I had trouble understanding your post so forgive me if I'm totally off-track here. The thermal resistance theory of heatsinking is quite well known and there are a lot of articles about it in the Internet (google for: thermal resistance heatsink). You have to read and search for a while as most information is pretty scattered. I'm pretty sure that most books teaching analog electronics design contain a chapter describing these principles as well.
Unfortunately, there's not much stuff written about heat tunnels or convection (free or forced) in general - and unfortunately some of the material is just plain opinions. I'm pretty sure you have to combine theory that you find from physics books covering fluid mechanics and heat transfer. In a nutshell, the basic principles you need to find out are a) the most efficient solutions to remove heat from a specific cooling element in a specific location and b) how to create non-obstructed paths for cool air to flow in/warm air to flow away.
Unfortunately, there's not much stuff written about heat tunnels or convection (free or forced) in general - and unfortunately some of the material is just plain opinions. I'm pretty sure you have to combine theory that you find from physics books covering fluid mechanics and heat transfer. In a nutshell, the basic principles you need to find out are a) the most efficient solutions to remove heat from a specific cooling element in a specific location and b) how to create non-obstructed paths for cool air to flow in/warm air to flow away.
I guess that would be along it. Think how air flows when it hits a round obstacle: If you blow air across the cylinder, the forced air is in contact with only about an half of the cylinder's surface while blowing air along the cylinder creates more contact area with forced air and the cylinder. However, the top of the cylinder area is a serious obstruction for air flow. This might not be as bad as it seems. I assume you talk about those capacitors, right? The typical conduction path for most heat in them is the closed, flat (metal) end so you likely wish to force a turbulent airflow to it. However, this might create some problems with channeling the "used", warm air away in a laminar, non-obstructed flow, though. Well, since I can't provide any exact science you got my opinion. Anyway, I have a strong feeling that the best solution depends a lot on the layout you use. Maybe you should add a temperature probe to the cylinder and test both configurations (along and across).
"Truth is one, paths are many ..." - The Dalhia Lama.
1) The best cooling technic is not to generate too much heat in the first place.
2) Fans make noise. Brushless fans make the least amount of noise (AC driven w/magnets). Fans with hard coated bearings and lower RPM are the quietest, ball bearing fans make more noise.
3) Thermal conductivity is enhanced by larger gaps between hot parts. The intake port should be larger than the exaust port on the box, otherwise your device becomes a "shop vac" and gathers dust.
4) If your caps are heating up, use a higher voltage rated caps and more of them.
I am building 1U (one up, single space) rack mounted amps and understand your problems. Another "trick" is to simply use an amp that is more thermally efficient. Example: I use AussieAmplifiers.com NX500 with +/- 58 VDC rails for an output of around 300 watts into 4 ohms, 200 watts into 8 ohms ... and no heat sinks and no fans at all!!. I just bolt 'em down to the bottom plate of the aluminum rack chassis. The reason this works is: reduced voltage on a thermally efficient amp of better than 40% efficient at full, continuous, "pink noise" power = higher efficiencies when driven by music and run 24/7/365 ...
The best science is lab science = real prototypes that perform as expected ...
1) The best cooling technic is not to generate too much heat in the first place.
2) Fans make noise. Brushless fans make the least amount of noise (AC driven w/magnets). Fans with hard coated bearings and lower RPM are the quietest, ball bearing fans make more noise.
3) Thermal conductivity is enhanced by larger gaps between hot parts. The intake port should be larger than the exaust port on the box, otherwise your device becomes a "shop vac" and gathers dust.
4) If your caps are heating up, use a higher voltage rated caps and more of them.
I am building 1U (one up, single space) rack mounted amps and understand your problems. Another "trick" is to simply use an amp that is more thermally efficient. Example: I use AussieAmplifiers.com NX500 with +/- 58 VDC rails for an output of around 300 watts into 4 ohms, 200 watts into 8 ohms ... and no heat sinks and no fans at all!!. I just bolt 'em down to the bottom plate of the aluminum rack chassis. The reason this works is: reduced voltage on a thermally efficient amp of better than 40% efficient at full, continuous, "pink noise" power = higher efficiencies when driven by music and run 24/7/365 ...
The best science is lab science = real prototypes that perform as expected ...
Yep all of that except I said "Rack gear" not amplifier. Its actually an industrial battery charger. It needs to charge and then dump that power. There is no choise but to make heat. Because of the number of batteries we need to mount them together in an inclosed space. They are long. I have the choise of blowing across of along.
I also need to figure out heatsinks about:- across or along , but I think that information is easier to find.
I also need to figure out heatsinks about:- across or along , but I think that information is easier to find.
The only way you can know for sure is by testing. That is what most of the science of fluid dynamics is based on anyway. They set up a lab with lots of temperature probes, look at the results, and develop empirical formulas for engineers to use.
You can look at fluid dynamics and heat transfer textbooks. There are actually a lot of good pictures in the textbooks that will show what happens to an airflow when it strikes a cylinder certainly from the side and probably the face of it as well -- but testing will still be the only way to know for sure.
JJ
You can look at fluid dynamics and heat transfer textbooks. There are actually a lot of good pictures in the textbooks that will show what happens to an airflow when it strikes a cylinder certainly from the side and probably the face of it as well -- but testing will still be the only way to know for sure.
JJ
" ... all of that except I said "Rack gear" not amplifier. ... actually an industrial battery charger. ..."
Same, same ... got sinks, got heat, got a tight space = reduce power output or get a quiet fan ... with inlet bigger than the outlet ... if you got hot caps, get higher rated caps ...
Anything that controls or "pinches off" electrons from their normal desire to flow from - to + is, in effect, an amplifier ... even power supplies with sophisticated current regulation ...
"Truth is one, paths are many ..." - The Dalhia Lama.
Same, same ... got sinks, got heat, got a tight space = reduce power output or get a quiet fan ... with inlet bigger than the outlet ... if you got hot caps, get higher rated caps ...
Anything that controls or "pinches off" electrons from their normal desire to flow from - to + is, in effect, an amplifier ... even power supplies with sophisticated current regulation ...
"Truth is one, paths are many ..." - The Dalhia Lama.
here comes opinion.
flow along a cap will be difficult to arrange.
flow across a cap seems more natural.
Mount the caps horizontally and allow natural convection to create a vertical flow pattern upwards. Unfortunately this will be substantially laminar and the boundary layers will insulate the caps from the cooling air.
Blow the air through the gaps in the caps and at least part of the flow will be turbulent. The turbulence will strip off most of the thick insulating layer and allow better cooling as well as more cool air passing by, so you win twice.
The faster you blow the larger the perimeter that will be in turbulent air flow.
flow along a cap will be difficult to arrange.
flow across a cap seems more natural.
Mount the caps horizontally and allow natural convection to create a vertical flow pattern upwards. Unfortunately this will be substantially laminar and the boundary layers will insulate the caps from the cooling air.
Blow the air through the gaps in the caps and at least part of the flow will be turbulent. The turbulence will strip off most of the thick insulating layer and allow better cooling as well as more cool air passing by, so you win twice.
The faster you blow the larger the perimeter that will be in turbulent air flow.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.