Hi folks,
Has anyone had trouble using fanned heatsinks to cool power amplifiers?
In the process of designing a class D amplifier using the TPA3221 IC.
This will be set to 18dB or 24dB gain dependant on SE or balanced input.
The power output can be up to 40W for dual channel into 3ohms. Heat dissipation of 15W max for my application.
Evaluation board (EVB) uses the passive heatsink ATS-TI10P-563-C1-R0. This would be sufficient for my design but it costs money and I have a load of fanned heatsinks. They are sufficient for the heat dissipation and similar to this one. The distance from fan to IC would be around 4mm.
I'm concerned about electrical noise being coupled into the amplifier. At audio frequency, magnetic coupling is mostly likely, especially being so close.
A 10mV noise voltage would be amplified to around 630mV for 18dB and 2.5V for 24dB. Into 4 ohms, 10mW only for 18dB but a massive 1.6W with 24dB gain.
Am I overthinking this or have you fallen over the same issue previously? I couldn't find information on this by searching the forums.
Thanks
Has anyone had trouble using fanned heatsinks to cool power amplifiers?
In the process of designing a class D amplifier using the TPA3221 IC.
This will be set to 18dB or 24dB gain dependant on SE or balanced input.
The power output can be up to 40W for dual channel into 3ohms. Heat dissipation of 15W max for my application.
Evaluation board (EVB) uses the passive heatsink ATS-TI10P-563-C1-R0. This would be sufficient for my design but it costs money and I have a load of fanned heatsinks. They are sufficient for the heat dissipation and similar to this one. The distance from fan to IC would be around 4mm.
I'm concerned about electrical noise being coupled into the amplifier. At audio frequency, magnetic coupling is mostly likely, especially being so close.
A 10mV noise voltage would be amplified to around 630mV for 18dB and 2.5V for 24dB. Into 4 ohms, 10mW only for 18dB but a massive 1.6W with 24dB gain.
Am I overthinking this or have you fallen over the same issue previously? I couldn't find information on this by searching the forums.
Thanks
Last edited:
The main problem with fan cooling is the acoustical noise as fans placed at the output are in the least sensitive place in a power amplifier .
I am looking at an amplifier build with -100db SNR/ 0.003 % dist.@ 1watt/IMD 0.006/20Hz-over 20kHz and two ultra low noise Noctua fans running at =9.5db(A) combined.
Personally I don't like fans but each to his own.
I am looking at an amplifier build with -100db SNR/ 0.003 % dist.@ 1watt/IMD 0.006/20Hz-over 20kHz and two ultra low noise Noctua fans running at =9.5db(A) combined.
Personally I don't like fans but each to his own.
"When fans are used you just know what will break down". Also the amplifier will be dirty quite fast and, as the heatsink is underdimensioned, the device will possibly overheat then. Passive cooling is more suitable in audio. No noise, less dirt, no maintenance. Heatsinks also belong to audio-porn 
The ATS-TI10P-563-C1-R0 costs around 6 Euro....
The ATS-TI10P-563-C1-R0 costs around 6 Euro....
Last edited:
Use the right fan at the right place, for which it was developed
Fans are not anytime usable for all devices. depends on how these are made, and what concept besides cooling is behind.
The Fan you have in that picture is a CPU Fan, or Graphic Card Fan of a computer.
Fans like that are LOUD, up to 20 and more DB.
You do not need to be afraid of Electrical Noise as long you do not share the power source of the fan with the power source of the Chip you use.
I use Active cooling, because the heat sink I would need to cool down my CLASS A Build would be more than 10KG and very large, not to talk about the Price,.
I use 4 Low Noise Noctua Fans in my Class A, running only that fast that the amplifier is cooled down to max 70 degrees Celsius.
Fans are running on 9 volts only and give excellent results, concerning the FAN NOISE from Rotation. I do not say that if you really listening to catch that noise that there isn't any FAN NOISE.
And there is nowhere a rule which forbids to run FANS with UNDER-VOLTAGES. and this on a low noise fan will cut the NOISE more than many expect.
I can go down with the SPL of the Speaker to about 500mwatts so that is really not loud and it can be real quiet in the room, no one will hear the noise of the fans.
About being anxious that the fans break down,
Class A amps get always hot, so a high quality Thermal switch on each channel, which will switch of the system if temp reaches more than 100Degrees Celsius is just as normal as that you need, a fork a spoon and and knife to eat your meals.
