Aleph X heat sink thread

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Sorry for the length.

BrianGT said:


If you consider that a 3" section is 0.5 and it is a linear scale, then it would be around 0.14-0.15.

I guess that the top picture is assuming a single mounting point, since it is logarithmic.

--
Brian

It is, especially in real life, not logarithmic nor linear but rather assymptotic. The way I understand it works is that since air heats up and cooling decreases with the difference in sink and air temp as well as heat distribution is not ideal the increase in length will not be linear and you will also reach a point where "diminishing returns" occur. This means that for an imaginary heatsink at 3" with 0.5 C/W we might have a C/W of 0.2 at 12 " but even at 48" (a VERY long sink) the value might only be 0.15 which is not very cost effective. That is why we take two of the 0.2 sinks and put them side by side rather than on top of eachother which yields a near perfect linearity (ref A75 article if you wish) as far as C/W goes. Peter's preference for mono blocks with one sink per side is perfect implementation of this better way of using the total length.

In this case it is, IMHO, clear that it is roughly 0.27 for the 11" as mentioned and the 0.14 figure is for forced cooling. But at the same time it would be about 0.14 for two 11" pieces but side by side.

/UrSv

Peter, the two sinks are at some 0.14 combined according to the diagram and the A75 is at about 170 W which gives 170*0.14=24 degrees increase which indeed is just where we would be when sizing is perfect with respect to temp and cost. Nelson calculated 0.1125 and 0.06 with fans for a stereo amp giving us some 0.12 for a mono. We are in the same arena. Your A75 was just right with regards to sinks.
 
Retired diyAudio Moderator
Joined 2002
UrSv said:
Peter, the two sinks are at some 0.14 combined according to the diagram and the A75 is at about 170 W which gives 170*0.14=24 degrees increase which indeed is just where we would be when sizing is perfect with respect to temp and cost. Nelson calculated 0.1125 and 0.06 with fans for a stereo amp giving us some 0.12 for a mono. We are in the same arena. Your A75 was just right with regards to sinks.

If I wanted a chassis that is capable of double that, 340W, and I wanted four heatsinks with two heatsinks per side, is there any way to mount them and achieve 340W of dissipation?

Also, due to the large number of devices, wouldn't the heatsink be more efficient due to the multiple mounting points?

Also by your logic above, if I had 4 lengths of this heatsink that were 3" in length, and they were all mounted on different sides of the chassis, would I achieve a C/W of 0.125?

Another factor that is ignored throughout all of this is the dissipation of the aluminum chassis. This is a good deal of surface area also.

Many questions, but no answers... :confused:

--
Brian
 
BrianGT said:


If I wanted a chassis that is capable of double that, 340W, and I wanted four heatsinks with two heatsinks per side, is there any way to mount them and achieve 340W of dissipation?

Also, due to the large number of devices, wouldn't the heatsink be more efficient due to the multiple mounting points?

Also by your logic above, if I had 4 lengths of this heatsink that were 3" in length, and they were all mounted on different sides of the chassis, would I achieve a C/W of 0.125?

Another factor that is ignored throughout all of this is the dissipation of the aluminum chassis. This is a good deal of surface area also.

Brian

Yes, two sinks side by side on each side would yield almost 1/4 of what one sink would with a slight loss due to ambient temp sharing per side. This implies the same number of devices per sink. Same thing with the second set-up which would be very near ideal and would yield almost exactly 1/4 of the individual sinks. The multiple mounting point is also true as a larger are of the sink is heated and losses due to heat distribution is somewhat reduced. Also, mounting closer might yield similar sink average temp but increases junction temp.

For the chassis it needs to be thermally coupled to the sinks if dissipation should be useful. There is a reason why we all hunt down the best Mica/Kapton/Silver paste/Silpads...

/UrSv
 
Peter-wrt your quote
"I just called R-Theta and asked for a quote on 100 and 150 pcs. I should get the answer tommorow. They would be shipping directly to US and shipping payment would be done on the receiving end."

