Please Welcome the PRIZM v5 sound boardThe manual is in good shape too but not fully finished, it should be there by the end of the week.
As for availability, we have already manufactured some (USA) so they will be for sale "soon". Like for the NB, I've been busy with this release, so EU manufacturing will start during May(disclaimer : you'll read similar text and statements if you compare with the Nano Biscotte V4 annoucement thread from yesterday. I'm guilty of copy / paste again but to my defense, that's because some features and new implementations are identical between the 2 boards ! )
dimensions : 40.5 x 23.5 x 5 mm - thinner with smaller yellow capacitor with better placement for easier hilt integration- [NEW]
Dual Digital Sensor motion detection. Identical to CF MotionFusion™, very high swing sensitivity and vibration / battery ripple immune- [NEW]
Mute during power off. Optional silenced power-off sound when MOTG is engaged- [NEW]
Reverse polarity & over voltage protection- [NEW]
Enhanced deep sleep with 0.1 mA consumption (12 months of shelf time with a 14500, 36 months with a 18650), no kill key really needed.- [NEW]
3 channel color "mixing" with FlexiBlend™ - 4A per channel- [NEW]
Pulse Fx- [NEW]
RGB Neopixel / WS2812 (and alike) support with power saving, adjustable timings and dedicated data pad and resistor footprint- [NEW]
Ledstrip power saving StripOff™- [NEW]
Ledstrip data output/line with onboard resistor (aka LS pad)- [NEW]
Ledstrip localized & Accurate Blaster Bolt deflection AccuBolt™ and MultiBolt™- [NEW]
Omnisabers™ support / Omnisabers™ ReadyLegacy FEATURES
- Up to 10 color profiles allowing the real time color change within a font. The use of a momentary is preferred as more natural to switch color, but I've added support for latching too.
- Up to 6 sound fonts. Now with automatic font amount detection, board ships with the PC3 usual set of 3 fonts, but you can have only 2 or 6 if you like - Comes with Lightmeat, Greymeat and Darkmeat
- Up to 16 swing and 16 clash sounds
- Multiple boot sounds
- Force Fx
- True 16 bit audio
- Designed for single li-ion cell
- FoC Mixing Techniques allowins mixed color FoC as well as traditional and popular "On-top" type Flash on Clash
- clash and swing motion flow limiters
- momentary and latching switch support
- off delay APOP and Legacy "NovaStyle" Anti power off (using Aux switch)
- progressive ignition / retraction (of course)
- quick on / quick off
- flicker / shimmer / pulse
- R.I.C.E. configuration system (wired or wireless)
- adjustable drive to remove the resistor in some (rare) cases
- changeable sound fonts, drop in support of CF/PC fonts, extra sound files are ignored.
- Override master file + individual configuration files for font specific settings[MuteOff™]
The Mute-on-the-go™ has become highly popular especially with saber classes (and was an idea and request from Novastar, for his saber & fencing lessons back in like 2007?). When implementing this, I left the power off sound playing to indicate the saber sound was still "ok", especially when it was being tested for a show.
Then we got recurrent requests from people with bad WAF ;-) especially when playing at night with their saber and after being threaten of divorce several times. I think Saber League teachers were in the same boat too. I personally like to sometimes hear the power off sound to be SURE the saber is still "ok" when leaving the MOTG mode (and because silent sabers have now become daunting but it's also cool to allow the saber to be silent on demand, especially when you have 25 students in the dojo or studio turning their saber off at the same time :-)[4 AMPs / channel HUMF FAP OMFG yumnomnomnom what an upgrade !!!!]
Well not really. Over the past years I've been fighting some lack of common sense with direct driven LEDs, in particular the fact most people don't measure their LED and don't calculate resistors and just expect it to work. Based on this, I tried to keep LED driving "reasonable" for beginners and rated the PRIZM (and NB) output up to 2A. Heck, as someone posted yesterday during this current discussion (pun intended), the plague "I just use 1ohm and it works" is still plastered everywhere on Facebook, so it's legitimate (IMO) to question this disinformation.
