Page 2 of 3 FirstFirst 123 LastLast
Results 11 to 20 of 30

Thread: adding TECs to my baseplate, advice needed

  1. #11
    Join Date
    Feb 2008
    Location
    East Sussex, England
    Posts
    5,212

    Default

    As has been said above, avoid TECs if you can.

    I put together a pair of 20W RGB last year using 7x 1W 520, 7x M462 and 4x overdriven 638 diodes and all on a 10mm baseplate fixed to the projector baseplate with a simple heatsink bolted on the bottom. Then fans to whisk the heat away.
    http://www.facebook.com/SubsonicSystems
    http://www.frikkinlazors.co.uk

    You are using Bonetti's defense against me, ah?

    I thought it fitting, considering the rocky terrain.

  2. #12
    Join Date
    Apr 2016
    Posts
    162

    Default

    Quote Originally Posted by norty303 View Post
    As has been said above, avoid TECs if you can.
    Noted.

    I put together a pair of 20W RGB last year using 7x 1W 520, 7x M462 and 4x overdriven 638 diodes
    That's interesting, 20W RGB with raw 7W green and 14W blue? How much were your optical losses?
    How much did you overdrive the reds?

    a simple heatsink bolted on the bottom.
    One big heatsink? Interested where you got one that big (I assume around 30cm x 30cm).

  3. #13
    Join Date
    Feb 2008
    Location
    East Sussex, England
    Posts
    5,212

    Default

    Quote Originally Posted by Nii
    One big heatsink?
    No, quite a small cheap one actually, supplied by the Chinese case manufacturer.

    In terms of power, I can't quite recall what we got out the end of each module now, and we never tested power pre and post optics, so couldn't say, but they were in the right ballpark. Given that at ~20W you'd struggle to see 2W difference, I was more focussed on getting the knife edging and overlaying correct. Beam quality starts to count more than outright power at those sorts of levels I've found.

    The quad reds were running around 2.8W after optics iirc, so we were doing over 750mW per diode, but still under the pulsed max from the data sheet.

    These projectors, although over $10k in budget, were built down to a 'best bang for buck' budget. Originally they were just going to be green/blue, but we decided to sacrifice a couple of 520 diodes (which were quite expensive at the time) in order to add the bit of red to each projector to make them more versatile. They worked out pretty well all things considered.

    Simplicity was also important as they were primarily intended to go to Burning Man and get full of dust! And they did.... the state they were in when they came back to me for cleaning was horrific.
    But unnecessary complications and budget and time sucking things like TECs were ruled out very early on.
    http://www.facebook.com/SubsonicSystems
    http://www.frikkinlazors.co.uk

    You are using Bonetti's defense against me, ah?

    I thought it fitting, considering the rocky terrain.

  4. #14
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    or have an aluminum module housing to cool the module from its top, which makes removing heat problematic since its on the optics
    Have you changed your mind about this kind of setup, or did I not explain what I intended properly?
    It was your explanation which made it sound as if you planned to cool the module where I show the acrylic sheet, ie the top.

    Simplicity was also important as they were primarily intended to go to Burning Man and get full of dust!
    This is an important concept to understand when you plan to build. Some of us (like me) enjoy the building process and the projectors are a platform to test ideas, they will not travel much and cost effectiveness is less of an issue. Those that build or buy are looking for ease of use, convenience and robustness as well as cost effectiveness.

    The TECs can be used to boost performance. For that matter so can liquid nitrogen if you get my point. You can construct some pretty radical systems, but you may not want to try to build a business around them.

  5. #15
    Join Date
    Apr 2016
    Posts
    162

    Default

    Quote Originally Posted by planters View Post
    It was your explanation which made it sound as if you planned to cool the module where I show the acrylic sheet, ie the top.
    Sorry I don't understand. Did you mean in your original post that it's hard to remove heat from the whole base plate and I should try removing the heat only from the modules, or the other way round?
    What I want is cooling the modules, not caring if the whole baseplate gets cooled in the process or not.


    Quote Originally Posted by norty303 View Post
    No, quite a small cheap one actually, supplied by the Chinese case manufacturer.
    About how much? What I wanted to ask was the width and length of the heatsink, not the height.
    If I try to use just heatsink(s) instead of peltier with CPU heatsinks I'd like to have a general idea what size to go for, and where to look. Maybe I'm overcomplicating the planning and should go with few small heatsinks instead of finding a big one.