Are these 11" sections of R-Theta 65340 (ie, 11"x11"x2.3", which appears to be the dimensions of your A75 heatsinks)?

I have ordered the Aleph-X boards and am also assembling the Aleph 2's as well. I would be interested in 4-8 of them assuming I could use 2 per amp.

SteveA
 
BrianGT said:
...is there any way to mount them and achieve 340W of dissipation?
-- Brian
Use of screws is the recommended method for mounting them:D

Seriously, what's the question? If two work well to dissipate 170W, then 4 work well to dissipate 340W.

And keep in mind.... The 340 watts will be dissipated, no matter what. The only question is, at what temperature for the devices? Heat is energy, more or less. Just multiply by time (and there's no stoppin' time!) You put heat into the heatsink via the semiconductor devices and the temperature goes up until an equilibrium is reached where it transfers to the air at the same rate that it goes into the sink. At that point the temperature will stop rising. There's no energy storage there other than the mass of the heatsink times its heat content, and that isn't much for the mass of aluminum involved.

Yes, multiple mounting points for the semiconductors would be better....But maybe not quite enough to offset the fact that the data we have is for differential temp btw heatsink and air of 50degrees C. We don't want to run them that hot (70-75C), right?

And yes, 4- 3in. lengths gets you 1/4 the thermal resistance or .125 in this case.

The chassis won't get you much additional heatsinking, but that's probably ok since it would make everything inside that much hotter.

I strongly recommend going to the Aavid site and reading a couple of their papers under the heading "Technical Info"
 
Hello Peter,

First of all thank you for looking into this. Could you please add me to the list as I could use 2 pieces of the 65340. Also I seem to recall from another thread that you bought many pieces many years ago and they were 12" long not 11". Am I correct and if so does 12" not make better yield of a standard 8 foot extrusion?

Thanks,

Martyn
 
Peter,
Showing Wayne's aleph and then your A75 is cruel and unusual punishment! ;) Santa told me no presents for you, you bad boy!
However, I don't think that assembly with Wayne's heastsinks would be more complicated. It seems to me that all you would have to do is to make the main rectangular bars longer to include both heatsinks.

You did get my email about the order, didn't you?
 
grataku said:
Peter,
Showing Wayne's aleph and then your A75 is cruel and unusual punishment! ;) Santa told me no presents for you, you bad boy!
However, I don't think that assembly with Wayne's heastsinks would be more complicated. It seems to me that all you would have to do is to make the main rectangular bars longer to include both heatsinks.

You did get my email about the order, didn't you?

I got your e-mail.

The reason I showed both pics is for you to avoid disapointments, when unpacking your presents.;) The assembly with R-Theta sinks and your boards doesn't really require any angles or bars. I will mount the boards directly to the heatsink using output devices and holes in PCB (for standoffs).

Now to hold both sinks together I can use 4 crossbars (like in my gold Aleph5) and attach 4 panels directly to those bars. With the other sinks it wouldn't be that easy.;)
 

Attachments

  • a5.jpg
    a5.jpg
    69.3 KB · Views: 1,079
Peter,

Can you get the R-Theta sinks in shorter lengths? How about an extrusion wider than 11 inch? I'm building low powered AX's and will only be dissipating around 100W per channel, so 2 of these large sinks is really too much. I want to go for a fairly compact chassis (maybe stereo chassis?), even if it means the sinks run slightly hot. My current choice was going to be the Conrad Engineering MF35 sinks from Australia, using only one per channel. From what I've seen, I may still go with these, even though the R-Theta's are so nice.
 
Peter,
I like your A75 chassis that is held together by the rectangular bars, what are you using? They look to be 2 1/2 in x 3/4. Since I don't need to mount transistors on the bars they could be 1/2 x 1 bars (about 1.5 bucks each at speedy metals).
I can mill them to size no problem. Milling the plates is a little bit tricky-er, it takes too long and there are many people that need to get on the machine. These plates have to be at least 3/16 or a 1/4 to get the wanted "powerhouse" effect. Anodization is also a big PITA. You do cut 1/4 inch plate with the some type of saw as I recall. I wish I could do that.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.