The Modified Black Box Thread

[Chrisanthellae]

Looks like you would not put enough volts into them, this causing less light output and par.

Looks like you're right. I found this thread that was quite helpful: https://www.electronicspoint.com/forums/threads/running-led-bulb-on-lower-voltage.275364/

TLDR: LED brightness does not scale linearly with voltage as is common with incandescents. A small change in voltage can make a huge difference in how much current the LED draws, thus having a huge impact on brightness. So running these 5W LEDs at around half their rated voltage would result in them being barely visible, if at all.

 

[oreo54]

Yea those "5w" ers are basically 2 "3W" chips in series.. so each one replaces 2...


Another thing about led's is watt efficiency increases w/ lower current. Sort of the less amps, less heating since more energy is released as photons.
Most if not all bb's are run low in amps ( about "2W" worth) anyways thus they can get away w/ flimsy heat sinks, even w/ the fans..

At 700mA you get 100% of rated output.
At 1400mA one expects 200% but gets say 185%
Of course gross amount is less..but 2 @ 700 exceeds output of one at 1400.
Lower currents are more linear though.

Err brightness scales w. voltage w/ LED's

SIMPLISTIC BULB VERSUS LED CONCLUSION:

1) A light bulb lumens (brightness) falls off a factor of 3.5 times faster compared to that of a LED for the same voltage drop in track voltage.

2) When running from a loosely regulated track voltage and you want the light source to have relative constant brightness, use a LED.

LED Vs Bulbs - Mark Gurries

Education about Digital Command Control (DCC) used in model railroads to run trains. This site is also a distribution point of my DCC presentations in PDF format. My Model Railroad Website

sites.google.com

Hmmm.. Seems contradictory (guy is an electrical engineer though).. but current draw increases more w/ voltage changes but efficiency decreases.. Must even out "in general"

Of course most "think" current not voltage.

But the voltage changes w/ LED are small for a large increase in current (and on/off or lifespan) w/ a fairly sharp on/off threshold that you noted

Got to think about this a bit.



Lots of interesting things w/ LED's..link below is just for "fun"

Thermal management of high power HB LED

The efficiency of transformation of electricity into the light is a key point for HB LEDs. Efficiency of the typical LEDs achieves 40% to 50%. That means that more than 50% of the electrical power …

3n-technology.com

 

[mrlavalamp]

Seems like going to 5w would require drivers to be upgraded to cope with the different voltage/current requirements even if you plan to underdrive them and extend the life and efficiency.

Probably not worth it. Or just keep going and change out everything (heatsink, lenses etc) and make your own fixture from scratch.

 

[JCOLE]

What would be a better white spectrum for SPS growth? 6500K, 7500k, 12000k, 20000k? I am assuming 6500k? I modified one of my black boxes a couple days ago and I think it looks good. It is a Mars Aqua 165W. I removed all whites down to 8. The remainder has a mix of 6500, 7500k, and 20000k. I am going to do another swap tomorrow and thinking of more 20000k but I don't want to sacrifice growth for a blue look if so.

 

[AbjectMaelstroM]

Got my spare Viparspectra in this is my plan for a modified layout, as it stands now I only use by White channel at 3%, feels like a waste, so I figured throw some more blue on there. Wish I could get PC Amber in this 3W egg-package.

Top is STOCK Viparspectra layout. Bottom is modified (anything with text is the modified diode).

Modifications to blue channel: Mostly untouched, replaced 4x470 Blue with 2xRB and 2x395.

Modifications to white channel: Replaced 12 whites with 4x490 cyan (I actually considered replacing the blues with the cyan as they "appear" to have a good bit of green in them), 2x395, 2xRB, 2x470, 2x420.

Thoughts?

 

[oreo54]

Got my spare Viparspectra in this is my plan for a modified layout, as it stands now I only use by White channel at 3%, feels like a waste, so I figured throw some more blue on there. Wish I could get PC Amber in this 3W egg-package.

Top is STOCK Viparspectra layout. Bottom is modified (anything with text is the modified diode).

Modifications to blue channel: Mostly untouched, replaced 4x470 Blue with 2xRB and 2x395.

Modifications to white channel: Replaced 12 whites with 4x490 cyan (I actually considered replacing the blues with the cyan as they "appear" to have a good bit of green in them), 2x395, 2xRB, 2x470, 2x420.

