Author Topic: Pixelnet strobe controller  (Read 6060 times)

Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 851
  • Kudos: 10
Pixelnet strobe controller
« on: December 27, 2015, 07:33:05 AM »
David/Corey/ANYONE, can you PLEASE design a strobe controller for dirk cheap strobes (1w leds run at 5v)???  Something simple that can be fed from a F16 port or hub.  4 RJ45 jacks to feed strobes.  Start channel random fires 16 channels OR maybe 3 channel operation where say 1 turns on, 2 is duration, and 3 is speed. 

Sam

Last year's video: https://vimeo.com/150560653

Offline corey872

  • Supporting Member
  • ******
  • Join Date: Oct 2013
  • Location:
  • Posts: 476
  • Kudos: 16
    • Link to the sale...
Re: Pixelnet strobe controller
« Reply #1 on: December 28, 2015, 12:04:20 PM »
Interesting thought.  I would be a lie to say I had not thought about some type of strobe driver for my Halloween decor.  I had a couple boards prototyped, but looking for time to put them together.  Though these were more of a board to use a 1-3W LED with the 'strobe' being just a short 'blip' or single frame illumination created in xLights.  Overall, it seems there would be several ways to implement a 'strobe'.  All would come with some trade off between cost, functionality, ease of use, etc.  To sketch out some ideas:

Basic (what I have right now):

This would simply be an adapter so a person could install 1, 3, 5W, etc LEDs (either white or RGB at the user choice). The adapter would have some simple components to drive the LED as one big node and the user would program effects through the sequencing program of choice.  If you wanted the LED to be constant on, then program that, if you want strobe, program single frame 'blips' for that channel...what ever pattern, speed, etc you want.  Number of strobes / nodes are governed only by the power supply, wire gauge and available channels.

Pros:  wide appeal, simple, cheap, dual use (Big RGB color node or strobe)  Cons: Might not be as bright as possible or as short as possible flash


Option 1:

Board similar to that above which would accept same high power LED, but would have a jumper to bypass current limit resistors and send 5V directly to the LED.  Upon construction, the user would have the option to choose 'normal' or 'strobe' mode via jumper / solder bridge.  The unit would fit into the show the same as a node, but would be an either / or proposition with respect to node or strobe usage.  This would also require some care by the user to not set the LED for 'strobe', then apply a long duration 'on' command....only short blips.

Pros: Would drive LED more as an actual 'strobe' with a bit higher output / brightness.  Cons:  either/or proposition, requires user responsibility.


Option 1a - Same as above, though with some sort of 'latch' circuitry integrated into the 'strobe' side.  This would insure the LED could only fire for a short duration before needing to have an 'off' command to re-set.

Pros - may save the LED in case of a user 'goof'  Cons:  would add more cost, complexity


Option 2:

Actual LED Strobe Driver.  As you mention, this might use 1-2 channels to set time and pulse duration, then use channel 3 to fire it off.

Pros Brightest output / shortest flash  Cons:  Limited / dedicated use, one single color, higher cost.


Option 3:

"2-Channel" RGB Strobe Driver

Some sort of driver which would actually consume 2 nodes worth of data - 6 total channels.  A 'mode' signal could be transmitted on the first few channels so the node would know whether to be in 'strobe' or 'constant on' configuration.  Color data transmitted on the last 3 channels so you would effectively have a colored strobe.

Pros:  most realistic 'strobe' effect  Cons:  Highest cost, highest complexity.



Overall, for my use and overall appeal, I think the Option 1 would suit well.  It would provide the basic principles of operation, plus a white or colored strobe.  If a person wanted higher brightness, they could simply install a higher wattage LED.  And/Or take the responsibility of putting 5V directly across the LED and being sure not to turn it full on. (which seems to be what the dirk cheap strobe is doing anyway?)

Offline jnealand

  • Developer
  • ******
  • Join Date: Aug 2013
  • Location: Kennesaw, GA
  • Posts: 2,036
  • Kudos: 37
Re: Pixelnet strobe controller
« Reply #2 on: December 28, 2015, 07:37:28 PM »
Zwiller take a look at this video on around the 35sec mark

https://www.youtube.com/watch?v=9OUr1X8SmA8
Jim Nealand
Kennesaw, GA
Lynx, Falcon, XL, FPP

Offline twooly

  • Supporting Member
  • ******
  • Join Date: Oct 2013
  • Location:
  • Posts: 982
  • Kudos: 13
Re: Pixelnet strobe controller
« Reply #3 on: December 28, 2015, 08:23:39 PM »
I would also be in for some type of usc type of controller for the 1-3 w strobes.  I've got C9 style strobes, I love them but I what to move away from ac style lights.  I also got some of the ACL strobes, they would have been perfect but I was not happy with them at all, takes ~20-30 seconds before they become random vs like my C9 strobes that are random from the start.

