I know the subject is a banality and that there should be answers in the diyaudio forums archive or somewhere else on the Internet, but I have failed finding them.
I'm currently researching for an active speaker project based on Lambda SB10 in a small enclosure using linkwitz transform. The SB10 is inefficient as ****, so an amp capable of delivering up to 500w short term (10ms, either 4 or 8ohm depending on the driver) would be desireable.
A standard class ab amp with an effiency of 50-60% rated Xw Yohm continously will dissipate Zw at full power. My question is, when used with a predictable load (as you have in an active speaker situation, where the amp will not see a load tougher than Yohm nominally) how effective should the heatsink be (how much heat should it be able to dissipate without reaching a temperature above ambient + 30-35 degrees celcius)? Is it really nessescary to dimension the heatsink according to maximum sine power? A typical music signal is pretty far from a sine wave signal when it comes to power density.
I'm looking forward to your input.
I think you pretty well cornered the issue. To start with, there is of course the quiescent dissipation with no signal. That's the easy one. Then, a designer would try to do an informed guess as to the worst case disspation to expect from the signal. designing for continuous 2/3 power output (that about were the max amp dissipation lies) is overkill, I think, at least for home use.
I would attack this as follows. You have the efficiency of the speakers, and assuming you'r listening to LOUD sounds in your home, say at 110 continuously, you can calculate the power required to drive the speaker to this. If your speakers are 80dB/W/m, going to 110dB at 4m means either 30+6 or 30+12 dB (I'm not sure about doubling distance, is that 3 or 6 dB?).
So, worst case 42dB above 1W. Calculate the amp dissipation at that output, add the quiescent, and you have the total. Say it is 60W, and if you allow 30degrees temp rise above ambient, you need a heatsink with a 0.5deg/W thermal impedance.
42dB above 1W is from PD(dB)=10log(Pi/P0)=
16KW. Nice amp on nice speaker.;)
Janneman, thank you. I'm pretty familiar with heatsink ratings and the different constants that apply to the rating dependig on the situation (a 100mm heatsink will not be able to dissipate two times the thermal energi compared to a 50mm based on the same extrusion and the C/W value does also depend on the difference between ambient and heatsink temperature).
What I'm after is a specific C/W value for 30-35 degree
kelvin/celcius temperature rise for an amp rated... lets say 300w into 8 ohm. Can you help me? The midwoofer and the tweeter amps will probably be rated for 80w into 8ohm each (class a/b).
When I look at different built-in plate amps for subwoofers I'm sometimes amazed by the indeed very small heatsink used considered the power output rating of the amplifier. Since heatsink extrusions are expensive (I would want something like 800x150x60 extrusions for the active three way speaker project) and hard to get I would certainly not want to spend more money than nessescary, that is, buying an "adequate" heatsink and nothing more.
By the way, what problems concerning vibrations and their effect on the amps should I expect when building an active loudspeaker and how can I avoid such? Has anyone seen the amp sections of Pass Labs' Rushmore? How has Nelson delt with the vibration issue?
Problem with high power amps (such as 300W) is that they still dissipate significant power at low output power. The reason is that efficiency on low output power is well bellow theoretically possible 70% for AB class.
After testing few songs in Sound Forge you can discover that average power is at least 10 db below maximum (I use a song from Guano Apes - as loud as possible and get -13 db result).
Calculation for 300W on 8 ohm speaker
To get 300W on 8 ohm speaker you will need 70v output amplitude. Amplifier rails must be a little higher (75v at least)
Assume that maximum average power is 50W (instead -10 db or 30W for safety margin).
50w on 8 ohm gives 3.5A peak output current.
Average (not RMS) current drawn from each rail is 3.5/pi = 1.115A
Power consumption is 1.115 x 75 x 2 nearly 170W !!
This result means that you need heatsinks able to dissipate 170w - 50w = 120W.
If you allow larger temperature rise over ambient temperature ie 60 deegrees then 0.5 C/w heatsink will be adequate.
BTW lot of amp manufacturers use these presumptions to calculate transformator power rating. For This kind of amp 200W toroid will do the job for them.
Thank you for the explanation, dulel. I think I understand the heatsink issue now. I start to wonder if a Zappulse module would actually save me some money due to efficience (both smaller heatsinks and transformers).
A professional amp design engineer I know, told me that because the energy content of music material is far lower than if you were playing sinewaves, you can get away with a far smaller PSU. In a typical consumer amplifier, the supply will be able to output around 1/8th of the current draw from the amp (output power/efficiency in %) continuously, while a professional amplifier for stage or PA-use typically has a PSU continuous output of 1/3 of the current draw. The calculations here are definately "worst case". You will be able to get away with using a much smaller heatsink.
There is an advantage to using a larger sink such one of size needed for the worst case sine situation. Keeping the output devices cooler during normal operation than they "need" to be places them closer to the "good: section of the derating curves on the spec sheet. This leaves you with more headroom for peaks and keeps the SOA larger.
Of course, there are advantages to both a larger heatsink and a beefier supply, but it's still nice to know what you could get away with if you had to :)
Any Sony-engineer who specified heat-sink and power supply like the avarage diy'er does would be handed a hira-kiri sword with instructions :smash: :D
Well, the last Sony receiver I had blew it's output transistors from overheating. So maybe swords shold be handed out to two guys - one who selected the itty-bitty heatsinks and one to the guy the designigned the thermal "Protection" circuit.
|All times are GMT. The time now is 02:49 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2015 DragonByte Technologies Ltd.
Copyright ©1999-2015 diyAudio