I talking about 125 - 200 Watts Heat dissipation / Channel. Present time it's about 136, the new build will be 182 Watts / Channel HEAT
So there is not much to be afraid of that if the Fan breaks, the amp will break as well., I do not have worries. I agree with everyone who doesn't like fans in there amplifiers, but sometimes there ain't no other way or chance to do so..
But in your case I would go with just a little bit larger Heat sink. On the internet there are many website to calculate the size for Heat Sinks without Fans,
That's my 5 cents to this matter.
Regards
Chris Hess
Fans are not anytime usable for all devices. depends on how these are made, and what concept besides cooling is behind.
The Fan you have in that picture is a CPU Fan, or Graphic Card Fan of a computer.
Fans like that are LOUD, up to 20 and more DB.
You do not need to be afraid of Electrical Noise as long you do not share the power source of the fan with the power source of the Chip you use.
I use Active cooling, because the heat sink I would need to cool down my CLASS A Build would be more than 10KG and very large, not to talk about the Price,.
I use 4 Low Noise Noctua Fans in my Class A, running only that fast that the amplifier is cooled down to max 70 degrees Celsius.
Fans are running on 9 volts only and give excellent results, concerning the FAN NOISE from Rotation. I do not say that if you really listening to catch that noise that there isn't any FAN NOISE.
And there is nowhere a rule which forbids to run FANS with UNDER-VOLTAGES. and this on a low noise fan will cut the NOISE more than many expect.
I can go down with the SPL of the Speaker to about 500mwatts so that is really not loud and it can be real quiet in the room, no one will hear the noise of the fans.
About being anxious that the fans break down,
Class A amps get always hot, so a high quality Thermal switch on each channel, which will switch of the system if temp reaches more than 100Degrees Celsius is just as normal as that you need, a fork a spoon and and knife to eat your meals.
I talking about 125 - 200 Watts Heat dissipation / Channel. Present time it's about 136, the new build will be 182 Watts / Channel HEAT
So there is not much to be afraid of that if the Fan breaks, the amp will break as well., I do not have worries. I agree with everyone who doesn't like fans in there amplifiers, but sometimes there ain't no other way or chance to do so..
But in your case I would go with just a little bit larger Heat sink. On the internet there are many website to calculate the size for Heat Sinks without Fans,
That's my 5 cents to this matter.
Regards
Chris Hess
Last edited:
Passive is good, but sometimes Price is very high
Dirt and needs of filters are in need with a not too clean environment.
all my Class A using Fans, mounting them in a manner that these pull off the heat and not blowing on the heat sink directly, not placing them in the BOX if possible then NO FILTERS ARE NEEDED.
Then also just let them RUN AS MUCH they need to cool down the system.
As larger the fan is in diameter the more you can go down with the speed, the less you will see dirt.
I calculated the HEAT SINK for the CLASS A I'm going to build, with a max heat of 80 Degrees after 2 Minutes which will not exceed 100Degrees Celsius after 2 hours run time, My amps running daily at least 4 - 8 hours, right now it's already 7Pm to now, almost 2 am 7 hours without stop. Heat 70degrees.
Present time the ACTIVE HEAT SINK is 9.5cm x 17.5 cm x 10 cm one Fan/side and a center fan to feed "fresh air" from outside the box, but blowing into a AIR CHANNEL not into the BOX itself, at the side rotating at 500rpm and then blowing outwards through the Active Heat sinks, not into the box.
The Passive Heat Sink would have following measure
400mm length 200mm height and 120mm thick with a pick up copper plate with a thickness of 10 mm. this plate alone weights 5 kilogramm.!
And Thermal switch still is needed for each side.
One Side of this Heat Sink cost more than the complete amp. is about 500USD inclusive milling and prepairing it with all it needs.
I'm talking about 1 SIDE, check out my threads about the Class A I build then you know what I mean. But still I'm going to build that PASSIVE AMP,. just to find out how long it last, till a Output Transistor blows up because of overheating.
Regards
Chris Hess
Dirt and needs of filters are in need with a not too clean environment.
all my Class A using Fans, mounting them in a manner that these pull off the heat and not blowing on the heat sink directly, not placing them in the BOX if possible then NO FILTERS ARE NEEDED.
Then also just let them RUN AS MUCH they need to cool down the system.
As larger the fan is in diameter the more you can go down with the speed, the less you will see dirt.
I calculated the HEAT SINK for the CLASS A I'm going to build, with a max heat of 80 Degrees after 2 Minutes which will not exceed 100Degrees Celsius after 2 hours run time, My amps running daily at least 4 - 8 hours, right now it's already 7Pm to now, almost 2 am 7 hours without stop. Heat 70degrees.