You can see the strobes right at the start of my video, that is my idea of a strobe (Bummer I need to move to vimeo since they muted my audio..... >:()
https://www.youtube.com/watch?v=zzeApuxvhcY

These are the ACL strobes I have....and the not being random at the start.
https://www.youtube.com/watch?v=JaxHkN8cZHA

What I liked about the ACL strobes was the small form factor, basically a pixel that it can be controlled by 5v (would love 12v but 5v is ok)



Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 851
  • Kudos: 10
Re: Pixelnet strobe controller
« Reply #4 on: December 28, 2015, 09:26:06 PM »
Corey,
Thank you for taking an interest and your well thought response.  You make some very good points.  Talking about leds beyond 1w is very encouraging to me as well.

I still think a dedicated strobe controller will outperform sequencing but it might be doable.  Hard to deny this in context of XL's strobe function and the ability to use the leds full on.  Am I right, we are basically looking at 20 blips/sec @ 50msec at best if we actually sequenced?  Never really tried to do it. 

Dammit Jim!  (always wanted to say that  ;D) those look good thanks for posting.  We might have something with that pattern spaced out farther with 1W leds.  I think the longer spacing is the key. 

Todd, I appreciate you posting the ACL strobes.  Took one for the team me thinks  ::)

Right now, there a couple of DIY options that could be coaxed to work, but they involve special firmware/5V instead of 12v/etc.  I hope others chime and offer their opinion but I would like BOTH the driver board like you have already and the strobe board in option 2 and both would be a welcome addition the the Falcon line. 


Offline pixelpuppy

  • Sr. Member
  • ****
  • Join Date: Aug 2015
  • Location: Dallas, TX
  • Posts: 485
  • Kudos: 10
Re: Pixelnet strobe controller
« Reply #5 on: December 28, 2015, 10:22:25 PM »
  Am I right, we are basically looking at 20 blips/sec @ 50msec at best if we actually sequenced?
Keep in mind that each "blip" requires and ON frame and an OFF frame.  At MINIMUM its two frames per "blip".  So at 50msec timing, the most you could do is 10blips/sec.

With that said, it's still probably much faster than you'd want.  At that rate I think it would look like a flickering bulb that was malfunctioning and not a strobe.  It needs to be more like 1-2 flashes per second with some randomization across the whole group of lights.
Vixen and xLights for sequencing / FPP for scheduling and playing / Falcon controllers for pixels / DIY controllers for everything else

Offline arw01

  • Hero Member
  • *****
  • Join Date: Oct 2013
  • Location:
  • Posts: 821
  • Kudos: 0
Re: Pixelnet strobe controller
« Reply #6 on: December 28, 2015, 10:46:15 PM »
Tale a look at the Komby Blinkin product. I've not hooked mine up but would do what you want.  Probably not a big deal to change to some pixelnet protocol from the dmx it would be now.

I've not used my ACL strobes, have a hundred not assembled, a hundred assembled.  I bought the Komby to be able to control them someday and added the blinkin to the order just for fun.  I figured I would have to split up the strobes on multiple channels anyways to get random.

How cheap of a pic could you get that would let you do a real random?

Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 851
  • Kudos: 10
Re: Pixelnet strobe controller
« Reply #7 on: December 28, 2015, 11:34:33 PM »
Good point pixelpuppy.  Once again never tried it in sequencing.

Komblinkin gets me close.  Another member offered to write code for me to defeat the wireless and make it dumb.  IE power it 5v and it strobes.  However, you gotta have an ftdi programmer to flash it, learn how to use it (my main concern) and then you need a dumb 5v controller to turn it on.  Another thing brewing is the Renard Plus RGB+W.  Mac might get lucky and find the random flash code he needs to implement.  So there are ways to hack strobes, but I thought there might be a better way like the Falcon Team does. 

Offline corey872

  • Supporting Member
  • ******
  • Join Date: Oct 2013
  • Location:
  • Posts: 476
  • Kudos: 16
    • Link to the sale...
Re: Pixelnet strobe controller
« Reply #8 on: December 29, 2015, 08:12:21 AM »
...
How cheap of a pic could you get that would let you do a real random?

