hey everyone. i was at a surplus shop today and i picked up some heatsinks cheap ($20). they are in great condition, just with a few holes. i found the profile from AAVID. here it is:
http://www.aavidthermalloy.com/bin/exdisp.pl?Pnum=66880&LengthUnits=in&ExLength=3&airflow=57.2
if that link doesnt work, its product #66880.
Its thermal resistance is 1.18 @ 3" length. I have two 13.25" lengths. I want to do a stereo amp either for rear channels or for bi-amping my fronts. could i pull off a stereo aleph 5?
http://www.aavidthermalloy.com/bin/exdisp.pl?Pnum=66880&LengthUnits=in&ExLength=3&airflow=57.2
if that link doesnt work, its product #66880.
Its thermal resistance is 1.18 @ 3" length. I have two 13.25" lengths. I want to do a stereo amp either for rear channels or for bi-amping my fronts. could i pull off a stereo aleph 5?
cowanrg said:
Are they anodized? The web applet gives a thermal impedance of a 13.25" length at .56 C/W. I'm pretty sure that's for the unfinished aluminum heatsink as Rod Elliot's spreadsheet comes up with pretty much the same number.
You really don't want the temp of the sink to be much over 50C (you can, but then the parts are running very hot, and it will HURT to touch it, I know it's time to stop steaming the milk for my morning Latte when I can't hold the pitcher...)
So, assuming at 25C (77F) room, that's a rise of roughly 25C. 25/.56 = 44 watts dissipation. No, an Aleph 5 will not fit on that heatsink without serious fan cooling.
If the sink is nicely black anodized the thermal impedance may be as low as .25 C/W 25/.25 = 100 Watts...and an Aleph 5 will still not work without fan cooling.
Scott
The size of those heat sinks sound similar to mine. My Sinks are 9 x 12. I operate at 1.8 amps bias per channel. Quite warm but can hold it for about 20 seconds. If it gets too hot put a small fan on it, change the bias, or change the number of fets, after all its DIY.
The driver circuits are about the same.
The driver circuits are about the same.
Attachments
cowanrg said:wow, i wasnt aware an aleph 5 had THAT much heat! would i be fine then with an aleph 3 in stereo? i was hoping to get a bit more power out of this then just 60 watts into 4 ohms, but it looks like thats the most i could get
oh, and yes, they are anodized black.
Thats all *if* you want it convenction cooled. Fan cooling helps A LOT. I have some truely large heatsinks 15.5X19x1.5 with 40 cooling fins running the long dimension, 23 pounds each. I purchased them before I found Rod Elliot's spread sheet. (It's pretty accurate I must say.)
I thought they would work for me, but I decided to test them before I started drilling. I bought 26 1.5K Ohm 25Watt Aluminim cast resistors from the local surplus place, which if you wire in parallel and plug into the wall (in the US) is good for ~250 watts of dissipation about that of an Aleph 5 if I remember correctly. I clamped them to said heatsink (I have 3) and off I went.
It took about 3 hours to reach thermal equalibrium, which was a temp rise of 42.5 degrees c, or .17 C/W each. Way too hot. By Rods spreadsheet if I get them anodized they will be .09 C/W, but it will be hard to realize that in real life I bet.
Now, I turned on the ceiling fan to see how much a little bit of air cooling would help was enlightening to say the least. Temp rise of the sink was 17C, for a thermal impedance of .068 C/W, and that was with general air movement, nothing directed at the heat sink itself.
Now, 250 W without the fan on generated quite a chimney (spelling?) effect on the sink, if you held your hand over it you could feel the hot air coming up off of the sinks.
When this was all said and done (and the temp of the room was 5 degrees hotter) I decided that an Aleph 3 on smaller sinks, fan cooled was for me..and I'd save the big sinks for a big AB biased amp where their long time constants could be put to use.
I also dug out my old power electronics college text to check out the equations that govern heatsinks. It turns out there are some terms in the denominator of the impedance equation that are a power of 4...that's N^4. Heatsinks have to get quite big to get a lower impedance and they must get bigger wayyyyy faster than you think.
Scott
Scott,
sounds like you are describing the Chatham heatsink. I use two
of those for 1 channel of Aleph 2... and they get toasty! I remember
thinking that it was overkill to use 2 sinks per channel, but it
actually worked out quite well. So the rule of thumb is, when in
doubt, double it! - you can never have too much heatsink. I think
the Chatham heatsink suffers because it's cooling fins are to small
at 1.5"... i bet it would help a lot if they were 3".
m.
sounds like you are describing the Chatham heatsink. I use two
of those for 1 channel of Aleph 2... and they get toasty! I remember
thinking that it was overkill to use 2 sinks per channel, but it
actually worked out quite well. So the rule of thumb is, when in
doubt, double it! - you can never have too much heatsink. I think
the Chatham heatsink suffers because it's cooling fins are to small
at 1.5"... i bet it would help a lot if they were 3".
m.
cowanrg said:wow, i wasnt aware an aleph 5 had THAT much heat! would i be fine then with an aleph 3 in stereo? i was hoping to get a bit more power out of this then just 60 watts into 4 ohms, but it looks like thats the most i could get
oh, and yes, they are anodized black.
Oops...I guess I didn't answer your question...
Going by the Aleph 3 service manual each fet is biased for 1A, making 2A of current. With 25V rails that works out to 100W dissipation...which if the spreadsheet is right should be reasonable for your heatsinks. If they run too hot, take one of those new fangled super quiet PC fans and run it under the recommended voltage. They are silent when doing this, and you'd be surprised at how much it helps.
Scott
cowanrg said:
Well actually it's 4. Two in the constant current source, two in the output stage. So 25Watts each isn't so bad. With a thermal impedance from junction to sink of 1.2 C/W or less youre going to keep the junction temp at 80 degrees...which is hot...but the IRF FETS do pretty good with temp so they should be okay.
(I should have said each FET leg, or something like that, sorry.)
Scott
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.