First of all, as a precision, the PRIZM doesn't "output" 2A or 4A, just like with Watt ratings of speakers, a lot of people think that it's forced out while it's "requested" by the load and the board would drive *up to xxx A*
Second, the driver part (that has been used for like 4 years now at least) has its maximum current rating at 6.5A and it's safe to use it 4A.
Overall, if someone wanted to use a BBB tri-cree on a PRIZM, I would however suggest to maybe split the load between a BB pair and the last B die or even all split on the 3 channels but it should be fine to handle all 3 on a single channel.[New motion sensor]
It's been a while since I wanted to harmonize the boards all together and it's been mentioned during the release of CF7. The real difficulty was finding a proper time slot to accommodate the board production (EU / USA), distribution (EU / USA) and overall growing demand (worldwide). While it's a cool update which brings the former super accurate motion detection brought to you via CF7 2 years and a half ago, I don't call it an innovation per say, and I'm happy to provide it to the whole range of board (planned for Pico Crumble too).
It also brings a very good immunity to power supply noise and infamous errant swings, but this shouldn't be an excuse for not measuring your leds nor verifying what your battery drain is (and not just from the "datasheet"). To be noted that some manufacturers use binning for both voltage and hue and reading the bin code is much more accurate than just using the typical value from the datasheet (which is just a ballpark).
The motion detection uses the simplified parameters from CF, a motion "gate" that we usually set between 50 and 70 (not critical), no more general sensitivity to mess up with, and a swing threshold separated from the clash threshold.[Omnisabers ?]
I've been using our former multi-saber wireless stage performance system for some year with Saberproject and during the period they designed their own wireless transmitter, I worked with them to enhanced the possibilities of wireless live performance with lightsabers. This has been achieved in the Crystal Focus firmware and is now progressively ported to other boards.
There's both a known demand for this for quite a while (saber leagues, saber groups, stage demo, booth demo, conventions, artistic fencing) and personal interest to pursue that work, I just couldn't address that so far but be sure I haven't ignored it :-)
How does it work ? Basically, as I consider trying to stream the sensor or the digital audio from the saber is a bad idea (mainly for latency issues but not only), I rather chose (in 2006) to send saber actions to the computer to control a polyphonic sampler.
I'm currently working on my new omnisaber "dongle" to revive both hardware and software initially designed 10 years ago but in the meanwhile, I'm ensuring the new gen of boards can be "omnisabers-ready". The saber actions are exported on the board serial port / RICE port, just like any of the diagnostic information and remote configuration.[Neopixel (and alike) support]
The NB allows to drive a WS2812 (and compatible) blade up to 150 pixels. The strip data line is handled by the pad L3 when the board is configured for using a ledstrip. We spend quite some time to build a neopixel driver that uses zero CPU along with an abstraction layer in the firmware so that the blade output / update is not different between high-power LEDs and a strip.
That means the strip gets the same exact refresh rate, flicker look & feel.
The driver has also adjustable timings, as some compatible pixels use a different chip (APA106, SK6812) have different timings, we therefore provide adjustable params in the override.txt file to define the strip protocol timing with a resolution of 25.2 nanosecond.
The board *addresses* 150 pixels max. With 2 strips you could have 300 leds.
So yes you can wire 3 strips NP. The current is crazy and there's more shadowing than expected but it's been tested by Alex Gordon.
I think you all got the point, I tried to separate the question of the addressed pixels vs the total number of pixels in the blade, which are 2 different things.
* To be noted that one significant difference between high-power LEDs and a LED strip when using the FlexiBlend™ engine is that the latter has a 10 bit resolution (0-1023) to define color with a significantly high resolution (billions of colors).
The strip uses only 8 bits to define color, therefore there's a much more noticeable quantification (ie "step") effect on smooth effects like pulse. Conversely, subtle effects like shy flicker become less visible or "obvious" due to the lower resolution of the strip. As a result, certain effects will need to be exagerated when using a ledstrip. Color remains defined on 10 bit just like our other direct drive boards however expect no visible color change by tuning channels with only 1 or 2 steps. The strip resolution being 4 times lower, steps of 4 are required to obtain a color change on the strip. For Pulse Fx (now on NB too) afficionados, the low brightness side of the ledstrip is also 4 times less smoother, hence digging too hard with the flicker depth will produce some stepped flicker. A depth of about 60-70% remains ok to the look (basically avoid the lower quarter).* Neopixels & Power saving
In the global effort to reduce the need of a kill-key (would it be during a convention or for some shelf time) we wanted to preserve our super efficient deep-sleep mode that uses only 0.1mA of current, and which allow more than 1 year of shelf time with a charged 18650 (tested).