For sake of discussion, I will assume "real random" to mean "pseudo-random to the point the casual observer would not notice a pattern".  (and spare the great mathematical / philosophical debate over "true" randomness :)  )

With that said, you could probably do it with the same $1.30ea PIC running the uSC's... (or similar depending on how many output channels you wanted) though you also need all the volt regulators/capacitors/output drivers, etc to go along with it.  ...plus firmware, ICSP provisions, etc.  Overall, I'm not terribly sure it's worth the hassle, but this does exemplify the two main forks in this road:

1) Offload the 'strobe' to a dedicated controller which just receives a 'go' command from the sequence and starts strobing.

Pros:  Would save a few channels in the sequence, maximum brightness for 'white' strobe  Cons: complexity of PIC and associated 'support' components, firmware to keep track of. Less flexibility in color / sequence...or substantially more complex PIC firmware to sort it all out (Dave!?!)

2) A board which allows high wattage LEDs to be added directly to the sequence as 'nodes' and triggered through the sequence program.

Pros: Single board solution, program as 'random' or as 'sequenced' as you want, program strobe color of choice or white  Cons: 3 channels required per RGB strobe / 1 channel per 'white' strobe


Regarding speed - as pixelpuppy said, 2 frames per blip, 10 blips per second...though I would agree - much faster than that and it starts to look like a flicker as opposed to a strobe.  Additionally, looking at some LED 'overdrive' guidelines from Cree, they state:

Quote
1. For duty cycles between 51-100%, do not exceed 100% of the maximum rated current;
2. For duty cycles between 10-50%, do not exceed more than 200% of the maximum rated current;
3. For duty cycles less than 10%, do not exceed more than 300% of the maximum rated current.

So this would have you at 50% duty...right on the border of Condition 1/2...so again, it sort of falls to the user to extrapolate...how hard do you want to push the LED and what 'risk' is acceptable.  A dedicated controller could lock you to 10% or less duty and provide max brightness...but again, you loose flexibility.  A more general controller could let you run the full range of strobe to constant-on, though would either require the user to be responsible with the 'power' or would lock you to '100%' rated current and a slightly dimmer flash. (Cree also mentions the power increases do not result in linear brightness increases...efficiency drops at higher levels and light output decreases with increased heat of the junction.)

Offline twooly

  • Supporting Member
  • ******
  • Join Date: Oct 2013
  • Location:
  • Posts: 982
  • Kudos: 13
Re: Pixelnet strobe controller
« Reply #9 on: December 29, 2015, 08:45:55 AM »
I myself like the idea of apply voltage and it does its thing (thats why I liked the ACL strobes)

Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 851
  • Kudos: 10
Re: Pixelnet strobe controller
« Reply #10 on: December 29, 2015, 09:39:21 AM »
Great info.  LOL "true randomness"

Forgot all about the constant current DMX boards overseas.  I think these would work for someone who wants to program blip strobing. 
1w: http://www.aliexpress.com/store/product/DM-100AC-300ma-3channel-dmx-constant-current-decoder-AC9-32V-input-300ma-3-channel-output/701799_1417374648.html  Was told my Ray that these work for DC as well as AC. 
For 3w: http://www.aliexpress.com/store/product/DM-101-4-channel-RGBW-dmx-constant-current-decoder-DC12-24V-input-600ma-4channel-output/701799_868054623.html

Both require a start channel programmer.  Then there are some newer boards with dipswitch addressing.  http://www.aliexpress.com/store/product/4CH-easy-dmx-constant-current-decoder-DC5-36V-input-350ma-4channel-output-L82-W65-H18mm/701799_32347810143.html  You still need to figure out the power and dmx data.  A custom cable would work. 

It also looks like there is newer generation of these.  Not sure what the difference.  http://www.aliexpress.com/store/product/AT2390-DMX-constant-current-decoder-DC12V-input-RGB-3-3w-output-pwm/701799_547318311.html

And here is something I just stumbled across and saw for the first time.  Looks like switchable 350 or 700ma CC. http://www.aliexpress.com/item/350ma-constant-current-12-channel-dmx-512-dimmer-LED-DMX512-decoder-RJ45-XRL-3P/32379521522.html?spm=2114.01020208.3.88.78hsn1&ws_ab_test=searchweb201556_2,searchweb201644_1_79_78_77_82_80_62_81,searchweb201560_1  Pretty sure I am getting one of these to fool with! 

