Jacco,
Why use mica when you could cut your thermal transfer loss in half by using the Thermalloy ceramic-alum. oxide insulators? If anyone here needs any the local surplus shop stocks TO-247 size... they have a couple thousand of them at .76 each, probably cheaper in quantities...... In fact if anayone need any just let me know or I can put you in direct touch with them. These are alaso made in TO-220 and TO-3.
Mark
Why use mica when you could cut your thermal transfer loss in half by using the Thermalloy ceramic-alum. oxide insulators? If anyone here needs any the local surplus shop stocks TO-247 size... they have a couple thousand of them at .76 each, probably cheaper in quantities...... In fact if anayone need any just let me know or I can put you in direct touch with them. These are alaso made in TO-220 and TO-3.
Mark
Mark,
Jacco wants MT-200 insulators.
Mouser has them 0.40 each and price decrease w/quantity, item #: 532-4170,
http://www.mouser.com/catalog/623/1374.pdf
Jacco,
I can send you about 20 pieces of Mt-200 mica.
Jacco wants MT-200 insulators.
Mouser has them 0.40 each and price decrease w/quantity, item #: 532-4170,
http://www.mouser.com/catalog/623/1374.pdf
Jacco,
I can send you about 20 pieces of Mt-200 mica.
Thanks a lot guys. They are nice, indeed
Khan-Man,
retain them for a next project, you've invested time and LCDs in those insulators.
I've got spare output boards, if 12 per channel can't cope with the heat i need to step up to 18 or 24 devices.
I'm considering using Assmann's E20 insulator sheet that can be punched in the required size, i've done that with mica sheet for MT200s in the past with a dedicated punch mold.
Assmann electronics (Germans are fun) produces ceramic filled insulators that have 1/3d of the thermal resistance of silpads,
4.5 W/m *K instead of 1.5 W/m *K.
Aluminum Oxide (Al2O3) does 25 W/m *K, but Assmann's ceramic filled 0.2 mm thick E20 insulator should outperform an Aavid Thermalloy's 1.5mm thick aluminum oxide insulator by nearly 0.1 C/W.
TO3 :
Aavid Thermalloy Al2O3 = 0.30 C/W
Assmann ceramic E20 = 0.22 C/W
www.Assmann.com
Thermal management catalogue, Chapter 8
Digikey carries Assmann products, could be they can deliver E20 on request for the US, E20 price tag in the US will probably also scare the pants off you.
$0.76 for a thin TO246 Aluminium Oxide insulator is darn right cheap. I can get them overhere, but at $3.30 each, $2.77 for a minimum of 10.
retain them for a next project, you've invested time and LCDs in those insulators.
I've got spare output boards, if 12 per channel can't cope with the heat i need to step up to 18 or 24 devices.
I'm considering using Assmann's E20 insulator sheet that can be punched in the required size, i've done that with mica sheet for MT200s in the past with a dedicated punch mold.
Assmann electronics (Germans are fun) produces ceramic filled insulators that have 1/3d of the thermal resistance of silpads,
4.5 W/m *K instead of 1.5 W/m *K.
Aluminum Oxide (Al2O3) does 25 W/m *K, but Assmann's ceramic filled 0.2 mm thick E20 insulator should outperform an Aavid Thermalloy's 1.5mm thick aluminum oxide insulator by nearly 0.1 C/W.
TO3 :
Aavid Thermalloy Al2O3 = 0.30 C/W
Assmann ceramic E20 = 0.22 C/W
www.Assmann.com
Thermal management catalogue, Chapter 8
Digikey carries Assmann products, could be they can deliver E20 on request for the US, E20 price tag in the US will probably also scare the pants off you.
$0.76 for a thin TO246 Aluminium Oxide insulator is darn right cheap. I can get them overhere, but at $3.30 each, $2.77 for a minimum of 10.
Parts GB'ers-
Well a quick update.
We are STILL awaiting the Welborne labs order completion.. They sent a "partial order" out 2 weeks late and no word on when the remainder will be shipped.
I originally rx’ed payment for 26 channels. I added 4 for myself and ordered 30.
I completely forgot the 6 channels for Al, Mark and Pavel as thanks for the designing the new boards. Also a friend needs 4 channels so I will be making a second order as soon as this one is completed.
Well a quick update.
We are STILL awaiting the Welborne labs order completion.. They sent a "partial order" out 2 weeks late and no word on when the remainder will be shipped.
I originally rx’ed payment for 26 channels. I added 4 for myself and ordered 30.
I completely forgot the 6 channels for Al, Mark and Pavel as thanks for the designing the new boards. Also a friend needs 4 channels so I will be making a second order as soon as this one is completed.
Howdy Mr. G,
Sure would like to have some TO3 versions of them for the KSA me and the grandson are working on.
BTW, any good folks have a line on the heatsinks like was used in the original? My boy seems to think that a flat back heatsink that's big enough without a fan would be another choice, but I dunno. Could be the money talking (his) for this choice. Or maybe he just wants to be different than the ole man.
Regards,
Elvin
Sure would like to have some TO3 versions of them for the KSA me and the grandson are working on.
BTW, any good folks have a line on the heatsinks like was used in the original? My boy seems to think that a flat back heatsink that's big enough without a fan would be another choice, but I dunno. Could be the money talking (his) for this choice. Or maybe he just wants to be different than the ole man.
Regards,
Elvin
For convection cooling, I recently found a TO-3 sink here:
http://www.surplussales.com/Heatsinks/HeatSink2.html
(labeled 750W) which would work with 3 sinks per channel, if not a bit close to the acceptable operating margins. At $25/sink that is a sizable investment. This sink appears to be similar to the following thermaloy extrusion:
http://www.aavidthermalloy.com/cgi-...th=6.75&LengthUnits=in&SButton1=Change+Length
I modified the length on the web page to show that in 6.75 length sections it would be about .22 degrees C/W per channel in sets of three.
It appears to me that Krell used this one on the KSA-80:
http://www.aavidthermalloy.com/cgi-...h=8.000&LengthUnits=in&SButton1=Change+Length
Using 8 inch lengths, two per channel, is just right with 3 devices on each. I might use these if I could find them for less than the surplus sinks above, but it is easy to see why forced convection is so attractive, as it is clearly a lot less investment for a fan than more metal to create the same effective surface area. Still, there are advantages to each: convection is silent, except for the thermal expansion and contraction noises, and looks really nasty. No-one will doubt that your amp mean business with those giant sinks hanging on it. Forced convection seems to offer better thermal stability, smaller size and, of course adjustability, but requires at least a minimal fan circuit. I'm not sure which way I am going yet. Kind of depends on how much I can justify spending on vanity, but thats what mid life is all about. ;-)
http://www.surplussales.com/Heatsinks/HeatSink2.html
(labeled 750W) which would work with 3 sinks per channel, if not a bit close to the acceptable operating margins. At $25/sink that is a sizable investment. This sink appears to be similar to the following thermaloy extrusion:
http://www.aavidthermalloy.com/cgi-...th=6.75&LengthUnits=in&SButton1=Change+Length
I modified the length on the web page to show that in 6.75 length sections it would be about .22 degrees C/W per channel in sets of three.
