I disagree about a fan being a viable option in a high end amplifier. You're spending megabucks to get low noise, blacker blacks, higher dynamic range and you're willing to sacrifice that to fan noise? That makes no sense at all. Bigger heat sinks and better heat management is the answer, not fan and its inherent noise.

Your post is a good example of information being used out of context to make a point that is biased in its conclusion. For example, the amp is overheating and dangerously hot. As I recall JA' informal method of temperature determination, he was unable to keep his hand comfortably on the amps chassis and heat sinks, which he usually states is about 80 degrees C., 176 Farenheit. Warm, and perhaps dangerous to a persons touch but certainly not hot enough to start a fire, the ignition point of paper being 450 F for example. But why is the amp at this temperature, playing music at one watt? No, it has just pumped out 100 watts RMS for one hour, per FTC testing standards. After that regimen, most amps tested are at that temperature, or have shut down before the hour had elapsed. The new Class D amps are not stressed by this test, but otherwise the reaction of the FBI was quite normal. To characterize the amps reaction to the FTC conditioning as an indication of its normal funtioning is just plain wrong.

Overheating is a relative term, a vacuum tube is not normally touchable in use, is it overheating? Are light bulbs dangerous if they cannot be touched while in use? I have touched voltage regulators on CD players and DACs that were uncomfortably warm, but operate at that temperature in normal operation. My Krell preamp has voltage regulators that run very warm and has done so for years without problems. I read that the wear and tear semiconductor devices undergo from heat is not caused by the heat itself, but from the repeated heating and cooling cycles it experiences when powered on and off. Fan cooling works of course, but is not used much because of fan noise and the need for maintenance. Check the insides of your PC and its power supply to see the build up of dust that will occur. I have a Crown amp I use on a subwoofer in a HT system, it has a fan that I can hear the moment the soundtrack changes from loud to quiet. Fan noise has long been an issue.

Krell is known for running their amps in class A or with high bias current that keeps the transistors running in class A. Class A operation means high heat dissipation. They've done that for years, and given their five year warranty coverage and the cost of shipping those monster amps around, they must not have many problems with heat damage.

It would be interesting to see how your AVA amps would react to the FTC conditioning procedure. Mosfets (used in your amps) are known to increase their internal impedance as their temperature rises, which limits how hot they will become (as well as their output capability.) Yet they still need fan cooling? I wonder what temperature they would be at after one hour at one third power? We know the Krell amp will survive, would the AVA? A mystery. . .

Quote:
do you also turn off the heat or AC in your house while listening. When the music is playing, the fans ain't heard. A $16,000 amp that runs what seems exceedingly HOT is a dumb design. It will certainly shorten it's reliabilty. Long term lifetime. Even rack systems have fans in place checkout most pro setups fan it baby. Smoldering amps and electronics can't sound better than ones that are running cooler with fans. Most fans are black so it shouldn't have any effect on.."blacker blacks" Low noise fans are also available.

I actually do turn off AC while listening. I find the AC noise to be extremely distracting. I have no problems with fans in a low to midfi systems, but if you're going to be paying a gazillion dollars for an integrated, there shouldn't be any fan noise associated with it. Otherwise, you may as well pay a lot less for a cheaper integrated because you won't be able to hear all the subtleties that a 16k integrated will be able to recreate.

I don't like outdoor concerts because the acoustics suck. I prefer indoor venues. I don't hear the AC at a live event indoors, so that doesn't apply.

I usually listen to music at night, so I turn off the AC during the time I listen. If/when it gets too hot, I stop, turn on the AC, cool off the room and then resume my listening. I want to experience the recorded environment when I listen to a recording, not my own environment intermixed with the recording venue. Hence, I minimize the effects of my room, I pick out audio components that do not detract from my listening experience, I minimize noise within my equipment and my listening environment. If you're not willing to do that, then I see no point in spending so much money on audio equipment that strives extract the Nth amount of detail and ambiance from the recording and take the noise level down to imperceptible using power conditioners, cables, etc, etc.

High end audio is about getting the last 5-10% out of your recordings. If you're not willing to minimize extraneous noise within your listening environment, then you're just wasting your money on high end equipment. At least in my opinion.

Perhaps if you turned off your fans and your AC you wouldn't need cascading amplifiers supplying 12 megawatts of power into your speakers.

Seriously though, I think that room noise defeats the purpose of a high fidelity musical reproduction. Noise is still noise no matter how you try to mask it.

Transistors fail between 150C and 175C. What you should consider is not just the temperature of the heatsink, but the temperature of the semiconductor junction.

For instance, in this example you could roughly assume that the transistors connected to the heatsink will have a temperature gradient of around 1C per watt. So if the amp has say 2 transistors in its output sage, and the amp is dissipating say 50watts then each transistor is getting rid of 25watts
This means that the junction temperature will be 25C above 80C or 105C.

How does this shorten the amps life? Well MIL217 is a book primarily written for the US military. It is a set of equations that is used to predict MTBF (Mean Time Between Failure)
As a transistors junction temperature approaches its failure level then MTBF reduces exponentially!

Thanks, Martin. Interesting info.

Yes, Krell has a very strange business model. Create products that self-destruct from heat damage and make them extremely heavy and expensive to ship, so the cost to do so is so high it uses up all the profits when they must be constantly repaired and shipped back and forth during the warranty period, and subsequently the company goes belly up. Once past the warranty period, the customers will be angered by the high cost and difficulty of shipment, not to mention repair, thus creating a disatisfied customer base that won't buy Krell in the future. Apparently this model is not working, as they have not closed their doors and have continually expanded their product line to include HT components and loudspeakers.

My ten year old Krell amp had a problem in one channel one day, as a small puff of smoke rose from its chassis vent upon turn on. So, did a cascade of component failures in a domino affect occur within the amplifier, or fuses blow or were wild AC waveforms or a DC power supply surge sent to my speakers? No, it quietly shut itself off with no further harm. If I tried to turn it on again, it would just shut down with no trace of any other component failure. The problem turned out to be . . . an electrolytic capacitor, a small one on an amp board. First time it ever had service and never has since.

I'm sure you'll spin this into some terrible design flaw or who knows what else. I'm pleased for you that AVA amps will never die and their impeccable design and performance is at the pinnacle of the amplifier art and all else are the equivalent of a boom-box IC amp chip taped to a beer can heat sink (polypropylene tape of course). So enjoy, and meanwhile I'll try to start a new thread regarding the FBI's review.

Gosh, bi-polar, how do you live with the day to day disappointment of an amp that just keeps working and working while producing excellent sound? It must be exhausting.

Quote:

Transistors fail between 150C and 175C. What you should consider is not just the temperature of the heatsink, but the temperature of the semiconductor junction.

For instance, in this example you could roughly assume that the transistors connected to the heatsink will have a temperature gradient of around 1C per watt. So if the amp has say 2 transistors in its output sage, and the amp is dissipating say 50watts then each transistor is getting rid of 25watts
This means that the junction temperature will be 25C above 80C or 105C.

How does this shorten the amps life? Well MIL217 is a book primarily written for the US military. It is a set of equations that is used to predict MTBF (Mean Time Between Failure)

Quote:

As a transistors junction temperature approaches its failure level then MTBF reduces exponentially!

Let me add my 2 cents in layman's interpretation....

Increased Operating Junction temp of 10 deg C above 85 deg C will reduce MTBF ( Life ) by 50% !!

Lets assume that a Power Transistor has an MTBF of 500,000 Hours ( not untypical )

If it is operated with its junction temperature of 20 deg Centigrade above 85 deg C, the Power Transistor MTBF will drop from 500,000 hr to 125,000 hours

Assuming that a Power Amplifier has 5 Pairs of output transistors in parallel per channel ( a total of 20 Output devices ) the MTBF for the Amplifier will be :

125,000 divide by 20 = 6250 Hours

That is just 260 days of continious use ( which could also translate to just 260 days of being left powered on, for a Class A amplifier... even with No signal.

THis example is NOT related to any specific commercial product, but is given to show that Heat KILLS ... rapidly, if temperatures are not adequately controlled by passive or active heatsinking.

Cheers

I do the same thing, Alex. I cool the room, then listen. Outdoor concerts? Not unless I'm overpowered by reefer madness. I attended my last Hollywood Bowl concert about 15 years ago...the sound of a bad AM radio. I attended my last Rolling Stones concert about 20 years ago, at the Oakland Coliseum. I didn't go there for high-end sound, but for the publicly naked babes and the, er, casual atmosphere.

DUP, your AVA stuff is no better or no worse than anything else out there in a similar price range. If you need that much grunt to "move air," you need some science to verify air movement as a measurement criterion. Here is what I suggest. Put a couple of anemometers in your room, and scatter a few wind socks around your listening chair (in a pinch, you could mount condoms on sticks). Then wear a propeller beanie, tied down with a chin strap, as you ramp up the velocity. Chart the results. How many watts does it take to get you airborne? Did the wind socks (or condoms) change directions with frequency? You could be a pioneer. This may be your last shot at fame and fortune -- you sure as hell ain't gonna win any spelling bees.

Happy tunes, all.

Quote:
(in a pinch, you could mount condoms on sticks)

Be sure to keep track of the style and brand of condoms; it matters.

Does AVA make condoms?

Yes, Elk, I am sure they do. I have heard that the bigger models are hybrids, with tubed output stages and fans.

There once was a man at Assizes
Whose balls were of different sizes --
The left was quite small,
Of no consequence at all,
But the right one was large,
And won prizes.