Maybe I should clarify what I have in mind as a stand alone strobe board.  Fits in TA200, is fed by 1 port of F16, 16 outputs via RJ45 jacks.  Something of a 4 port hub merged with uSC that runs strobe firmware.  "Innocence and mayhem at once". 



Offline t.jo13

  • Supporting Member
  • ******
  • Join Date: Dec 2014
  • Location:
  • Posts: 184
  • Kudos: 2
Re: Pixelnet strobe controller
« Reply #11 on: December 29, 2015, 06:20:25 PM »
Sam,
Ray offers 2 different types of strobe that can run off of the f16. ( pic attached ) the round plastic domed ones are 12v, and  i was not that  happy with them. The node style ones are 5v and have some good pop, especially when you put a c9 bulb cover on. I  tested them with the th2010 controller that I have used for few years and the 5v get some good  randomness playing with the speed as they are a strobe bulb but you have to make them strobe  they did not make it into the show this year but I can run a couple of tests in the morning using  xlight strobe function and adjusting the time of the effect. To see what that does. I wonder if there could be an addition to the usc utility to help make the strobe effect, or a modification to the v2 usc to allow it instead of adding another board to the mix. Not sure  if it is possible or not . Just cant beat a xeon ac strobe.  :)

Offline mararunr

  • Hero Member
  • *****
  • Join Date: Dec 2013
  • Location:
  • Posts: 509
  • Kudos: 12
    • Bentonville Heart Lites
Re: Pixelnet strobe controller
« Reply #12 on: December 29, 2015, 06:49:37 PM »
Sam,
Ray offers 2 different types of strobe that can run off of the f16. ( pic attached ) the round plastic domed ones are 12v, and  i was not that  happy with them. The node style ones are 5v and have some good pop, especially when you put a c9 bulb cover on. I  tested them with the th2010 controller that I have used for few years and the 5v get some good  randomness playing with the speed as they are a strobe bulb but you have to make them strobe  they did not make it into the show this year but I can run a couple of tests in the morning using  xlight strobe function and adjusting the time of the effect. To see what that does. I wonder if there could be an addition to the usc utility to help make the strobe effect, or a modification to the v2 usc to allow it instead of adding another board to the mix. Not sure  if it is possible or not . Just cant beat a xeon ac strobe.  :)
I use the 5v and wrap them in bundles of 7 to make 'em really bright (treat all 7 as one pixel with virtual string setting on F16v2).  Very nice and great that they can be inline with the rest of my 5v stuff.

Sent from my XT1254 using Tapatalk

Bentonville Heart Lites (www.facebook.com/bheartlites)

This is just my opinion/suggestion/viewpoint.  Others with other viewpoints/experiences may have different advice.  I am a hobbyist with a couple years real world experience, not an expert.

Offline AussiePhil

  • Sr. Member
  • ****
  • Join Date: Dec 2014
  • Location:
  • Posts: 495
  • Kudos: 11
Re: Pixelnet strobe controller
« Reply #13 on: December 30, 2015, 02:37:14 AM »
Lots of great discussion but I don't think one idea gives you a true independent string able strobe, it's actually neither cheap nor simple to drive 1w or better LEDs as a strobe.

@twooly, I think I worked out your issue with the ACL strobes, as far as I'm aware the ones that got made still use the base firmware I wrote years ago for the original units. The firmware at turn on immediate seeds a random value for delay and will immediately randomise a string ... Except... If the voltage does not stay up and the pic resets and has to reseed.
I noticed this when testing the new strings running from a dc-dc solid state switch, until it turned on 100% they wouldn't randomise

Offline twooly

  • Supporting Member
  • ******
  • Join Date: Oct 2013
  • Location:
  • Posts: 982
  • Kudos: 13
Re: Pixelnet strobe controller
« Reply #14 on: December 30, 2015, 07:09:16 AM »



@twooly, I think I worked out your issue with the ACL strobes, as far as I'm aware the ones that got made still use the base firmware I wrote years ago for the original units. The firmware at turn on immediate seeds a random value for delay and will immediately randomise a string ... Except... If the voltage does not stay up and the pic resets and has to reseed.
I noticed this when testing the new strings running from a dc-dc solid state switch, until it turned on 100% they wouldn't randomise

I'll pm you, so I don't derail this thread since I have strings that don't randomize at least how I'm hooking them up.

 

Back to top