Present time the ACTIVE HEAT SINK is 9.5cm x 17.5 cm x 10 cm one Fan/side and a center fan to feed "fresh air" from outside the box, but blowing into a AIR CHANNEL not into the BOX itself, at the side rotating at 500rpm and then blowing outwards through the Active Heat sinks, not into the box.
The Passive Heat Sink would have following measure
400mm length 200mm height and 120mm thick with a pick up copper plate with a thickness of 10 mm. this plate alone weights 5 kilogramm.!
And Thermal switch still is needed for each side.
One Side of this Heat Sink cost more than the complete amp. is about 500USD inclusive milling and prepairing it with all it needs.
I'm talking about 1 SIDE, check out my threads about the Class A I build then you know what I mean. But still I'm going to build that PASSIVE AMP,. just to find out how long it last, till a Output Transistor blows up because of overheating.
Regards
Chris Hess
All this would of course have been unnecessary if one builds a less powerfull amplifier (let alone a class A device). Like the OP in this thread does, his efficient class D amplifier just needs a tiny heatsink to cope with the maximum so worst case 15W power dissipation for 2 channels. The whole device will be around the size of the transformer of a class A device and weigh less
Now the funny part of this is when you 2 would meet and compare the amplifiers only to discover they are almost indistinguishable from eachother
Now the funny part of this is when you 2 would meet and compare the amplifiers only to discover they are almost indistinguishable from eachother
Last edited:
This is one of the best Reply I ever heard. Thank you.
I have to acknowledge that only heat is big in my Amp.. this CLASS A has only MAX 26 - 28 watts into 8 Ohm.
I have heard Class D, have you heard a good CLASS A, with this I do not say that mine is good. But Class A is from Purity the best. same as compairing Vinyl with a CD. OR Vinyl with a MP 3 and so on.. Makes no sense..
I'm happy with mine and he is happy with his, until he listen to a class a for a while.. then no more..
Cheers
Check out my threath about class a on Sunday.. or next week, depends on my work.. Prototype Version 9.6.10.21 is going to be Boxed. The only bad thing is you can't listen to it. But if you pass Swiss (see you are in Germany) your welcome anytime to listen to Class A. I'm living in Capital City of Switzerland Bern. You can find me most time in my Thai Restaurant - Gourmet Thai Restaurant & Take Away Bern Switzerland, this will lead you if you use Google.
Regards
Chris Hess
I have to acknowledge that only heat is big in my Amp.. this CLASS A has only MAX 26 - 28 watts into 8 Ohm.
I have heard Class D, have you heard a good CLASS A, with this I do not say that mine is good. But Class A is from Purity the best. same as compairing Vinyl with a CD. OR Vinyl with a MP 3 and so on.. Makes no sense..
I'm happy with mine and he is happy with his, until he listen to a class a for a while.. then no more..
Cheers
Check out my threath about class a on Sunday.. or next week, depends on my work.. Prototype Version 9.6.10.21 is going to be Boxed. The only bad thing is you can't listen to it. But if you pass Swiss (see you are in Germany) your welcome anytime to listen to Class A. I'm living in Capital City of Switzerland Bern. You can find me most time in my Thai Restaurant - Gourmet Thai Restaurant & Take Away Bern Switzerland, this will lead you if you use Google.
Regards
Chris Hess
Granted there may not be conducted noise, I can mitigate against this fact by using a filter on the power supply rails, and using higher CMRR parts on the rails.
The issue I'm questioning here is radiated noise via the means of magnetic coupling.
Fans have a place in electronics, even for audio electronics. It seems everyone in audio is far too concerned with dust and audible noise than anything else.
Coming from a professional background designing digital electronics, my first thought is radiated and conducted electrical noise especially for analogue electronics where amplification is involved.
Any high power amp professionally made for PAs will use fans, there is just a little maintainance required to ensure that the dust doesn't encumber the performance of the cooling solution.
I bought in a Marshall 100 watt guitar amp.
The fan was squealing. Probably caught it just before it failed.
So bought in another one and fitted it.
Dead quiet now.
For some reason the fan has an input vent but no output vent !
Not sure how much good the fan is doing as the case well sealed except for rear vent where fan is.
The fan was squealing. Probably caught it just before it failed.
So bought in another one and fitted it.
Dead quiet now.
For some reason the fan has an input vent but no output vent !
Not sure how much good the fan is doing as the case well sealed except for rear vent where fan is.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.