hey all.
im getting along quite nicely with the aleph2's and im at the part where i need to start putting them together. im wondering exactly how you calculate heat transfer properties of a heatsink. i found some i would like to use on aavid.com, but they only have a C/W rating of 0.54 (for a 3 inch section). from my calculations, with a monoblock, 1 heatsink per side, using 3 8" sections of this heatsink, i would get a 11.4 C temperature increase per heatsink... is this ok? is this calculation even right?
in trying to figure out which heatsink would be needed for the 3 2-channel aleph2's, i am unable to calculate any reasonable values... basically, the aleph2 generates 300 watts of heat PER side, so there needs to be a heatsink that can get rid of 300 watts... what is an appropriate temperature gain? could 100F increase in temp be fine? what would be a nice SINGLE heatsink that could tolerate 300 watts of heat?
im getting along quite nicely with the aleph2's and im at the part where i need to start putting them together. im wondering exactly how you calculate heat transfer properties of a heatsink. i found some i would like to use on aavid.com, but they only have a C/W rating of 0.54 (for a 3 inch section). from my calculations, with a monoblock, 1 heatsink per side, using 3 8" sections of this heatsink, i would get a 11.4 C temperature increase per heatsink... is this ok? is this calculation even right?
in trying to figure out which heatsink would be needed for the 3 2-channel aleph2's, i am unable to calculate any reasonable values... basically, the aleph2 generates 300 watts of heat PER side, so there needs to be a heatsink that can get rid of 300 watts... what is an appropriate temperature gain? could 100F increase in temp be fine? what would be a nice SINGLE heatsink that could tolerate 300 watts of heat?
cowanrg said:
Are you insane? One heatsink? That would have the be a very large heatsink, unless you are considering active cooling.
What is your desired chassis size? You could use multiple extrusions to make the sides of the case, like:
http://www.passdiy.com/images/gallery/a2-p1-f2.jpg
but taller and deeper, giving you 600W total dissipation.
I am using 4 of these per channel:
http://www.aavidthermalloy.com/bin/exdisp.pl?Pnum=65615&LengthUnits=in&ExLength=8&airflow=57.2
I picked them up surplus for $25 each
If you do in for a large order of heatsinks, I wouldn't mind joining in on the order to get a set for myself if the prices are reasonable. Let me know if you are interested in doing this.
--
Brian
well, when i say one heatsink, i mean that there is one block of heatsinks.
the chassis dimensions depend on the heatsinks 🙂
ill first have to figure out what will cool the amp, then see where i could find it, and how much. could anyone answer my questions above? that would help me out a LOT.
the chassis dimensions depend on the heatsinks 🙂
ill first have to figure out what will cool the amp, then see where i could find it, and how much. could anyone answer my questions above? that would help me out a LOT.
What i have read somewhere an increase
of max 40 degrees Celsius (C/W) is desirable.
I would guess you can accept up to 60 degrees Celsius. But I refuse to count it in Farenheit. To much trouble for me.
Look at each section.
How much Watt will this section dissipate?
Watt x C/W-value= degrees Celsius (temp increase)
If you have 8 sections and 300 watt.
Each section takes care of 300/8= ? (help me please) watt
By the way, Anders Celsius was a Swedish scientist.
😉 Hurray for Sweden! Buuh, for Fahrenheit. Sounds German
of max 40 degrees Celsius (C/W) is desirable.
I would guess you can accept up to 60 degrees Celsius. But I refuse to count it in Farenheit. To much trouble for me.
Look at each section.
How much Watt will this section dissipate?
Watt x C/W-value= degrees Celsius (temp increase)
If you have 8 sections and 300 watt.
Each section takes care of 300/8= ? (help me please) watt
By the way, Anders Celsius was a Swedish scientist.
