CXA3590 Build Plan

Doer

Doer

Well-Known Member
churchhaze said:
This is a simulation of a 6' long 4" profile, 1.3" tall with 20mph wind across one long even 360W heat source (60% of 600W). The view doesn't fit the whole 6' bar, but it shows what I mean about temperature distribution.

View attachment 3369002

Honestly, if that's the temperature of the red part, it shouldn't be too much of an issue, although it is something to keep in consideration.

This is all I was trying to say. Sorry for trying to help. It's not like I haven't thought about designing exactly what you're thinking of.
Click to expand...
All are welcome to help. I love this diagram, thanks. I am sure you have thought about it. I didn't come in here to compete, I live here.

upload_2015-3-10_15-59-46.jpeg
 
Doer

Doer

Well-Known Member
churchhaze said:
This is a simulation of a 6' long 4" profile, 1.3" tall with 20mph wind across one long even 360W heat source (60% of 600W). The view doesn't fit the whole 6' bar, but it shows what I mean about temperature distribution.

View attachment 3369002

Honestly, if that's the temperature of the red part, it shouldn't be too much of an issue, although it is something to keep in consideration.

This is all I was trying to say. Sorry for trying to help. It's not like I haven't thought about designing exactly what you're thinking of.
Click to expand...
Well, you are in the ballpark at around 1760 LFM. Can you see what happens when you double that speed? I am thinking about dissipating 540w.

So, you think it is feasible also. With this sim, what has stopped you from finding the optimal set up? You could add a tiny bleed air port to that end. It would raise the pressure and lower the airspeed over all but would balance of the heat on the sink.
 
Sativied

Sativied

Well-Known Member
bicit said:
Eta: I'd love to see what this software is and what its capable of regardless though. New things are always fun.

Eta: since a know tone isn't communicated properly through text I just want to clarify. This is genuine interest. Im not trying to be snarky.
Click to expand...
It's amongst others used to create photorealistic impressions of architectural lighting (down to the photon count if needed). I honestly don't understand more than half of its capabilities because I don't need/use them, but using different planes at different heights representing the canopy could give some insight in hotspots and uniformity. Would also show the effect of using reflective walls nicely. That's not something I can slap together in a few minutes and if I put time in that it will be for something useful and positive. Surely there are cheap or even free raytracers that can do the same thing. May not be as realistic but good enough to experiment with different light source density and height (and angles... ideally all cobs in Doer's module would be able to tilt and rotate individually, or at least the cobs that end up along the edges.). Perhaps, using semi-transparent cones in sketchup one could do something similar...
 
Doer

Doer

Well-Known Member
churchhaze said:
This is a simulation of a 6' long 4" profile, 1.3" tall with 20mph wind across one long even 360W heat source (60% of 600W). The view doesn't fit the whole 6' bar, but it shows what I mean about temperature distribution.

View attachment 3369002

Honestly, if that's the temperature of the red part, it shouldn't be too much of an issue, although it is something to keep in consideration.

This is all I was trying to say. Sorry for trying to help. It's not like I haven't thought about designing exactly what you're thinking of.
Click to expand...
So, it cools off a bit at the end, downwind? That can be exploited.
 
Doer

Doer

Well-Known Member
Sativied said:
It's amongst others used to create photorealistic impressions of architectural lighting (down to the photon count if needed). I honestly don't understand more than half of its capabilities because I don't need/use them, but using different planes at different heights representing the canopy could give some insight in hotspots and uniformity. Would also show the effect of using reflective walls nicely. That's not something I can slap together in a few minutes and if I put time in that it will be for something useful and positive. Surely there are cheap or even free raytracers that can do the same thing. May not be as realistic but good enough to experiment with different light source density and height (and angles... ideally all cobs in Doer's module would be able to tilt and rotate individually, or at least the cobs that end up along the edges.). Perhaps, using semi-transparent cones in sketchup one could do something similar...
Click to expand...
Dude! Draw us a photon. Please please@!! :)
 
Doer

Doer

Well-Known Member
Seems to me Mr. @churchhaze, this sim tells us we need a bit more heat sink but only toward the end. The heat sink needs to get "better" there? Thicker?
 
Doer

Doer

Well-Known Member
This illustrates the concept of forced air. If the fan is drawing at 1000 LFM in 36 in^2, and the channel is 8 in^2 , that is the theoretical gain. It is not over 4 times faster due to turbulence, parasitic drag, etc, but it could be 3 times faster. There is likely an optimal combination of slope, (distance apart) with the ratio of the fan area and the channel area.

airflow.JPG
 
Sativied

Sativied

Well-Known Member
Doer said:
Dude! Draw us a photon. Please please@!! :)
Click to expand...
Not sure I can zoom in far enough to draw 'a' photon :) The more photons on a given location, the lighter the pixel(s) representing that location becomes. It doesn't draw individual photons, but it does calculate and simulate them realistically.

Even if I would put something pretty together today it would take my computer even longer to finish. I guess I've said too much to not back it up so... so just a quick example, not based on realistic specs, scale is not realistic and no volumetric lighting so you don't see the light, but the lit surface (doh...).

I run it on an beast not in my network so some pics (photos..)
doerled.jpgwithout light calc

15x900lumen 3000K spots (not led, these probably have more falloff, I'd need more input for realistic results)
doerled15x900.jpg
15x90000lumen HPS :D
doerled15x9000hps.jpg

Again, illuminating non-existing material, not meant to be a realistic representation of intensity or anything. That second image however does give an idea of what I was referring to earlier about using it to create a uniform spread.
 
Last edited:
C

churchhaze

Well-Known Member
Here's 6 feet of the 5.886" profile at 10 and 20mph with 1 big even 360W heat source.

10mph5886.jpg
20mph5886.jpg

Not too bad if you can actually keep that speed. Of course if the fan goes out, it's going to be a wreck. That's part of the reason I have larger heat sinks than I really need. Obviously it's not going to be perfectly spread like this too. The temperature of the base closest to the cobs will have higher temperature than the outsides, obviously.

(all this is assuming ambient temperature of 25C)
 
Last edited:
Scotch089

Scotch089

Well-Known Member
Sativied said:
Not sure I can zoom in far enough to draw 'a' photon :) The more photons on a given location, the lighter the pixel(s) representing that location becomes. It doesn't draw individual photons, but it does calculate and simulate them realistically.

Even if I would put something pretty together today it would take my computer even longer to finish. I guess I've said too much to not back it up so... so just a quick example, not based on realistic specs, scale is not realistic and no volumetric lighting so you don't see the light, but the lit surface (doh...).

I run it on an beast not in my network so some pics (photos..)
View attachment 3369161without light calc

15x900lumen 3000K spots (not led, these probably have more falloff, I'd need more input for realistic results)
View attachment 3369160
15x90000lumen HPS :D
View attachment 3369162

Again, illuminating non-existing material, not meant to be a realistic representation of intensity or anything. That second image however does give an idea of what I was referring to earlier about using it to create a uniform spread.
Click to expand...
What did you use to make those pictures. o_O
 
bicit

bicit

Well-Known Member
Sativied said:
Not sure I can zoom in far enough to draw 'a' photon :) The more photons on a given location, the lighter the pixel(s) representing that location becomes. It doesn't draw individual photons, but it does calculate and simulate them realistically.

Even if I would put something pretty together today it would take my computer even longer to finish. I guess I've said too much to not back it up so... so just a quick example, not based on realistic specs, scale is not realistic and no volumetric lighting so you don't see the light, but the lit surface (doh...).

I run it on an beast not in my network so some pics (photos..)
View attachment 3369161without light calc

15x900lumen 3000K spots (not led, these probably have more falloff, I'd need more input for realistic results)
View attachment 3369160
15x90000lumen HPS :D
View attachment 3369162

Again, illuminating non-existing material, not meant to be a realistic representation of intensity or anything. That second image however does give an idea of what I was referring to earlier about using it to create a uniform spread.
Click to expand...
The problem I see with running three tubes like this is getting even airflow through all three tubes without bogging down the fan with excessive 90* turns. You don't want one tube running hotter than the others, as that's bad for distribution(temperature droop).

2 tubes and one fan would work fine. Three tubes and one fan might be problematic. Are you planning on using a vortex fan to suck air through these? Or are you just using the computer fans to move the air?
 
Doer

Doer

Well-Known Member
Sativied said:
Not sure I can zoom in far enough to draw 'a' photon :) The more photons on a given location, the lighter the pixel(s) representing that location becomes. It doesn't draw individual photons, but it does calculate and simulate them realistically.

Even if I would put something pretty together today it would take my computer even longer to finish. I guess I've said too much to not back it up so... so just a quick example, not based on realistic specs, scale is not realistic and no volumetric lighting so you don't see the light, but the lit surface (doh...).

I run it on an beast not in my network so some pics (photos..)
View attachment 3369161without light calc

15x900lumen 3000K spots (not led, these probably have more falloff, I'd need more input for realistic results)
View attachment 3369160
15x90000lumen HPS :D
View attachment 3369162

Again, illuminating non-existing material, not meant to be a realistic representation of intensity or anything. That second image however does give an idea of what I was referring to earlier about using it to create a uniform spread.
Click to expand...
That is so cool. It is like seeing my dream. Amazing.

I was making a little Uncertainty jest, of course. It is why they call it ray tracing. :)

Overlapping. There it is.
 
Doer

Doer

Well-Known Member
churchhaze said:
Here's 6 feet of the 5.886" profile at 10 and 20mph with 1 big even 360W heat source.

View attachment 3369237
View attachment 3369238

Not too bad if you can actually keep that speed. Of course if the fan goes out, it's going to be a wreck. That's part of the reason I have larger heat sinks than I really need. Obviously it's not going to be perfectly spread like this too. The temperature of the base closest to the cobs will have higher temperature than the outsides, obviously.

(all this is assuming ambient temperature of 25C)
Click to expand...
Oh, it has to have a fail safe. No flow, no go. I made one for my water light. It takes a couple of interacting relays and a sensor. You have think about what if the fan is turning but something is blocking the air.

The DC side has a little relay that is activated by a flow sensor. I would use something solid state here for safety. Fan on, sensor detects air movement, relay is held closed.
That relay closed, allows the DC magnet to close in a larger AC relay for the light circuit.

If there is no flow at the sensor at any time, the DC relay opens and drops the AC. It saved my butt on several occasions.
 
Doer

Doer

Well-Known Member
bicit said:
The problem I see with running three tubes like this is getting even airflow through all three tubes without bogging down the fan with excessive 90* turns. You don't want one tube running hotter than the others, as that's bad for distribution(temperature droop).

2 tubes and one fan would work fine. Three tubes and one fan might be problematic. Are you planning on using a vortex fan to suck air through these? Or are you just using the computer fans to move the air?
Click to expand...
I was just thinking one fan per light bar. And I see the point now of just recirculating room air, no bulky ducts, so not a cool tube after all. :)

The fan choice is wide open. It needs to have more than 2 pounds of back pressure and pull better than 1000 LFM. I have not started looking but it need to be a pretty good suction fan.

It can certainly be hung from the ceiling next to the end of the light. But, you don't want too big. As small as possible really. The calculators say 250 CFM is 1000 LFM @ 6" x 6"
That gets at least doubled going thru the small area of the heat sink channel.

It is always a matter of paying the piper and never robbing Peter to pay Paul.

I wonder what gains are had from an upward rise angle to the light bar? What angle?
 
Last edited:
Doer

Doer

Well-Known Member
A computer fan won't work for this since I I need an exhaust fan or a bilge blower or something that can move air.

Been looking around. With China a player now, the Internet has gone back to the good old days. The Library of the world but the books are piled on floor. Serious, I watched this all slip into Google page 1 and follow me Ads. A bit sickening, to be frank.

IAC, we can still look far afield into other industries that are not so expensive. Take Boating for example. JSTkidding! But, it does have one of the world largest marketplaces.

Here is a hefty bilge blower on 12v DC 270 cfm in 4" cross section.
`$30.00
http://www.ebay.com/itm/12V-270CFM-IN-LINE-BOAT-BILGE-AIR-BLOWER-SUIT-4-HOSE-Boat-Marine-Caravan-Inline-/310795412035

    • Voltage: 12V DC
    • Flow capacity: 7,645 LPM (Litres per minute) / 270 CFM (Cubic Feet per minute)
    • Current draw: 6 Amp
  • Dimensions: Length 204mm (8") x Width 127.4mm (5") x Height 135mm (5.3") excluding mounts
  • Intake 105mm (4.1") - Suits 100mm (4") diameter blower hose
 
Last edited:
Doer

Doer

Well-Known Member
The cross section of the tunnel is slightly smaller than the cross section of a 4 inch circle. So we get a nice little speed boost. We have to see if this is still maybe too big. But, then it can be voltage throttled, to an efficient value.

LFMcalc.2JPG.JPG
 
Last edited:
You must log in or register to reply here.
Top