Two questions....
1. Is it unwise to place filter caps close to a large heatsink?
2. Does placing heatsink horizontal ( fins pointing downwards ) a bad idea? Will it affect the performance of the heat sink?
FYI : Transistors that will be mounted on the heatsink are 2SC2922. The amp is a clone of Naim NAP 200
1. Is it unwise to place filter caps close to a large heatsink?
2. Does placing heatsink horizontal ( fins pointing downwards ) a bad idea? Will it affect the performance of the heat sink?
FYI : Transistors that will be mounted on the heatsink are 2SC2922. The amp is a clone of Naim NAP 200
For optimal cooling, the fins have to be placed such that hot air can easily go up and cold air can easily replace it. Horizontal with fins downwards sounds like the worst possible placement for natural convection cooling.
Not good idea to heat up the capacitors. It will reduce their life...
Bad idea. It's doubling the thermal resistance, which halving the maximum possible power to dissipate.
Sajti
It also depends how hot the heatsink is going to get. The NAPs should have very low bias current and run cool normally. Hard to comment without seeing the heatsink and enclosure. Obviously, for good dissipation the heatsink fins need free air convection current.
The biasing current is around 30ma so I guess the transistors shouldn't run too hot.
Here is the heatsink that I plan to use - DAC - Heat Sink Black Anodized - 196-50
Here is the heatsink that I plan to use - DAC - Heat Sink Black Anodized - 196-50
The heatsink shown is simply too small to permit full power in class AB. As already suggested, this may be OK for this low-bias class AB type amplifier, as long as you don't use the amplifier at medium to high power. It's self-defeating to build the amplifier in this way but if you have made a mistake in your choice of model, there will have to be compromises somewhere and probably in several important areas.
You may realise that NAP200 is rated for 2 x 100W/4R but you won't get near that safely. It also needs a high quality power transformer of at least 400VA rating. The original amplifier uses the whole case as the heatsink which is why the output transistors can be mounted horizontally. To use a heatsink instead, both the output transistors and the heatsink fins will need to be vertical, as shown in the first image of your product link.
If you must use a heatsink, it will still need to be much larger than your example and the case will likely need to be taller (perhaps >100mm high) to accommodate it. Then, there is still the problem of full and easy air flow from beneath the amplifier though large holes, up through all the fins then out to free air again. The slot in the PCB may be enough to allow air flow with a matching cross-section size heatsink but the case must also be slotted or meshed above and below.
Overall, you would be wasting money by building this expensive 2 x 100W/channel amplifier but only fitting a (approximate) 2 x 30W heatsink. Building it as original, in a thick aluminium case, doesn't require any heatsink and should be cheaper, assuming you were going to follow the original design correctly and add a heat spreader bar between the transistors and the bottom of the case.
You may realise that NAP200 is rated for 2 x 100W/4R but you won't get near that safely. It also needs a high quality power transformer of at least 400VA rating. The original amplifier uses the whole case as the heatsink which is why the output transistors can be mounted horizontally. To use a heatsink instead, both the output transistors and the heatsink fins will need to be vertical, as shown in the first image of your product link.
If you must use a heatsink, it will still need to be much larger than your example and the case will likely need to be taller (perhaps >100mm high) to accommodate it. Then, there is still the problem of full and easy air flow from beneath the amplifier though large holes, up through all the fins then out to free air again. The slot in the PCB may be enough to allow air flow with a matching cross-section size heatsink but the case must also be slotted or meshed above and below.
Overall, you would be wasting money by building this expensive 2 x 100W/channel amplifier but only fitting a (approximate) 2 x 30W heatsink. Building it as original, in a thick aluminium case, doesn't require any heatsink and should be cheaper, assuming you were going to follow the original design correctly and add a heat spreader bar between the transistors and the bottom of the case.
Thanks for your inputs.i have a spare heatsink with me
196mm long and 80cm height. Fins are the same
Here is the link - 196MM width/100MM length extruded aluminum heat sink.