It appears to me that Krell used this one on the KSA-80:
http://www.aavidthermalloy.com/cgi-...h=8.000&LengthUnits=in&SButton1=Change+Length
Using 8 inch lengths, two per channel, is just right with 3 devices on each. I might use these if I could find them for less than the surplus sinks above, but it is easy to see why forced convection is so attractive, as it is clearly a lot less investment for a fan than more metal to create the same effective surface area. Still, there are advantages to each: convection is silent, except for the thermal expansion and contraction noises, and looks really nasty. No-one will doubt that your amp mean business with those giant sinks hanging on it. Forced convection seems to offer better thermal stability, smaller size and, of course adjustability, but requires at least a minimal fan circuit. I'm not sure which way I am going yet. Kind of depends on how much I can justify spending on vanity, but thats what mid life is all about. ;-)
Hi,
I also thought of going with passive cooling, and the cheapest heatsinks I came up with, that seem to be able to dissipate the heat, are 6 pieces of 6" length of these heatsinks:
http://cgi.ebay.com/HEATSINK-ALUMINUM-EXTRUSION-24-WILL-CUT-FREE-RAW_W0QQitemZ7551992595QQcategoryZ31489QQrdZ1QQcmdZViewItem
which have already been mentioned in this thread a few pages back.
From Rod Elliott's calculator, and what it says in the auction, it seems that 6 x 6" can dissipate enough heat for one stereo amp (at 3.53A total bias), or a monoblock (at 5.29A-6.18A total bias).
What do you guys think? Will it suffice? Or am I missing something?
Thanks.
Paul
I also thought of going with passive cooling, and the cheapest heatsinks I came up with, that seem to be able to dissipate the heat, are 6 pieces of 6" length of these heatsinks:
http://cgi.ebay.com/HEATSINK-ALUMINUM-EXTRUSION-24-WILL-CUT-FREE-RAW_W0QQitemZ7551992595QQcategoryZ31489QQrdZ1QQcmdZViewItem
which have already been mentioned in this thread a few pages back.
From Rod Elliott's calculator, and what it says in the auction, it seems that 6 x 6" can dissipate enough heat for one stereo amp (at 3.53A total bias), or a monoblock (at 5.29A-6.18A total bias).
What do you guys think? Will it suffice? Or am I missing something?
Thanks.
Paul
the numbers on the auction are hinky. That extrusion looks a lot like:
http://www.aavidthermalloy.com/cgi-...h=8.000&LengthUnits=in&SButton1=Change+Length
However, the dimensions are slightly different. So, by my estimation to compensate for 3 less fins worth of surface area, you can add 2 lineal inches to the above thermalloy sink to get a fairly accurate estimate. According to the calculator, comparing the C/W figure for 6" from the auction sink to 8 inches of the Thermalloy extrusion shows a pretty significant difference: 0.75 vs. 0.93 degrees C/W. At 12" vs. 14" the difference is even more stark at a claimed 0.375 degrees C/W vs. 0.70 for the thermalloy. Draw your own conclusions. By my calculation you would need three 18" sections of the auction sink, per side, in order to attain a good result, but you might get away with three 12 inch sections per side if there is air moving in your room.
Ebay is the worlds largest market for caveat emptor.
http://www.aavidthermalloy.com/cgi-...h=8.000&LengthUnits=in&SButton1=Change+Length
However, the dimensions are slightly different. So, by my estimation to compensate for 3 less fins worth of surface area, you can add 2 lineal inches to the above thermalloy sink to get a fairly accurate estimate. According to the calculator, comparing the C/W figure for 6" from the auction sink to 8 inches of the Thermalloy extrusion shows a pretty significant difference: 0.75 vs. 0.93 degrees C/W. At 12" vs. 14" the difference is even more stark at a claimed 0.375 degrees C/W vs. 0.70 for the thermalloy. Draw your own conclusions. By my calculation you would need three 18" sections of the auction sink, per side, in order to attain a good result, but you might get away with three 12 inch sections per side if there is air moving in your room.
Ebay is the worlds largest market for caveat emptor.
When I plug the ebay sink into Mr. Elliott's calculator, I found you can halve the c/w resistance if you opt for anodizing. The minimum emissivity is ten fold when anodized. Although it is not advertised, I would guess that is where the seller came to the .375c/w figure for a 12" section.
I would prefer the fins to be oriented the other way, so that the end result isn't a foot tall and relativley narrow amp. I have been watching ebay for a couple of months now, trying to find the perfect heatsink, but nothing seems to be an extraordinary deal. I somewhat regret missing the seifert GB a while back.
