Hi,
I was kindly posted this by Hugh Dean of AKSA, when I enquired about a couple of heatsinks
Have you considered fan cooling? On a Class A amp (Aleph by
any chance??) it is really is well nigh impossible to
achieve good cooling with still air, particularly in a high
ambient situation - yea verily, even the Brit summer.
Fan cooling with a large (120 x 120mm) fan rated for 2700rpm
or so and reduced to half voltage with a fluoro ballast is
VERY effective. If you bring a 32dB fan to half voltage,
the flow drops to about a third, yes, but the speed halves
and the noise drops to around 22dB; scarcely audible, even
in a quiet room, and MUCH quieter than a PC fan, which we
all have to live with constantly........ Most 120x120 fans
can easily generate 90cfm flow.
Fan cooling at 22cfm (12.7l/s) into a well designed heatsink
such as the flat backed MF18-151.5 will give an effective
thermal performance of 0.13C/watt. That should do it, as it
means for 100W input, only 13C above ambient, which should
guarantee MTBF on an IRF mosfet of around 50,000 hours!!
Two such heatsinks facing each other can have the semis on
opposite sides, will be 120mm x 120mm end on, and 151.5mm
long. For minimum noise, it is suggested you use a 4" long
120mm x 120mm plenum chamber between fan and heatsink; this
stops turbulent noise of fan blades against the fins.
Unfortunately I do not have photos, but from the description
you should get the picture. There are 18 fins on this sink.
Hope this is helpful, Happy New Year, and sorry about the
delayed response.
Cheers,
Hugh
Hugh R. Dean
Sales & Research
Aspen Amplifiers P/L
www.aksaonline.com
Problem is I'm having difficulty in visualisng how it's done like this, anyone know where there's any pictures?
Thanks
Raj
I was kindly posted this by Hugh Dean of AKSA, when I enquired about a couple of heatsinks
Have you considered fan cooling? On a Class A amp (Aleph by
any chance??) it is really is well nigh impossible to
achieve good cooling with still air, particularly in a high
ambient situation - yea verily, even the Brit summer.
Fan cooling with a large (120 x 120mm) fan rated for 2700rpm
or so and reduced to half voltage with a fluoro ballast is
VERY effective. If you bring a 32dB fan to half voltage,
the flow drops to about a third, yes, but the speed halves
and the noise drops to around 22dB; scarcely audible, even
in a quiet room, and MUCH quieter than a PC fan, which we
all have to live with constantly........ Most 120x120 fans
can easily generate 90cfm flow.
Fan cooling at 22cfm (12.7l/s) into a well designed heatsink
such as the flat backed MF18-151.5 will give an effective
thermal performance of 0.13C/watt. That should do it, as it
means for 100W input, only 13C above ambient, which should
guarantee MTBF on an IRF mosfet of around 50,000 hours!!
Two such heatsinks facing each other can have the semis on
opposite sides, will be 120mm x 120mm end on, and 151.5mm
long. For minimum noise, it is suggested you use a 4" long
120mm x 120mm plenum chamber between fan and heatsink; this
stops turbulent noise of fan blades against the fins.
Unfortunately I do not have photos, but from the description
you should get the picture. There are 18 fins on this sink.
Hope this is helpful, Happy New Year, and sorry about the
delayed response.
Cheers,
Hugh
Hugh R. Dean
Sales & Research
Aspen Amplifiers P/L
www.aksaonline.com
Problem is I'm having difficulty in visualisng how it's done like this, anyone know where there's any pictures?
Thanks
Raj
Raj,
This will give you an idea of what he is talking about. Picture two heatsinks that when faced fin to fin they create a cube similar to the picture...
The only other tidbit is that he recommends using a 4" plenum. The plenum in this case is nothing more than a 4" square spacer that's hollow between the fan and the heatsinks...
You then mount your semiconductors on the "ledges" that you see in the picture... In the case of the heatsinks he is referring to, I believe that there would be one mounting "ledge" on each side instead of two like in the picture...
If there is no ledge on the heatsinks you choose then you would simply distribute the devices down the flatback of the heatsink on each side...
Hope this helps!
This will give you an idea of what he is talking about. Picture two heatsinks that when faced fin to fin they create a cube similar to the picture...
The only other tidbit is that he recommends using a 4" plenum. The plenum in this case is nothing more than a 4" square spacer that's hollow between the fan and the heatsinks...
You then mount your semiconductors on the "ledges" that you see in the picture... In the case of the heatsinks he is referring to, I believe that there would be one mounting "ledge" on each side instead of two like in the picture...
If there is no ledge on the heatsinks you choose then you would simply distribute the devices down the flatback of the heatsink on each side...
Hope this helps!
Attachments
Fan cooling
Raj1
I constructed this case with fan cooling for my A75. The fan is a 24 volt run at 12volt, very quit and cooool. It is pretty much out of the A75 article, the heatsinks were surplus and this seemed the best way to provide for the heat dissipation. If I want to increase the bias I just have to turn up the fan. Plus I think it is a great looking toaster, or that is what most people call this amp.
Bryan
Raj1
I constructed this case with fan cooling for my A75. The fan is a 24 volt run at 12volt, very quit and cooool. It is pretty much out of the A75 article, the heatsinks were surplus and this seemed the best way to provide for the heat dissipation. If I want to increase the bias I just have to turn up the fan. Plus I think it is a great looking toaster, or that is what most people call this amp.
Bryan
Attachments
I never thought of using an AC fan but it should be quieter than a brushless DC one. The DC ones make a brrrr sound at low speeds because of torque pulsations whereas an AC one running off a sinewave should be much quieter. I have heard of using a series capacitor to reduce the voltage to an AC fan but the fluoro ballast method sounds good.Raj1 said:
hmm im not sure if i agree with this - yes it is a great way for cutting costs BUT it adds noise and interferance - the former is definatly bad for any sort of audio produce and the latter caused by the magnet) isn't goin to help much
if u are planning to use a fan, look for a better way to control the noise - such as this Pulse Width Modulation controler : http://www.bit-tech.net/article/51/
also bear in mind that it is possable to get very low c/w heatsinks for not much - i infact have two 0.2c/w heatsinks comming in the post for me that i got for free as samples from aavid
if u are planning to use a fan, look for a better way to control the noise - such as this Pulse Width Modulation controler : http://www.bit-tech.net/article/51/
also bear in mind that it is possable to get very low c/w heatsinks for not much - i infact have two 0.2c/w heatsinks comming in the post for me that i got for free as samples from aavid
further details etc
Hi,
I have no idea what a fluoro ballast is! Can anyone tell me, where can I get it/one from?
Her's the web address for the fan I have in mind.
http://www.tekheads.co.uk/tekheads/product?id=600801
SPeed selectable 12v fan which runs at 21db whilst churning out 25cfm.
Bryan what type of fan did you use? Was it dc 24v or ac?
Also what type of transformer did you use?
I was thinking of using a walwart dc regulated psu for the supply to the fan
The heatsink is 151.5mm long
120mm wide
and the fins are 60mm long
I think the base is 7mm thick, the standard dissipation of this heatsink without air cooling is 0.46 c/w
If you apply 13.5 cfm this drops to 0.13 c/w
At 27 cfm it drops to 0.11 c/w
however I am keen to do this as quietly as possible so if you think the fan I've chosen is no good please let me know!
Thanks
Raj
Hi,
I have no idea what a fluoro ballast is! Can anyone tell me, where can I get it/one from?
Her's the web address for the fan I have in mind.
http://www.tekheads.co.uk/tekheads/product?id=600801
SPeed selectable 12v fan which runs at 21db whilst churning out 25cfm.
Bryan what type of fan did you use? Was it dc 24v or ac?
Also what type of transformer did you use?
I was thinking of using a walwart dc regulated psu for the supply to the fan
The heatsink is 151.5mm long
120mm wide
and the fins are 60mm long
I think the base is 7mm thick, the standard dissipation of this heatsink without air cooling is 0.46 c/w
If you apply 13.5 cfm this drops to 0.13 c/w
At 27 cfm it drops to 0.11 c/w
however I am keen to do this as quietly as possible so if you think the fan I've chosen is no good please let me know!
Thanks
Raj
Fezz,
I tried out the pwm stuff (with an MIC502 IC) but wasn´t pleased cause the pwm signal turned up everywere despite using a seperate supply.
Raj1,
22dB(A) isn´t that silent inmy opinion. I use 12dB(A) fans at a lower voltage (i guess around 10dbA for the two) and only now I can say that I can´t hear them. At 15 dB(A) for the two of them at 12V you could hear them at over 3m distance.
william
I tried out the pwm stuff (with an MIC502 IC) but wasn´t pleased cause the pwm signal turned up everywere despite using a seperate supply.
Raj1,
22dB(A) isn´t that silent inmy opinion. I use 12dB(A) fans at a lower voltage (i guess around 10dbA for the two) and only now I can say that I can´t hear them. At 15 dB(A) for the two of them at 12V you could hear them at over 3m distance.
william
220 or 240VAC fans runs very quietly at 117VAC.
DC brushless fans don't all "cog". If the blade tends to snap into specific positions when you turn it with your fingers, it will cog. Some are real smooth. A lot of the cheap ones have noisy bearings that will make more noise than the air they blow when run at low speeds.
A series cap or 1 or 2 uF (or inductance supplied by a flourescent tube ballast) or dropping resistor work fine with AC fans, and you don't have to provide an extra power supply to drive them.
The cap can be left charged and if you open the amp to do some work, you can get zapped pretty good. If you put a cap in series with a fan, be sure to put a 20-100k resistor across it to discharge it when you turn off the power.
MR
DC brushless fans don't all "cog". If the blade tends to snap into specific positions when you turn it with your fingers, it will cog. Some are real smooth. A lot of the cheap ones have noisy bearings that will make more noise than the air they blow when run at low speeds.
A series cap or 1 or 2 uF (or inductance supplied by a flourescent tube ballast) or dropping resistor work fine with AC fans, and you don't have to provide an extra power supply to drive them.
