Project: 5W full color

[flecom]

Hehe, that rather flies in the face of reports from people who insist on using TECs.

I'm a fan of the KISS principle (there's enough complication going on in a projector already - one less thing to fail) so passive cool wherever possible.

Did you monitor temp and output during that period at all? Interested to see if you saw any dropoff in power as the temp rose (significant or otherwise)

power stayed about the same, I just checked it with my lpm every now and then

as far as temperature I just touched it to see how hot it was... without a fan it would get too hot to touch after a while but a little bit of forced air did the trick... the 40mm fan kept it pretty much cool to the touch

while making a sealed projector is a great idea its just not really realistic for me... I was thinking about how I could pull off sealing it and the only thing I could come up with was water cooling... still considering it

[LaNeK779]

I was thinking about how I could pull off sealing it and the only thing I could come up with was water cooling... still considering it

if you check computer enthusiasts' and overclockers' forums you'll find lots of info on this, even ready made kits

go for a daisy chain design from radiator to waterblock 1 then 2 then 3 etc, then reservoir then pump and so on. this keeps the flow rate neater and the tubing simpler. tygon tubing is a good idea as it does not bend or kink easily. a reservoir is necessary, to help you get air out of the system easily when refilling and will allow you to use more water, so the system will be more resistant to temperature peaks

don't let the pump run too quick. fast flowing water will just pass through the waterblocks and won't have enough time to absorb enough heat. just get a smooth, relaxed flow rate for better results (i.e use 1/2'' tubing than 3/8'')

a temp sensor in the reservoir will help you optimise the flow rate. an alarm circuit would make it easier to detect pump failure (but even if this happens, the system has a large failsafe window, before all the water gets boiling hot, maybe even several hours)

just don't mix copper and aluminum in the same cooling circuit to avoid corrosion and punctures. if the radiator is aluminium, make the blocks out of aluminium too

don't use distilled water as it is not very thermally conductive. tap water is better, even if you'll face salt residue problems some time (in some years time)

if you have access to a mill you could make your own waterblocks, no need for an elaborate design, just a piece of metal with a groove running through it a couple of times and a lid with 3/8 or 1/2 push-fit fittings http://webx.dk/oz2cpu/pcmod/waterblock-gf2-b.jpg a more complex design will restrict flow rate, but could make the water more turbulent, so that a higher volume of water would contact the block, thus yielding better cooling capacity

fishtank pumps are very suitable for the job (eheim for example)

you can search for a used motorcycle radiator, which is relatively small, or you can resort to the many times more expensive ready made pc watercooling radiators (like black ice or others)

external fittings to drain the system before transportation and refill it after you arrive could be a good idea

(i hope you don't plan to watercool the optics as well, but even for this there are suitable mounts from industrial CO2 systems )

that's the general idea, back from the days that i used to tinkle my poor PCs

[flecom]

hrmm, I think I may know what your talking about

[dnar]

My only objection to using TEC's when not necessary is that ultimately you wind up having to move twice as much heat... Granted, fan cooling is better outside of a module.

[White-Light]

hrmm, I think I may know what your talking about

Hmm Ribena. Invisible characters

[laserwinkel]

back to topic
so parts have arrived



i have the following dichros:
Reflect Red
Reflect Blue
Reflect Blue + Green

any tips suggestions or something else before i go mount everything ?

[mccarrot]

Are your red lasers selected for H and V polarisation?

If not you need a waveplate before you can combine them with a PBS.

[laserwinkel]

Are your red lasers selected for H and V polarisation?

If not you need a waveplate before you can combine them with a PBS.

not sure about this, is there a way to check how they are polarised?

[mccarrot]

99% chance they have both the same polarization, if you shoot the trough a PBS and they react the same they have the same polarization.

Be prepared to buy a waveplate to turn the polarisation, or the less prevered option, put one of the red lasers on his side.

http://store.laserextreme.nl/catalog...products_id/59

[laser_freak]

99% chance they have both the same polarization, if you shoot the trough a PBS and they react the same they have the same polarization.

Be prepared to buy a waveplate to turn the polarisation, or the less prevered option, put one of the red lasers on his side.

http://store.laserextreme.nl/catalog...products_id/59

Or use two mirrors to rotate the beam.