~Brad
I would prefer the fins to be oriented the other way, so that the end result isn't a foot tall and relativley narrow amp. I have been watching ebay for a couple of months now, trying to find the perfect heatsink, but nothing seems to be an extraordinary deal. I somewhat regret missing the seifert GB a while back.
~Brad
Hi,
I read that orienting the sink with fins horizontal reduces the dissipation by about 20%. I think very long sinks will derate more than -20%.
For a passively cooled ClassA that would be disastrous.
An earlier reference to 0.2C/W is a bit low for a passively cooled KSA50 (145W * 0.2 = 29Cdeg) add in ambient and Rthc-s and your Tc could be 65degC, some run as hot as this and a few even hotter. 0.16C/W would be a better target.
If you use a 30Vac transformer then 0.14C/W would probably be needed.
I read that orienting the sink with fins horizontal reduces the dissipation by about 20%. I think very long sinks will derate more than -20%.
For a passively cooled ClassA that would be disastrous.
An earlier reference to 0.2C/W is a bit low for a passively cooled KSA50 (145W * 0.2 = 29Cdeg) add in ambient and Rthc-s and your Tc could be 65degC, some run as hot as this and a few even hotter. 0.16C/W would be a better target.
If you use a 30Vac transformer then 0.14C/W would probably be needed.
niles said:
Niles, can you explain that to me ?
One thing not to be underestimated is the dust accumulation of fan cooled amplifiers.
Airflow induced by the fan transports dust to the amplifier, statics makes it stick.
Dust reduces the capability of components to get rid of heat dramatically. The average temperature of devices will increase and lead to failure, the number 1 cause of electronics decease.
A fan cooled amplifier will have to be cleaned much more frequently.
Tiresome, unless someone has the guts of JanDiddenMan and places the amplifier in the dish washer.
Problem with class A amplifiers is that the regular way to cool other components is to add ventilation holes in the chassis to accomodate airflow within the amplifier.
Sealing the chassis will keep the dust out, but requires an alternate way of cooling amplifier components.
Using vertical natural convection to cool main components and horizontal forced cooling for the heatsinks seems to be a way of reducing dust growth.
People with allergies and plenty space better opt for fan-less.
(i completed a 4 year desensibilisation treatment, employed a number of electronic air ionisers in the house for years.)
jacco:
I'd love to but I haven't had thermal science in more than a decade, and I wasn't very good at it then. ;-)
It seems to me, albeit without any equations to back it up, that the thermal mass of the sink can be plotted as a function air velocity over the sink assuming turbulent flow. Yes? I would think that at a static flow rate, that the thermal conductivity is more consistent with forced convection than with natural convection, which would all other things the same, damp fluctuations in temperature swing, therby allowing for greaterthermal stability. I now defer to the professor, who will likely set me straight.
I'd love to but I haven't had thermal science in more than a decade, and I wasn't very good at it then. ;-)
It seems to me, albeit without any equations to back it up, that the thermal mass of the sink can be plotted as a function air velocity over the sink assuming turbulent flow. Yes? I would think that at a static flow rate, that the thermal conductivity is more consistent with forced convection than with natural convection, which would all other things the same, damp fluctuations in temperature swing, therby allowing for greaterthermal stability. I now defer to the professor, who will likely set me straight.
temps
If you look at the temperature chart in Post 4375, the red and yellow lines are the temp of large fan cooled heatsinks, while green and blue are the temperature of smaller convection heat sinks totally isolated from any fan induced air movement. The temperature stability over time seems to be about the same to this uneducated eye.
If you look at the temperature chart in Post 4375, the red and yellow lines are the temp of large fan cooled heatsinks, while green and blue are the temperature of smaller convection heat sinks totally isolated from any fan induced air movement. The temperature stability over time seems to be about the same to this uneducated eye.