😉 Hurray for Sweden! Buuh, for Fahrenheit. Sounds German
I use the 250mm(H) x 230mm(W) x 65mm(D) 18-fin heat sink for 40W dissipation at 15C temp increase. If I want to keep the 15C temp increase with 300W, I might need 7 or 8 of this size for one side. 🙂
JH
JH
I would go with three tall ones of these per channel/side:
http://www.aavidthermalloy.com/bin/exdisp.pl?Pnum=65615&LengthUnits=in&ExLength=12&airflow=57.2
They go together nicely and are heavy:
http://brian.darg.net/ax-chassis
--
Brian
http://www.aavidthermalloy.com/bin/exdisp.pl?Pnum=65615&LengthUnits=in&ExLength=12&airflow=57.2
They go together nicely and are heavy:
http://brian.darg.net/ax-chassis
--
Brian
The essential temperature is inside the transistor.
Not on the case of the transistor, not on the heatzink.
Between the inner of a T and the case, there is a t-resistance.
It is also in C/W. Can be like 0.7-1.0 C/W
The mounting-isolation also have C/W. Usually like 0.2-0.4
Then heatzink. Which is in contact with the air.
So temp inside the transistor is= temp of air + total resistance x Watt
Total temp resistance is:
Rth of heatzink
Rth of Isolation
Rth of Transistor (case to inner)
All this added is the total temp-resistance
between air and inside the transistor
Not on the case of the transistor, not on the heatzink.
Between the inner of a T and the case, there is a t-resistance.
It is also in C/W. Can be like 0.7-1.0 C/W
The mounting-isolation also have C/W. Usually like 0.2-0.4
Then heatzink. Which is in contact with the air.
So temp inside the transistor is= temp of air + total resistance x Watt
Total temp resistance is:
Rth of heatzink
Rth of Isolation
Rth of Transistor (case to inner)
All this added is the total temp-resistance
between air and inside the transistor
if i have read you correctly, after calculation, less than 40 C overall change in heatsink temp is acceptable?
if this is the case, i was looking at around 20 C, therefore it should work out FINE.
if this is the case, i was looking at around 20 C, therefore it should work out FINE.
99% sure he will have a comment for us
i hope, because im really worried about this! if you say 40 is ok, and im sticking around 20, that should be MORE than fine... but we will wait for the great one to have his say...
i still need to FIND these things after i figure out what i need. 🙁 i think im gonna start a heatsink/capacitor company. just make huge power supply caps and heatsinks. it seems like its the two things no one can ever find!
i still need to FIND these things after i figure out what i need. 🙁 i think im gonna start a heatsink/capacitor company. just make huge power supply caps and heatsinks. it seems like its the two things no one can ever find!
Absoluely right
Seems you have to hire a private investigator
to find those "big" things.
At least here in Sweden
Not to mention Guatemala, Tahiti or Tunisia.
Heatzink profiles are much cheaper, when bought in 1000 mm lenght.
But then you need some equipment to cut them to suitable size.
Seems you have to hire a private investigator
to find those "big" things.
At least here in Sweden
Not to mention Guatemala, Tahiti or Tunisia.
Heatzink profiles are much cheaper, when bought in 1000 mm lenght.
But then you need some equipment to cut them to suitable size.
i have the equipment to cut them, i just need to find them, in any length...
i guess ill give aavid a call/email and see where i could find some of their particular styles. does anyone else know of a place in the US that is actually a distributor, not just a supplier?
i guess ill give aavid a call/email and see where i could find some of their particular styles. does anyone else know of a place in the US that is actually a distributor, not just a supplier?
Former discussions on this topic
You could do a search for posts with "heastink" or "temperature" by Nelson Pass, but....
here are a few snips that I saved from previous discussions. I think that the conclusion is that 50deg. C on the sink next to the device is a good target. That would be a 25-30 deg. C rise, not 40.
1) Grey Rollins: "Don't overlook the obvious...Nelson runs his devices (in the Alephs, at any rate) at about 22-25W dissipation each. Take the rail, that general dissipation figure, and calculate the bias current accordingly. "
Nelson Pass: " As usual, Grey nails it down. At 25 watts or
so, the transistor "never" fails. At 50 watts,
we see some failures. "
2) (Discussing biasing) NP: "As the temperature reaches 50 deg C. or so over the first hour, this tapers off about 10% to the final static value. As the temperature rises from about 25 deg C to the 50 deg figure..."
3) And the ultimate: NP: "I try to run the heat sinks around
50 deg C, which is the point at which you can only touch them for
a few seconds. "
Hope that helps.
You could do a search for posts with "heastink" or "temperature" by Nelson Pass, but....
here are a few snips that I saved from previous discussions. I think that the conclusion is that 50deg. C on the sink next to the device is a good target. That would be a 25-30 deg. C rise, not 40.
1) Grey Rollins: "Don't overlook the obvious...Nelson runs his devices (in the Alephs, at any rate) at about 22-25W dissipation each. Take the rail, that general dissipation figure, and calculate the bias current accordingly. "
Nelson Pass: " As usual, Grey nails it down. At 25 watts or
so, the transistor "never" fails. At 50 watts,
we see some failures. "
2) (Discussing biasing) NP: "As the temperature reaches 50 deg C. or so over the first hour, this tapers off about 10% to the final static value. As the temperature rises from about 25 deg C to the 50 deg figure..."
3) And the ultimate: NP: "I try to run the heat sinks around
50 deg C, which is the point at which you can only touch them for
a few seconds. "
Hope that helps.
Heatsink source
Forgot to mention:
There has been discussion here about M&M Metals that makes sinks and is willing to sell to individuals. Look at the Aleph 2's in the gallery. Somebody there used their sinks. Can't just now recall the name of the fellow.
http://www.mmmetals.com/index.html
Forgot to mention:
There has been discussion here about M&M Metals that makes sinks and is willing to sell to individuals. Look at the Aleph 2's in the gallery. Somebody there used their sinks. Can't just now recall the name of the fellow.
http://www.mmmetals.com/index.html
And Lastly...
Read this thread too. It pretty well covers it.
http://www.diyaudio.com/forums/showthread.php?s=&postid=21829&highlight=temperature#post21829
Read this thread too. It pretty well covers it.
http://www.diyaudio.com/forums/showthread.php?s=&postid=21829&highlight=temperature#post21829
Re: And Lastly...
thank you vpharris. this gives me a great idea of what needs to be done. now i just need to find those damn heatsinks! called a few places, and so far the best best is $950 for an 8ft section. needless to say im still looking...
thank you vpharris. this gives me a great idea of what needs to be done. now i just need to find those damn heatsinks! called a few places, and so far the best best is $950 for an 8ft section. needless to say im still looking...
vpharris said:
Transistors break down at a certain temp.halojoy said:
What is the max temp for the working space inside a transistor?
It is expressed as max Tj in the datasheeet.
j stands for junction and T for temp.
Most power transistor have 150-200 degrees C.
That value a transistor can survive.
That is not to say that it is working well near that temp.
How can I try to calculate Tj if I cannot messure it?
Heatsink temp can be messured. Not on the spot where transistor is in direct contact,
but very near.
-The heat travels through transistor to transistors surface, the case.
-Then it travels through the isolation material.
Then heat hits the heatsink.
The transistor have a thermal resistance, in same way as heatsinks.
Termal resistance is written Rth in datasheet. th stands for thermal.
The isolation material have also Rth.
Rth is messured in C/W. Temp increase per Watt.
If we want to find out how much temp is higher/lower after heat has traveled
through a material, we get: Watt x Rth= temp difference.
Sample:
heatsink messured: 80 degrees Celsius
Isolation Rth : 0.4 C/W
Transistor Rth : 1.0 C/W
Power in transistor : 50 Watt
temp inside transistor will be
[50 x (0.4 + 1.0)] + 80 = 150 degrees
That is high for TO220,
but a TO3, I imagine, can work rather well at that temp.
Nelson have to comment on that. He should know.
This is an accurate enough calculation with realistic
numbers.
Personally, I try to not run heat sinks greater than about
50 deg C, and cases above 25 watts, which gives me about 90,
maybe 100 deg C for chip temperature.
Most Pass Labs products run semis in the 10-15 watt range.
numbers.
Personally, I try to not run heat sinks greater than about
50 deg C, and cases above 25 watts, which gives me about 90,
maybe 100 deg C for chip temperature.
Most Pass Labs products run semis in the 10-15 watt range.
- Status
- Not open for further replies.