Also I can have 2 cooling fans mounted close to the heatsinks, to facilitate heat dissipation.
I believe this heatsink plus 2 fans should do the job. Isn't it?
196mm long and 80cm height. Fins are the same
Here is the link - 196MM width/100MM length extruded aluminum heat sink.
Also I can have 2 cooling fans mounted close to the heatsinks, to facilitate heat dissipation.
I believe this heatsink plus 2 fans should do the job. Isn't it?
Fans help a lot. I'm using 3 mm x 80 mm x 140 mm AL angle, no fins, as a heat sink for one pair output transistors, 72 W avg; the two fans make it run cool.
It will depend on room temperature as to whether you can use any suggestion about minimum heatsink ratings and there are large differences in daily air temperatures and cooling needs between say, London, Mumbai and my location. However, it is best to avoid fans because they are are typically too noisy to permit optimum good sound, even when they seem silent. Dust accumulation also becomes a problem inside the amplifier and failure of the fan or its power supply is potentially disastrous.
In a suitable ventilated case, where air can flow up over the full heatsink width, the larger (100 mm?) size heatsink should be fine for 8 ohm use as is. Just ease-off on the maximum continuous power at lower load impedance. However, I don't have the PCB in front of me and unfortunately, I don't think the large slot in the PCB is long enough for full width airflow.
It is also important that air temperature in the case is only partly representative of the output stage temperature, to control bias safely as designed. If all air is blown either straight out or throughout the amplifier, the bias control circuit may need alteration. This is really something to avoid and the original assembly, which has the sensor transistors located near the bottom plate of the case, may not work correctly or at least respond fast enough.
Build it as the amplifier was designed and you won't have troubles. Trying to adapt it to use standard hobby bits and pieces may indeed work out but are they worth the compromise and hassle of getting all the extras to work correctly?
In a suitable ventilated case, where air can flow up over the full heatsink width, the larger (100 mm?) size heatsink should be fine for 8 ohm use as is. Just ease-off on the maximum continuous power at lower load impedance. However, I don't have the PCB in front of me and unfortunately, I don't think the large slot in the PCB is long enough for full width airflow.
It is also important that air temperature in the case is only partly representative of the output stage temperature, to control bias safely as designed. If all air is blown either straight out or throughout the amplifier, the bias control circuit may need alteration. This is really something to avoid and the original assembly, which has the sensor transistors located near the bottom plate of the case, may not work correctly or at least respond fast enough.
Build it as the amplifier was designed and you won't have troubles. Trying to adapt it to use standard hobby bits and pieces may indeed work out but are they worth the compromise and hassle of getting all the extras to work correctly?
I second Ian Finch's prudent comments.
This larger heatsink is rated at 1C/W which is fine provided the heatsink can get rid of the heat. The heat doesn't get absorbed forever by the metal of the heatsink, it has to transfer the heat into something else, like the surrounding air or more metalwork that the heatsink is attached to.
This is why a heatsink is only effective if it has cool air moving between its fins. Normally, heatsink fins are oriented vertically in free air so natural convection causes an air flow from bottom to top of the fins. In the absence of convection flow, a fan will be very effective and solve all your problems and you could get away with a smaller heatsink.
You can get away with the 1C/W heatsink and nothing else if you don't play very loud music into low impedance speakers on a hot day for a long time, etc. I would have thought for normal room listening levels and 8 ohm, moderately efficient speakers and a NAP low bias design and plenty of air holes in your case it would be ok. You can always attach a thermal cutout mains switch to the heatsink to avert calamity.
As for electrolytic caps, their electrolyte tends to dry out when hot over 1000s of hours of use. This is because they are never perfectly sealed. I would recommend 105C+ rated caps in hifi boxes where heat is generated. More expensive but then they maintain performance for many years or even decades. Consider high temperature rated electrolytic caps on the PCBs too, if there are any, as the PCB parts will be warmed by the heatsink beneath them.
Last edited:
If you are going to use fans think about the size of PC CPU coolers. Consider how small a fan and heatsink comes with a stock Intel CPU, which can dissipate 100W or so. With a NAP the transistors will dissipate about half the power of the speakers (or something like that). Very loud room speaker power is in the 30W average region and your NAP boards would dissipate 15W, give or take. so you don't need much fan/heatsink.
If you use a fan. Think of running it at low speed.
A fan at low rpm already drastically improves cooling on a heatsink.
With PCs we are used to high speed noisy CPU fans needed to extract watts in the 100s, as well as case fans that scream like hurricanes.
However going from heatsink with no fan to heatsink with a light breeze makes a large diference.
All my PC case fans are wired on 5volt instead of the 12v. This gives: About no noise and cooling good enough.
The idea is to break the hot air layer that would stay at the surface of the heatsink fins in still air. A gentle breeze is enough to get this effect that drastically improves cooling.
PS To monitor Temperture and Fan speed on a PC, I use "SpeedFan"
A fan at low rpm already drastically improves cooling on a heatsink.
With PCs we are used to high speed noisy CPU fans needed to extract watts in the 100s, as well as case fans that scream like hurricanes.
However going from heatsink with no fan to heatsink with a light breeze makes a large diference.
All my PC case fans are wired on 5volt instead of the 12v. This gives: About no noise and cooling good enough.
The idea is to break the hot air layer that would stay at the surface of the heatsink fins in still air. A gentle breeze is enough to get this effect that drastically improves cooling.
PS To monitor Temperture and Fan speed on a PC, I use "SpeedFan"
Last edited:
Hi Jrp27
I would suggest that you have a look at some of the articles on Rod Elliott's sites about heatsinks. You can find a good starter here.
Heatsinks And Amplifiers
Also, as one Mchamobin has suggested it is better to run a larger fan at low speed - it is difficult to hear the music over a screaming fan 🙂
Hope this helps,
Mike
I would suggest that you have a look at some of the articles on Rod Elliott's sites about heatsinks. You can find a good starter here.
Heatsinks And Amplifiers
Also, as one Mchamobin has suggested it is better to run a larger fan at low speed - it is difficult to hear the music over a screaming fan 🙂
Hope this helps,
Mike
Last edited:
Great insights...thank you everyone.
Here is what I have finalized. I would be using a tunnel heatsink with a fan mated at the other end.
Something like this...
I will also have a thermistor based fan speed regulator.
I don't play music at high volumes all I guess this should suffice.
What are your views?
Here is what I have finalized. I would be using a tunnel heatsink with a fan mated at the other end.
Something like this...
I will also have a thermistor based fan speed regulator.
I don't play music at high volumes all I guess this should suffice.
What are your views?
Will that heatsink fit between the output transistors and the power supply capacitors or is your PCB a different design to the original?
An interesting idea.
What fan size ?
If it is the small size as typically seen on removable hard drive trays, that's no good.
All of them are terrible about noise and reliability, including those claimed to be "silent".
P.S.to #16.. Note that the bias temperature sensing on the original PCB at least, is a long way from the output stage and uses the the case to conduct heat toward the sense transistors, so it actually settles a bit faster than older models. Any independent heatsink will most likely impede this thermal link and could result in a loss of bias control. Even the typical (and tedious) 30 min settling time for a stable bias with Naim's old products may never happen at all.
NAP and many other quasi-complementary designs are fairly forgiving of bias setting but if there is no effective bias servo action, you may also find yourself in trouble with thermal runaway, even with a perfectly working fan and heatsink.
NAP and many other quasi-complementary designs are fairly forgiving of bias setting but if there is no effective bias servo action, you may also find yourself in trouble with thermal runaway, even with a perfectly working fan and heatsink.
Yup the transistors will be mounted horizontally. So the heatsink will be below the PCB. There isn't enough space between o/p transistors and filter caps. So the best option is to fit it below the PCB.
Fan is a 80mm brushless fan.. very little noise. Infact u won't even hear it if running on low speeds.
- Status
- Not open for further replies.