Airspeed of cooling fan and input electrical power amount / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Airspeed of cooling fan and input electrical power amount

previous 1 2

Move Thread

LAN_403

BigBad

Sat Nov 08 2014, 10:58PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

Putting them in series increases the airspeed, but putting them in parallel increases the volume. But the airspeed is usually plenty anyway.

Ash Small

Sat Nov 08 2014, 11:09PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

BigBad wrote ...

Putting them in series increases the airspeed, but putting them in parallel increases the volume. But the airspeed is usually plenty anyway.

Putting them in series doesn't double the airspeed, in my experience, while putting them in parallel seems to (EDIT: double the volume). I find that putting them on the inlet and outlet is better than just putting them in series, though.

EDIT: Please see my comment below

Andy

Sun Nov 09 2014, 01:25AM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Worked out assuming that 1mm of tangent surface area removes about 0.1W I'll need about 72kph airspeed for the surface area per LED(copper wires), as the drag at that speed will be significant I'll need to overcome the back pressure, ie will need some pressure.

I think by what you lot posted and more research I should beable to work out the fan to variables.

Thanks all.

Side note LEDs

But might be planning on using ones like this

LEDs are close to this data

BigBad

Sun Nov 09 2014, 10:02PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

Sounds a bit odd. 72 kph sounds too fast.

You normally would want add a heatsink to the LED to increase the surface area rather than rely on the wires.

Also, as a good rule of thumb the wires beyond a few diameters don't contribute to the cooling.

Ash Small

Sun Nov 09 2014, 10:25PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I think I may have misunderstood what you are trying to do here. If you are simply trying to cool the LED's then heatsinks with fans on them would be more appropriate. I originally thought you just wanted to move air in and out of the enclosure. Please disregard my earlier comments.

Andy

Sat Nov 15 2014, 02:45AM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Ash, I was trying to work that out, the light cells has one fan ,which sucks, to make up for the drag, with a high pressure high velocity air flow for the plumbing into the cells.

The goal is to make a grow light that make hardly any heat, to the plants, but is in the multiable of 200Watt(sun light power), say 10X, 20X etc/ meter. A sparky will sort out the power supply and wiring to the device, but the device has to last, 10k hours.

BigBad, I was planning on using the copper track of a PCB as the heat sink, with a solder join to the copper wire, about the area, I think that might be back wards, as circle area increases area by 3, and area in a second equation increases the length of power to a smaller area sample, heat transfer. If area effects two equation positively it should be a good thing to increase, you lot have helped out as I didn't think that adding more fans, would increase pressure or velocity.

Cheers all

Sulaiman

Sat Nov 15 2014, 07:38AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

LEDs die very quickly if they are run hot
(of course you know this, hence the question)
so you have a common semiconductor problem, just like cooling transistors etc.

For a given power you can use one high power device and intensive cooling,
or you can use several lower power devices in parallel.

Since LEDs emit all of their light in a forward direction you do not need a reflector
so I think that you could consider a large heatsink with many LEDs
all running at much less than rated power, say 1/2 or 1/4 rated power
this would reduce the semiconductor temperature rise above heatsink temperature to 1/2 or 1/4
with a suitably large heatsink you would not need a fan which is a potential point of failure.

IF you do use a fan then consider designing the luminaire
so that the cooling fan also pulls air from the growing area to circulate air around the plants,
reducing problems such as mildew, fungus etc.

side note;
About a year ago I put six axial fans in series just because I had them handy,
there was no noticeable increase in airspeed and very little increase in pressure
I think that this was because there was a significant gap between blade tips and housing
so any back pressure just caused air to 'leak' around the blades.
It was just a quick lash-up so maybe with the right type of axial fan and better construction and measurement there would be a significant pressure increase
but I doubt that pressure increase will be proportional to the number of fans due to the 'leakage' problem.

BigBad

Sat Nov 15 2014, 04:51PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

Andy wrote ...

BigBad, I was planning on using the copper track of a PCB as the heat sink, with a solder join to the copper wire, about the area, I think that might be back wards, as circle area increases area by 3, and area in a second equation increases the length of power to a smaller area sample, heat transfer.

Sorry, that won't work worth a damn. a thin track can't carry heat well at all. Beyond a length of a few thicknesses the track does nothing. So if the track is (say) 0.25 mm thick, beyond a length of ~2mm the copper track isn't contributing to the cooling. All the heat will be dumped into the air within a few mm and the air will still act as a nice insulator and your LED will cook. Or you will need very, very high airspeeds: lots of power for the cooling fan.

The point of a heatsink is to represent a low thermal resistance to carry the heat well away from the thing you're trying to cool; and for that you need relatively thick copper or aluminium (ratios of 6:1 length:thickness type thing)

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob