Hello!
We are a couple of students from UIS (University of Stavanger), who wants to build an aleph amplifyer each. Some of us wants to build Aleph 4, and some wants to build Aleph 5. The problem is (as usual with class A designs), the choosing of heat sinks.
We have looked into some heatsinks delivered through elfa (http://www.elfa.se/no).
My problem is that i can't really figure out how to read the datasheet, because the graph showing the temperature rise vs effect is only shown for defined lengths, and we want to use 1 meter of the heat sink (4x25 cm)
Could anyone bee so kind to look into the datasheet, to see if one of the element called KS160 or KS200 is suitable for the aplication?
The link to the datasheet for both heatsinks (each in 100mm length):
http://www.elfa.se/pdf/75/07562267.pdf
Regards Vidar
We are a couple of students from UIS (University of Stavanger), who wants to build an aleph amplifyer each. Some of us wants to build Aleph 4, and some wants to build Aleph 5. The problem is (as usual with class A designs), the choosing of heat sinks.
We have looked into some heatsinks delivered through elfa (http://www.elfa.se/no).
My problem is that i can't really figure out how to read the datasheet, because the graph showing the temperature rise vs effect is only shown for defined lengths, and we want to use 1 meter of the heat sink (4x25 cm)
Could anyone bee so kind to look into the datasheet, to see if one of the element called KS160 or KS200 is suitable for the aplication?
The link to the datasheet for both heatsinks (each in 100mm length):
http://www.elfa.se/pdf/75/07562267.pdf
Regards Vidar
Hello Vidar
There are several ways of "calculating" the heatsink needed for an amp - one is to screw the calculations and make it really big - NP is a fan of big heatsinks for sure, but he also needs to run a buisness, so he needs his calculator, as sinks are expensive - but I think you have figured that out by now ;-)
Beeing engineering students you migth want to follow the calculator way.
I cant give you precise conclusions based on the datasheet, but mabye others can. What I can tell you, is that you want a max temp incline on your amps sinks at around 30 kelvin/centigrades. I assume a ambient temp @ 20 (can be lower in norway I know ;-) )and a max temp om the sinks @50 degrees centigrades. This will mean chip temps in the order of mabye 60 degrees. This you donw want higher for long term operation. Search the forum for this subject, as it is of high relevance in here, and often discussed.
Ok - 30 degrees rise - that means for an amp dispassing say 100 watts you need a sink with a C/W value around 0,3 - that means for every watt sinked, you get a 0,3 degree rise in temperature. This curve is not linear, so imply the safe rule of thump. Also, the datasheets are often optimistic, and I have read in here that a 0,3C/W heatsink is more often 0,35 or 0,40 - but again, search.
This means that for a A5 that throws 300 watts of heat (excluding losses in trannie and bridges) the total C/W value of the amp needs to be in the order of 0,1c/W - here an aluminium chassis, mabye with holes in top and bottom plate helps the heatsinks considerely
I hope this helped you along the way - if you search www and this forum, you should find what else you need.
Cheers !
Buhl
By the way - do you have any snow yet ???
There are several ways of "calculating" the heatsink needed for an amp - one is to screw the calculations and make it really big - NP is a fan of big heatsinks for sure, but he also needs to run a buisness, so he needs his calculator, as sinks are expensive - but I think you have figured that out by now ;-)
Beeing engineering students you migth want to follow the calculator way.
I cant give you precise conclusions based on the datasheet, but mabye others can. What I can tell you, is that you want a max temp incline on your amps sinks at around 30 kelvin/centigrades. I assume a ambient temp @ 20 (can be lower in norway I know ;-) )and a max temp om the sinks @50 degrees centigrades. This will mean chip temps in the order of mabye 60 degrees. This you donw want higher for long term operation. Search the forum for this subject, as it is of high relevance in here, and often discussed.
Ok - 30 degrees rise - that means for an amp dispassing say 100 watts you need a sink with a C/W value around 0,3 - that means for every watt sinked, you get a 0,3 degree rise in temperature. This curve is not linear, so imply the safe rule of thump. Also, the datasheets are often optimistic, and I have read in here that a 0,3C/W heatsink is more often 0,35 or 0,40 - but again, search.
This means that for a A5 that throws 300 watts of heat (excluding losses in trannie and bridges) the total C/W value of the amp needs to be in the order of 0,1c/W - here an aluminium chassis, mabye with holes in top and bottom plate helps the heatsinks considerely
I hope this helped you along the way - if you search www and this forum, you should find what else you need.
Cheers !
Buhl
By the way - do you have any snow yet ???
Hi Vidi,
The graphs in the data sheet give you the relationship between power dissipation and temperature rise (Temperature of Heat Sink - Room Temperature) when a certain length of the heat sink is used.
Take KS 300.6 as an example:
If the total power dissipation of all transistors mounted on the heat sink is 300W and the Temperature rise should be less than 40 Deg.(just for example, usually we would like to keep the temperature of the device below 70 deg C), you should choose a 500 mm long heat sink. Why? From the graph, take the x axis and find P = 300, the 500 graph gives you 33 deg. and the 300 graph gives you 42 deg. You may choose a 300 mm heat sink, but the devices will run a little hotter. At least this gives you some ideas how big the heat sink should be.
This is only the theoretical figures. There are other factors that will increase the thermal resistance, like the contact between transistor and heat sink, coatings in heat sink etc. Usually you will select something bigger to ensure the devices are running cool.
Heat sink spec. is usually given in terms of "deg. C/watt" for a certain length of that kind of heat sink. For example, a heat sink of 75mm long is 0.5 C/W. If you have 300W dissipation, the temperature rise will be 300W * 0.5 = 150 deg C. You can use a longer heat sink to lower the temperature. If you use 300mm long heat sink, the temperature rise will be 150 / 300 * 75 = 37.5 Again this is only theoretical figures.
Hope this helps.
The graphs in the data sheet give you the relationship between power dissipation and temperature rise (Temperature of Heat Sink - Room Temperature) when a certain length of the heat sink is used.
Take KS 300.6 as an example:
If the total power dissipation of all transistors mounted on the heat sink is 300W and the Temperature rise should be less than 40 Deg.(just for example, usually we would like to keep the temperature of the device below 70 deg C), you should choose a 500 mm long heat sink. Why? From the graph, take the x axis and find P = 300, the 500 graph gives you 33 deg. and the 300 graph gives you 42 deg. You may choose a 300 mm heat sink, but the devices will run a little hotter. At least this gives you some ideas how big the heat sink should be.
This is only the theoretical figures. There are other factors that will increase the thermal resistance, like the contact between transistor and heat sink, coatings in heat sink etc. Usually you will select something bigger to ensure the devices are running cool.
Heat sink spec. is usually given in terms of "deg. C/watt" for a certain length of that kind of heat sink. For example, a heat sink of 75mm long is 0.5 C/W. If you have 300W dissipation, the temperature rise will be 300W * 0.5 = 150 deg C. You can use a longer heat sink to lower the temperature. If you use 300mm long heat sink, the temperature rise will be 150 / 300 * 75 = 37.5 Again this is only theoretical figures.
Hope this helps.
Hello Vidar,
I just looked at the datasheet. The KS 200 is a bit thin and has a thermal resistance that is too high.
The KS160 in 200mm pieces (a convenient height for an Aleph) will have a thermal resistance of around 0,73°C/w.
For an Aleph 5 with 34V supply and 2A of bias you will need to dissipate 34x2x2x2=272watts
For 35°C above ambient this means an Rth of 0,13
So six pieces of KS160 (200mm high) would be just enough. The chassis would then be 480mm deep and 200mm high (plus feet).
Making it higher won´t reduce Rth very much more.
If you look at the KS300.6 this would be better suited. Here 2 200mm pieces could keep the amp cool.
For an Aleph 4 you will need to dissipate a bit more than 400 watts. Building two monos with KS300.6 x150 would be more practical in this case.
The heatsinks do look very much like the Fisher parts......
William
I just looked at the datasheet. The KS 200 is a bit thin and has a thermal resistance that is too high.
The KS160 in 200mm pieces (a convenient height for an Aleph) will have a thermal resistance of around 0,73°C/w.