    The quad reds were running around 2.8W after optics iirc, so we were doing over 750mW per diode, but still under the pulsed max from the data sheet.
    Interesting you got that high with passive cooling. Thank you for the info.

    BTW, heatsink plaster or thermal grease?
    Last edited by Nii; 07-13-2016 at 13:43.

  6. #16
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    Sorry I don't understand. Did you mean in your original post that it's hard to remove heat from the whole base plate and I should try removing the heat only from the modules, or the other way round?
    What I want is cooling the modules, not caring if the whole baseplate gets cooled in the process or not.
    You only want to cool the modules. The TEC that is placed between the modules and the base plate below them will transfer the heat into the base plate. Cooling the entire base plate with TECs is a bad idea. It will require huge amounts of power and will overwhelm any finned heat sink if you try to reduce the temperature of the base plate more than a few degrees below ambient. Have you ever worked with TEC's? They are useful, but they are VERY inefficient.

    A quad red producing 2.8W (700mW each) with passive cooling is very reasonable.

    Always use the largest and most densely finned heat sinks you can fit, afford and locate. The efficiency will continue to rise. Make sure your heat sink interface surface is as flat as possible. Grind it if necessary. A good flat surface will generate a suction effect if you try to pull it away quickly.

  7. #17
    Join Date
    Oct 2012
    Location
    Germany
    Posts
    1,059

    Default

    ... if your "thermal mass" is big enough, then passive cooling should be enough - I'm mounting 9Watt-diodes with their drivers on the aluminium housing of my CNC-mills Z-axis, what's give enough thermal stability to hold the drivers and the diode below 25 degC with 20 degC ambient temperature (in my basement).

    I'm using TEC's only when the ambient temperature is above 25 degC and/or the "thermal mass" is not sufficient ...

    Viktor

  8. #18
    Join Date
    Feb 2011
    Location
    New Hampshire
    Posts
    3,513

    Default

    Thermal mass will only slow the equilibration rate. The ultimate operating temperature will depend on the rate of heat transfer by the fans into the environment (given a fixed laser power).

  9. #19
    Join Date
    Oct 2012
    Location
    Germany
    Posts
    1,059

    Default

    ... the surface of the Z-axis profile (and thermal conduction to the other parts) is big enough, to hold the temperature near to room-temp -- otherwise I'll need a fin-cooler and a fan ...

    Viktor

  10. #20
    Join Date
    Apr 2016
    Posts
    162

    Default

    Quote Originally Posted by planters View Post
    You only want to cool the modules. The TEC that is placed between the modules and the base plate below them will transfer the heat into the base plate. Cooling the entire base plate with TECs is a bad idea. It will require huge amounts of power and will overwhelm any finned heat sink if you try to reduce the temperature of the base plate more than a few degrees below ambient. Have you ever worked with TEC's? They are useful, but they are VERY inefficient.
    I don't disagree with you, like you assume I haven't worked with TECs that much and I'll take your word for it. My only question is, how would you go about using TECs for cooling diodes, in cases when passive cooling isn't enough (eg. ambient temp is too high).
    Where could you put the TECs if not under the baseplate? I assume in the video you put it that way even though it wasn't efficient just because you were experimenting?
    If you make an aluminum case for your module you could have it on top of the module and then heatsink and fan blowing air out of the top of the case. Don't know how well that would work, haven't seen a projector doing that. What I have seen is some DPPS modules having this kind of design for the modules and blowing air in the optical floor.

    I was hoping to drive the reds at about 1W each.

    Always use the largest and most densely finned heat sinks you can fit, afford and locate. The efficiency will continue to rise. Make sure your heat sink interface surface is as flat as possible. Grind it if necessary. A good flat surface will generate a suction effect if you try to pull it away quickly.
    Thanks. I can also apply thermal grease between the heatsink and baseplate if its known to make noticeable difference when moving beyond CPU cooler surface area. I know Arctos uses heatsink plaster between the module base and the baseplate but I don't know how much that improves heat transfer.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •