is not true..

bounce 'can' be a desired trait....

usually you de-bounce.. switches.. or sensors where the 'bounce'.. will trigger another interrupt/function/event...

'bounce' can (and is) a useful occourence in some features..

just sometimes you need to code to eliminate it..

I doubt any kind of HOLDER mount would work..because of the flex/bend (and hence false positives) of the blade.

Also mounting to the BH wouldnt really be a wold shoot an IR (or whatever) beam up along the blade.. be more like a motion detection (alarm type trip) than it would be for point of contact..

IMHO.. only a pressure type sensor would give you the options you desire.. however mounting them in a way/place where you can still receive/gauge the pressure.. AS WELL AS still shine light around/through all these new components..

[lectricpharaoh]

Wow. I give you points for originality. But even without knowing anything about proximity sensors, I foresee numerous problems. Even assuming that you could adjust the sensors with such accuracy that they would react only to something within, say, 2 centimeters, that would mean you would have to space them no more than 3cm or so apart, to avoid gaps. Extending the range would allow you to space them farther apart (and therefore use fewer), but that would result in a lot more false positives, since anything within, say, 10cm would trigger a reaction.

If you're using a microcontroller, you can use multiple sensors in tandem. Say an object is registered 6 cm from proximity sensor A, and 4 cm from sensor B when the clash is triggered. Assuming sensors A and B are adjacent, and knowing the distance between them, we can determine the location of the clash with an acceptable amount of accuracy.

Do you have experience programming microcontrollers? Because either you know a lot more than I do and know something I don't, or you are just guessing based on intuition. If it's the latter, I can tell you that it's a lot more complicated than that. Just explaining why bounce cannot be ignored would take a lot of space and time.

Programming, yes. Programming microcontrollers, no. However, I think the bounce is a non-issue in this particular situation for two reasons. First, I am assuming that the sound board has a mechanism to cope with multiple clashes in quick sequence, possibly by terminating the currently-playing clash sound sample/light sequence and starting it over. Second, for the boards that have an available auxiliary input, where the clash effect can be triggered by a button, we can assume that any necessary debouncing is handled by the card. After all, that aux button will exhibit bounce, no?

We tend to overdrive out sabers as it is. Show our set-ups to a real light technician and they would be be appalled at how we're shortening the life of our LEDs. But those people are dealing with LEDs that are intended to stay on for hours at a time, or even 24 hours a day, whereas as we rarely run ours for longer than 30 minutes a day, and just want them to be as bright as possible. So to get a flash that is really noticeable, you would have to have your LED(s) running pretty dim when not in clash mode.

True, the clash brightness issue is something that is likely to be a significant hurdle. However, I've seen enough clever solutions to various problems in my life to write it off as impossible.

[Novastar]

I say build a proof-of-concept (which is actually how the FoC "flash on clash" thing became reality in the first place) and go from there. You seem dedicated enough to the idea, and that is commendable for sure!

All you'd have to start with is a "non-lit" main LED-string blade that used pressure-sensitive material to detect collision/clash in specific areas... pass that data on to your CPU/micro-controller/PIC/whatever, and then see if you could light up the correct "hot spots" on your LED string.

Color wouldn't matter (for the test), and having an already-lit "blade" would be a waste of time for now... this would keep the cost down (although later on, I guess they'd all have to be RGB(?) LEDs in order to produce different-colored responses for the flash on any given choice of your "main blade" color)...

[pointoforigin]

Yeah, that's exactly the idea. Again, time and money constraints. It would start with the most basic concept - so an LED string blade with the clash sensors from the store, see how well that works and go more complex from there - involving pressure sensors, microcontrollers etc etc.

Ideally it'd end up being either an RGB string, like you suggest Nova, or two strings - one being main and the other being flash.

Darth Real Life postpones a lot of things doesn't he?

I'm totally open to someone else trying the idea. I WILL do it, but I will do it when I can, which unfortunately might not be anytime soon.

[xl97]

I posted some sensors you should probably start with.. (not a clash sensor from the store)..

IMHO.. if you want to try and develop something like this...(or other ideas you may have)..

get yourself a dev. platform.

most publicaly available, easy to use with a large community base, would be an Arduino..

I suggest the Arduino Duemilanove, ATmega 328 version.. (this is your DEV platform.. NOT what you will try to stuff in a saber in you final project)..

runs you about $30.00


the sensors.. I posted before: http://www.sparkfun.com/commerce/pro...oducts_id=8685

maybe not THIS specific one..but thats the link to them..


suggestion:

get a few of the smaller 5 in pressure strips...and wire them up in an empty tube (in they fit)...

get your Arduino out and start testing...

not sure what resistance ro pressure (100lbs) you woudl need? that seems like but who knows..especially when dueling?

you could write a simple program that lights up an LED that matches that 'sensor' feedback...as well as 'half light' the leds on the left/right of the led you lighting in correspondance with the sensor..

there is one way to approach it.. good luck.

[TroyO]

This got me thinking about lighting the blade with an end mounted LED (The usual way) and combining it with a LED string style blade to provide the "flash on point of clash" effect.

The two don't seem to be mutually exclusive, it seems to me you could use a base mount LED in combination with string style lighting and potentially double your brightness and/or add in some effects.

It seems like the usualy base mounted LED shines "Up the tube" through and around the plastic and/or cellophane wrap but I don't think it's so much involved in the central open area. (But I could be wrong.) If I'm right, then a string could be in that space and only minimally effect the base lighting.

It could be kind of fun to play with.