For an Aleph 5 with 34V supply and 2A of bias you will need to dissipate 34x2x2x2=272watts
For 35°C above ambient this means an Rth of 0,13
So six pieces of KS160 (200mm high) would be just enough. The chassis would then be 480mm deep and 200mm high (plus feet).
Making it higher won´t reduce Rth very much more.
If you look at the KS300.6 this would be better suited. Here 2 200mm pieces could keep the amp cool.
For an Aleph 4 you will need to dissipate a bit more than 400 watts. Building two monos with KS300.6 x150 would be more practical in this case.
The heatsinks do look very much like the Fisher parts......
William
Hi everybody!
Thank's for all your advice! I'm feeling a bit smarter now
I was not shure if I could just add the value of the heatsinks together, but giving it a closer thought it seems logically.
Then my new question will be:
Building an Aleph 5 would be a little cheaper, and considering the minimal effect gained (40w, that would be below 3dB extra volume?), you would not notice the limitations of power between an Aleph 5 and an Aleph 4? Does anyone have any thoughts about this?
Is there any reason to build the Aleph 4 in stead of the Aleph 5?
William: Then I will take your advice, and use 6 KS160 at 200 mm if we choose to build Aleph 5 (The rest of the chassi will be in aluminium, so I'm guessing that Rth will be slightly better than 0,13), and if we choose Aleph 4, I guess we have to use 4 KS300.6 at 200mm. And for insulation, I'm planning the use of metaloxide insulators, wich has a Rth less than 0.3 degrees/W.
This could be very exciting!
Buhl: I'm sorry to disappoint you. We don't have snow in Norway yet (except in the highest mountains). None of the ski resorts (downhill) has opened yet, because of the lack of snow. All we get is rain. ALOT of rain! We have 10 degrees 6th december! So, no snow on christmas eve this year! Blame the global warming!
Thank's for all your advice! I'm feeling a bit smarter now
I was not shure if I could just add the value of the heatsinks together, but giving it a closer thought it seems logically.
Then my new question will be:
Building an Aleph 5 would be a little cheaper, and considering the minimal effect gained (40w, that would be below 3dB extra volume?), you would not notice the limitations of power between an Aleph 5 and an Aleph 4? Does anyone have any thoughts about this?
Is there any reason to build the Aleph 4 in stead of the Aleph 5?
William: Then I will take your advice, and use 6 KS160 at 200 mm if we choose to build Aleph 5 (The rest of the chassi will be in aluminium, so I'm guessing that Rth will be slightly better than 0,13), and if we choose Aleph 4, I guess we have to use 4 KS300.6 at 200mm. And for insulation, I'm planning the use of metaloxide insulators, wich has a Rth less than 0.3 degrees/W.
This could be very exciting!
Buhl: I'm sorry to disappoint you. We don't have snow in Norway yet (except in the highest mountains). None of the ski resorts (downhill) has opened yet, because of the lack of snow. All we get is rain. ALOT of rain! We have 10 degrees 6th december! So, no snow on christmas eve this year! Blame the global warming!
Hi Vidi,
you could also have a look at the kapton insulators. Their Rth is only 0,07 and they won´t break as easily as the metaloxide ones.
I´ve used them in my Aleph5 and X with good results. They come ready for use with a thin film of thermal grease and keep your fingers and everything you touch clean
William
you could also have a look at the kapton insulators. Their Rth is only 0,07 and they won´t break as easily as the metaloxide ones.
I´ve used them in my Aleph5 and X with good results. They come ready for use with a thin film of thermal grease and keep your fingers and everything you touch clean
William
Hi,
I have built the Aleph 1.2 monoblocks. For this I used the Fischer SK535 heatsinks. I payed about 8000-10000Nkr for four of these including VAT and freight. Two of these would be enough for one A4. I used a lot of time finding good heatsinks. Their dimensions are: 500x200x80 mm.
If these are to expensive Seifert have somewhat cheaper heatsinks. However, for the A4 it will be big.
Best regards,
Folke
I have built the Aleph 1.2 monoblocks. For this I used the Fischer SK535 heatsinks. I payed about 8000-10000Nkr for four of these including VAT and freight. Two of these would be enough for one A4. I used a lot of time finding good heatsinks. Their dimensions are: 500x200x80 mm.
If these are to expensive Seifert have somewhat cheaper heatsinks. However, for the A4 it will be big.
Best regards,
Folke
Hi!
I'm pretty shure that my speakers are not excessivelly powerhungry (they are self built with Scan Speak speakers), they have an min impedance of 4 ohm, and about 90 dB/W.
But, regarding cabinett to the amplifiers. I have built two The End Millennium power amplifiers from LC Audio a couple of years ago (still very happy with the result!), and just for fun I visited their web site today. I was very excited when I found that they are selling the cabinet alone (without powersupply and amplifier), at 5000 DKK.
Looking into this, I see that the cabinet uses four heatsinks rated at 0,33K/W. That would make a cabinet that has a total Rth of 0.33/4 = 0,0825, pluss the Rth of the cabinet itself (front plate, and top and bottom plates)? This would be enough for either an Aleph 5 or an Aleph 4!?
Here follows the link to the cabinet:
LC Audio Cabinet
I'm pretty shure that my speakers are not excessivelly powerhungry (they are self built with Scan Speak speakers), they have an min impedance of 4 ohm, and about 90 dB/W.
But, regarding cabinett to the amplifiers. I have built two The End Millennium power amplifiers from LC Audio a couple of years ago (still very happy with the result!), and just for fun I visited their web site today. I was very excited when I found that they are selling the cabinet alone (without powersupply and amplifier), at 5000 DKK.
Looking into this, I see that the cabinet uses four heatsinks rated at 0,33K/W. That would make a cabinet that has a total Rth of 0.33/4 = 0,0825, pluss the Rth of the cabinet itself (front plate, and top and bottom plates)? This would be enough for either an Aleph 5 or an Aleph 4!?
Here follows the link to the cabinet:
LC Audio Cabinet
Attachments
Steen: Im very interested - I always wanted to build one of those...how does that class a ZAP sound compared to our sweet Zens, Alephs ect. ? Did I miss anything ?
Vidi: S**t !!! - I have booked 5 days in lillehammer/hafjell i uke 5....damn, I dont want to be sitting in the "hytte", reading old pass labs service manual those days....
The LC cabinet is 530 euros, I calculate it to roughly 4000,- Dks. - but that is still a sum of money....
Cheers!
Buhl
Vidi: S**t !!! - I have booked 5 days in lillehammer/hafjell i uke 5....damn, I dont want to be sitting in the "hytte", reading old pass labs service manual those days....
The LC cabinet is 530 euros, I calculate it to roughly 4000,- Dks. - but that is still a sum of money....
Cheers!
Buhl
Hej Hans. The Zap sounds great, those big Sanken transistors are not bad at all. The amp is also well build. I bought it just before starting building amps myself. I wouldnt buy it today, but i am not much for selling it either
It doesnt quite have the sweet soundsignature as Mr. Pass's amps do. It is very powerfull though, its the strongest 50 watter i have heard. Overall a very nice amp. Also the powerdown function is pretty handy, no need to burn a couple hundred watts when doing the dishes
I am using an Aleph 5 in the main system right now and that is going nowhere. The Zen's doesnt really cut it regarding the power needed for the ProAc clones, but the Aleph 5 handles them very well. Maybe I will sell the Zap after all. Here is the Aleph 5:http://www.diyaudio.com/forums/showthread.php?postid=1068284#post1068284 Steen
Attachments
Steenoe: Do you mean that I have to use two of the cabinets (one for each Aleph 4 monoblock)? According to WuffWaff, (and Aleph 4 user manual), I will need to get rid of about 400w of heat. Using the mentioned cabinet, with an Rth= 0,0825, will give me a temperature of 33 degrees above room temperature (to a total of about 58 degrees). Isn't this enough cooling? I have not considered the insulators and junction Rth, is this the reason to choose two cabinets?
The reason that I prefer to use a pre-built cabinet, is that I do not have any experience doing metal work, so buying a pre-fabricated cabinet would enshure me a good result. (But, this is not in the spirit of a good DIYer.. )
Time will tell! I will ask the students of mechanical engeneering if they can help me out with the metal work!
Buhl: It may not be to late to get som snow in Lillehammer! Lets hope for the best!
I asume you know about the webcam on hafjell, but I post the link here!