The cap can be left charged and if you open the amp to do some work, you can get zapped pretty good. If you put a cap in series with a fan, be sure to put a 20-100k resistor across it to discharge it when you turn off the power.
MR
further details etc
Raj1,
24v DC FBK Panasonic. I used a pass transistor off of the regulated DC power supply with a variable pot at the base to adjust the voltage to the fan. Its been 7 years since I finished the amp and the fan is as quite as when it was new.
I would agree with other posts about the potential for electrical noise to be introduced into the front end, but with proper layout it should'nt be a problem.
Bryan
Raj1,
24v DC FBK Panasonic. I used a pass transistor off of the regulated DC power supply with a variable pot at the base to adjust the voltage to the fan. Its been 7 years since I finished the amp and the fan is as quite as when it was new.
I would agree with other posts about the potential for electrical noise to be introduced into the front end, but with proper layout it should'nt be a problem.
Bryan
MRehorst said:
Further experiments with one particular fan indicate that maybe a series cap isn't the way to go. When the fan was vertical, a 2.2 uF cap had it running nice and slow and quiet. As soon as I turned the fan horizontal, it slowly accelerated and eventually reached (after about 30 sec) its normal speed and noise level.
1 uF alone doesn't allow the fan to start turning.
If the thing is going to be so tweeky that the cap has to be somewhere between 1 and 2 uF, it probably isn't a very good way to control the speed. What will happen at higher or lower temperatures, or one fan to the next? If the fan fails to start turning when power is applied, you can quickly have a disaster.
Back to the resistors...
MR
Lowering the voltage to AC fans doesn't reliably decrease
the rotational speed, and so isn't that great at lowering
the noise.
Myself, I start with a big quality DC fan with ball bearings,
and I run it at variable DC voltage. Not all such fans will
be quiet, so you have to test them before you buy a bunch.
The big blades move air with less noise and at slower
speeds, and once you get them turning you can lower
the voltage way down for slow rpms.
Marlin P Jones had some terrific Nidec fans for $3 last month,
and I bought 100. They probably have plenty left, and you
can find them online at www.MPJA.com
Whoops, I just checked, and the price went up to $8
the rotational speed, and so isn't that great at lowering
the noise.
Myself, I start with a big quality DC fan with ball bearings,
and I run it at variable DC voltage. Not all such fans will
be quiet, so you have to test them before you buy a bunch.
The big blades move air with less noise and at slower
speeds, and once you get them turning you can lower
the voltage way down for slow rpms.
Marlin P Jones had some terrific Nidec fans for $3 last month,
and I bought 100. They probably have plenty left, and you
can find them online at www.MPJA.com
Whoops, I just checked, and the price went up to $8
At present I try not to use fans, but I used to take the fans out of mainframe and minicomputers (5" Boxer fans?) and run the 240V units on 120V, or series two 120V ones. Worked well, and since I got the fans for free by stripping old gear, the price was right. Lots of people probably have a few in their junk boxes if they dig deep enough.
More modern fans are going to be quieter, but it's a nice, cheap alternative.
Grey
More modern fans are going to be quieter, but it's a nice, cheap alternative.
Grey
Variable Inductor....
"Fan cooling with a large (120 x 120mm) fan rated for 2700rpm
or so and reduced to half voltage with a fluoro ballast is
VERY effective."
Also usefull is a speed control box for overhead ceiling fans.
These controllers are a multitapped series inductor and a rotary switch, and can be found for virtually free at garage sales or demolition yards.
I have three 240V fans and controller mounted together as solder fume extraction setup.
Eric.
"Fan cooling with a large (120 x 120mm) fan rated for 2700rpm
or so and reduced to half voltage with a fluoro ballast is
VERY effective."
Also usefull is a speed control box for overhead ceiling fans.
These controllers are a multitapped series inductor and a rotary switch, and can be found for virtually free at garage sales or demolition yards.
I have three 240V fans and controller mounted together as solder fume extraction setup.
Eric.
Actually, after having written the above, I remembered that I'm on the verge of having to put fans on an Aleph-X, 'cause a buddy of mine wants to hear one in his system and you can bet your sweet bippy that water-cooled stuff ain't portable.
Now all I've got to do is dig out my box of fans. Problem is, it's buried under a mound of other things.
<i>Sigh</i>
Grey
Now all I've got to do is dig out my box of fans. Problem is, it's buried under a mound of other things.
<i>Sigh</i>
Grey
cfm
Hi,
I've looked at a number of ac fans and have found that, anything that has a db rating of around 35db churnes out about 70cfm or so. If you go for round a 9ocfm fan it runs at about 40db.
Any ideas what cfm I would get out of a fan that provides 70cfm at 230v if the volatge was dropped to half, both in terms of db and cfm?
thanks
Raj
Hi,
I've looked at a number of ac fans and have found that, anything that has a db rating of around 35db churnes out about 70cfm or so. If you go for round a 9ocfm fan it runs at about 40db.
Any ideas what cfm I would get out of a fan that provides 70cfm at 230v if the volatge was dropped to half, both in terms of db and cfm?
thanks
Raj
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.