The pixel chips use some current even when off, about .9mA each. With a trimmed blade using for instance 2 strips of 140 pixels, we have a good 250mA of current draw when the blade is off, which we found ridiculous to not address for deep-sleep.
As a result, we provide a "remote power control" over the strip by hooking the strip(s) negative to 2 of the on-board LED drivers (aka on-board PEx). Both channel 1 and 2 are enabled at the same time to split the load of the strip when the blade is on. Once the blade goes off, channel 1 and 2 will keep providing the idle current to the blade until deep-sleep is reached, then the blade will be disconnected.
We call this feature StripOff™, a naming that has received quite a positive response, for some reason :-D
* Obviously, a more radical solution is always possible, by removing the blade from the emitter socket to prevent any current to be used, however it's great for component blade sabers to stay with their blade in and secured when used often, especially for duelling (the blade being heavier) and we think it's great to have such a saber always ready to strike without disassembling it just for the sake to save on runtime when there's a smarter way to do it.* Neopixels LS pad (data line output)
For Prizm we had the room to implement a specific data line output to control the ledstrip. As required per the datasheet, a resistor is required between the output of the controller (sound board) and the strip to reducing ringing and keep the signal with clean edges & states. We added a 0603 footprint on the board to include the resistor on-board and save space in the hilt. The resistor is preferred to be "as close as possible" from the strip but it's definitely not critical with the scale of a saber hilt and we experienced no issue with on-board + 20cm of wires or thru-hole resistor next to the strip.* Neopixel Special Fx
Based on our former CF-LS design, we combine a spatial scrolling with the fade effect of the high-power LED for a more dramatic scrolling effect. Two parameters lsfadeon and lsfadeoff define if you use the fade effect during ignition or retraction.
I've also added a parameter called lightstick that illuminates the blade all the pixels at once (no spatial scrolling) and which can be combined with the lsfadeon and lsfadeoff parameters.
That feature might sound a little "off" compared to the lightsaber but can be useful for some other pikes or staff, and I got a few request in the past for my electronics to go in devil sticks for instance.
Finally, the spatial scrolling is quite different to the eye compared to the progressive fade in / out of a high-power LED. As a result the sample-accurate synchronization between the power on / off sounds and the extension / retraction scrolling speed is rarely required or looks too slow. Like on our previous boards, the power on / off timing can be adjusted to an eye-pleasing speed
using the quick-on / quick-off parameters in the configuration file.
* Neopixel AccuBolt™ special Fx
Designed initially in 2011 (and not in 2013 as posted elsewhere by myself, it actually got implemented in CFLS v5, therefore before CFLSv6) it recreates the bolt impact on the blade, and stands for Accurate Bolt. Thanks to the neopixels, we've for sure propped it up
- The bolt is no long a segment flashing (obvious technology limitation of the 6 segment strip, back in the day) but now a fully defined, color-size-duration-location configurable impact simulator
- It replicates the blade energy alteration due the impact. It's not a random chopped-slice/segment of blade that changes of color, but it has smooth edges with a brighter central spot where the impact is.
- It progressively returns to the blade color in a realistic "temperature effect". Think hot metal hue changing when you watch a blacksmith hitting glowing metal on an anvil.
- Specific color or randomized/altered hue or totally random color (configuration based, font specific)
- And at last but not least, thanks to the fruitful discussion with Alex Gordon, we've added MultiBolt™ for extended realism. MultiBolt™ is a system that allows to trigger the bolt impact on the blade when it's *actually* happening, not when the sound starts playing. As a result, multiple blaster sounds can be initiated and further blade-blocked with associated visual impact