Thoughts?

Why not just get some cheap amber and replace the greens?
Collective whites have lots of green..

 

[AbjectMaelstroM]

Why not just get some cheap amber and replace the greens?
Collective whites have lots of green..

Closest to pc Amber I guess would be what ebay calls yellow is 585-595nm range?

I'm waiting on more RBs. Will see how things look. Ideally I'd like to get rid of green and red entirely, replacing with cyan amber. Since I already have cyan, may dump green for blue and replace red with amber (yellow).

 

[oreo54]

Closest to pc Amber I guess would be what ebay calls yellow is 585-595nm range?

I'm waiting on more RBs. Will see how things look. Ideally I'd like to get rid of green and red entirely, replacing with cyan amber. Since I already have cyan, may dump green for blue and replace red with amber (yellow).

How about IR for red, Amber for green?
Would be nice to find a lower than 595 "yellow" but, honestly, and from what I remember from a Dana test, nm are all over the board anyways w/ the cheaper diodes.. so who knows exactly what you get.

ATI blue plus peaks..

 

[AbjectMaelstroM]

How about IR for red, Amber for green?
Would be nice to find a lower than 595 "yellow" but, honestly, and from what I remember from a Dana test, nm are all over the board anyways w/ the cheaper diodes.. so who knows exactly what you get.

ATI blue plus peaks..

The spectrum crap shoot with these cheap LEDs is the only thing keeping me from "doing too much" as you really don't know what you're getting.

Should probably just go with KISS on this and just add blues/rb to the white channel leave eve thing else be. Theory and "on paper" spectrums are nice and all but lack of control over the actual led wavelength makes it somewhat pointless.

 

[oreo54]

The spectrum crap shoot with these cheap LEDs is the only thing keeping me from "doing too much" as you really don't know what you're getting.

Should probably just go with KISS on this and just add blues/rb to the white channel leave eve thing else be. Theory and "on paper" spectrums are nice and all but lack of control over the actual led wavelength makes it somewhat pointless.

There is no fun in that... ;)

Why not at least attempt to get a TRUE 20000k white channel or close and fill in a few holes in the process

* MIXING LIST
----------------------------------------
LED Violet (420nm) [120°] x4
LED RoyalBlue (450nm) [120°] x3
LED Cyan (490nm) [120°] x4
LED CoolWhite (8000K) [120°] x16
LED NeutralWhite (4000K) [120°] x2
LED WarmWhite (3000K) [120°] x2
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,763 lm
Radiant flux : 12,004 mW
PPF : 50.4 umol/s
TCP : 21740 K
CRI : 56
λp : 453 nm
Color : #979AFF
----------------------------------------

* PERFORMANCE @ 90cm
----------------------------------------
Irradiance : 1.6 W/m²/s
Illuminance : 361 lx
PPFD : 6.6 umol/m²/s
----------------------------------------

by SPECTRA 1.0β @ 1.023world

SPECTRA

SPECTRA can perform the optical simulation of many required for coral breeding. SPECTRAはサンゴ飼育に必要な多くの光学シミュレーションを行う事が可能です。

spectra.1023world.net

There are some extra RB diodes to make the whites higher than 8000K..so ignore the rb count.
28 total. swap count is 8 and a lot less possible disco, well at least in R/G spots
W/ luck a TRUE 20000k-ish channel

 

[Nicholas Dushynsky]

I'm running lensless on my viparspectra at 60 blue 5 white. I want to maybe add some violets and royal blues to my white channel, I'm not sure if it would make much difference, and is it actually worth changing the leds out?

 

[oreo54]

I'm running lensless on my viparspectra at 60 blue 5 white. I want to maybe add some violets and royal blues to my white channel, I'm not sure if it would make much difference, and is it actually worth changing the leds out?

You need a reason to I suppose.
Add missing/low parts of the spectrum.
Increase PAR by being able to increase channel
Change the look..
With you running at 60/5 it isn't par, You can go 100/9 and not change the look.

Just since I don't think it's been done.. Channel breakdown per channel. Not all peaks/spread will be exact due to differences in diode electrical and spectral charactheristics so take it for what it is or isn't..

Keep in mind these are relative charts so if you combine them your blue range peak goes way up while your "not blue" range gets squished down. Doesn't mean it is low just relatively less in comparison to blue.