Hafjell Webcam
The reason that I prefer to use a pre-built cabinet, is that I do not have any experience doing metal work, so buying a pre-fabricated cabinet would enshure me a good result. (But, this is not in the spirit of a good DIYer..
)Time will tell! I will ask the students of mechanical engeneering if they can help me out with the metal work!
Buhl: It may not be to late to get som snow in Lillehammer! Lets hope for the best!
I asume you know about the webcam on hafjell, but I post the link here!
Hafjell Webcam
Looking at the chassis, I wouldnt think that it can dissipate the heat from a 100 watt stereo Aleph
These amps do get very varm. An Aleph 5 would be (just) doable. I bet that will get warm too. I didnt calculate anything, but having build several of these amps, i cant recommend using only one chassis for the Aleph 4.
But you can ofcourse try it out and see how it goes.
Steen
These amps do get very varm. An Aleph 5 would be (just) doable. I bet that will get warm too. I didnt calculate anything, but having build several of these amps, i cant recommend using only one chassis for the Aleph 4.But you can ofcourse try it out and see how it goes.
Steen
Hi,
somehow heatsinks always get hotter than calculated with the specified Rth. This has something to do with the temperature at wich Rth is measured (mostly higher than 55°)
I also think that the chassis would be ok for an Aleph5, maybe you can even up the bias a bit wich would be good for your 4 ohm loudspeakers.
If you manage to set the bias to 2.6A as in the Aleph4 you will have just as much power into 4 ohms and dissipate a lot less heat.
William
somehow heatsinks always get hotter than calculated with the specified Rth. This has something to do with the temperature at wich Rth is measured (mostly higher than 55°)
I also think that the chassis would be ok for an Aleph5, maybe you can even up the bias a bit wich would be good for your 4 ohm loudspeakers.
If you manage to set the bias to 2.6A as in the Aleph4 you will have just as much power into 4 ohms and dissipate a lot less heat.
William
Hi,
there is a good heatsink paper in the Wakefield site.
The height of a heatsink changes the Rth s-a by about the square root of the height proportional change.
If you want to double the dissipation then you need to quadruple the height. If you have double the height you will get about 140% dissipation.
The Rth s-a is usually quoted at a particular delta T (difference between ambient and sink backplate temperature). Delta T is usually quite high, 50Cdeg to 70Cdeg. Much higher than we normally use. The Rth s-a can be corrected for a different Delta T.
The sink is tested with the whole back plate at the test temperature. An installation can only approach this theoretical dissipation IF the devices are spread around the backplate to try and achieve a nearly constant back plate temperature. A general rule I have found is that the radius from the device to the extreme corner of the backplate should be less than ten times the thickness of the backplate. A very wide or very tall heatsink will not meet this requirement. This is due to the far parts of the sink running cooler (sometimes much cooler) than the hot part under the device.
Orientation of the fins affects dissipation ability.
Colour affects dissipation.
restriction of air flow affects dissipation.
Hope this helps.
BTW. most heatsinks meet the manufacturers specification IF the corrections to Rth s-a are applied. Read Wakefield.
there is a good heatsink paper in the Wakefield site.
The height of a heatsink changes the Rth s-a by about the square root of the height proportional change.
If you want to double the dissipation then you need to quadruple the height. If you have double the height you will get about 140% dissipation.
The Rth s-a is usually quoted at a particular delta T (difference between ambient and sink backplate temperature). Delta T is usually quite high, 50Cdeg to 70Cdeg. Much higher than we normally use. The Rth s-a can be corrected for a different Delta T.
The sink is tested with the whole back plate at the test temperature. An installation can only approach this theoretical dissipation IF the devices are spread around the backplate to try and achieve a nearly constant back plate temperature. A general rule I have found is that the radius from the device to the extreme corner of the backplate should be less than ten times the thickness of the backplate. A very wide or very tall heatsink will not meet this requirement. This is due to the far parts of the sink running cooler (sometimes much cooler) than the hot part under the device.
Orientation of the fins affects dissipation ability.
Colour affects dissipation.
restriction of air flow affects dissipation.
Hope this helps.
BTW. most heatsinks meet the manufacturers specification IF the corrections to Rth s-a are applied. Read Wakefield.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.