I know you said add violet blue but why do you think you need it?

 

[JCOLE]

I dont know if this was already discussed in this thread but I figured I would provide an update on the newer Mars Aqua boards. I have 6 of these and I am in the process of changing out the diodes on the white channels.

Completed layout over frag tank on new light. Spectrum looks really good. On my DT I am going to use this layout except I will change out the 4 7500K diodes for 20000K for a crisper white. These are the newer Mars Aqua boards with COB diodes so I had to get a little creative.

Prepped and attached new diodes to heat sinks.

What I did first was break the first diode on the incoming negative side so I could use my meter and test voltage for every diode before testing lights to make sure I had a proper connection.

Then I scraped the old diode off the board to break the circuit. Important to do this one at a time for easy testing. Cleaned all old glue, residue, etc off board. Then scraped black paint off both incoming and outgoing circuit to expose the bare copper.

Soldered buss wire from the board to the heat sink. This is where you can use your volt meter to check which side is positive and negative if you are unsure. Once new diode is mounted then you can test the first diode that was scraped off in the beginning with a volt meter for voltage drop. If the voltage is correct then you know you have a good connection. Switch the meter to continuity and this will cause the circuit to light up when applied to the copper of the first diode.

Took a little time but once started it was about 2 hours from start to finish.

 

[Nicholas Dushynsky]

I know you said add violet blue but why do you think you need it?
[/QUOTE]
I would like to add some royal blues to the white channel as I dont run the whites very high as it looks quite yellow if I run it high. I think I should just leave well alone to be honest. Knowing my luck I'd either ruin the fixture or really annoy the corals. I only have an lps tank with a few zoas. I like the look of it under just blue but under white its just meh.

But looking on screen the white doesn't look too bad.

 

[Nicholas Dushynsky]

Good call on the multimeter. I ran to Home Depot picked one up. One of the red diodes was bad. Replaced and presto. Thanks again.

How do you like the colour over your tank with the mods? Is it noticable difference? I have ordered 20 new leds consisting of 10 450k royal blues, 4 20000k super bright lights and 6 x 420k violets, I'm not sure which ones I'm going to change as of yet.

 

[Nicholas Dushynsky]

Got my spare Viparspectra in this is my plan for a modified layout, as it stands now I only use by White channel at 3%, feels like a waste, so I figured throw some more blue on there. Wish I could get PC Amber in this 3W egg-package.

Top is STOCK Viparspectra layout. Bottom is modified (anything with text is the modified diode).

Modifications to blue channel: Mostly untouched, replaced 4x470 Blue with 2xRB and 2x395.

Modifications to white channel: Replaced 12 whites with 4x490 cyan (I actually considered replacing the blues with the cyan as they "appear" to have a good bit of green in them), 2x395, 2xRB, 2x470, 2x420.

Thoughts?

I'd like to know what a good combination of what to change on the viparspectra light is.

 

[AbjectMaelstroM]

I'd like to know what a good combination of what to change on the viparspectra light is.

Honestly, without access to a spectrometer, this is all "Kentucky windage" at best. Meaning, there is no way to tell. All i know is that viparspectras grow SPS just fine and that the white channel is under-utilized; so I'm adding more of what seems to work on the blue channel, to the white channel.

 

[Billsreef1]

I'd like to know what a good combination of what to change on the viparspectra light is.

Following I have 2 of the vipar 300w over my 125g. Running at present 28% blue 1% white with no mods, lenses are on. Verified par with the mq-510. Very interested in mods/diode swaps if needed...

 

[Billsreef1]

Honestly, without access to a spectrometer, this is all "Kentucky windage" at best. Meaning, there is no way to tell. All i know is that viparspectras grow SPS just fine and that the white channel is under-utilized; so I'm adding more of what seems to work on the blue channel, to the white channel.

Agreed!!! Very interested in your results with the changes!!!

 

[Billsreef1]

Honestly, without access to a spectrometer, this is all "Kentucky windage" at best. Meaning, there is no way to tell. All i know is that viparspectras grow SPS just fine and that the white channel is under-utilized; so I'm adding more of what seems to work on the blue channel, to the white channel.

Another thing I wonder is why did Vipar build a light where the white channel is so under utilized.....I am running mine at 1%...it just doesn't make much sense!!