Robert Harley's TAS Editorial

[Robh3606]

It takes a friggin long time to get them at good operating temperature.

Really?? Why worry about operating temps?? You want them to run cool. My amps barely ever get noticably warm except when I am running them harder than normal. The distortion figures always seem to be lower at lower wattages so why would they be worse when run at lower levels.

Rob

 

[vinylphilemag]

The 30W amp performing as I mentioned in the previous post is equivalent to a vehicle being able to maintain its maximum speed on 3 different steepnesses of hill, this is true headroom -- the amp will always perform with a sense of ease. The key selling point for Krells, for example, the power doubling as you double the load.

EDIT: A simple way of expressing the above is to say headroom as you think of it = ability to double the power when doubling the load, irrespective of what the measured maximum power number actually is.

Frank

Just a quick correction: you presumably meant "doubling the power when halving the load" (not doubling the load).

 

[Gregadd]

So Harley hears things that can't be measured and is not bothered by things that can. Despite his unfortunate reapeted use of the term "magic" I find myself in agreement with him. Efficeint speaker are plentiful and there's also horn speakers aplenty.

 

[microstrip]

So Harley hears things that can't be measured and is not bothered by things that can. Despite his unfortunate reapeted use of the term "magic" I find myself in agreement with him. Efficeint speaker are plentiful and there's also horn speakers aplenty.

Greg,

First, the word was not magic , the expression was sound so magical.
Second, IMHO, it was not unfortunate. The words were properly used in the context of the full editorial. Taken out of their context,or used by those who do not understand or do not want to understand, it can be used to discredit the author. I know many people disagree, but I hope that the high-end writings keep their freedom of speech, and are not obliged to restrain from using words just because their interpretation is not political correct and a few people can dislike it.
Third, again IMHO, RH seems really bothered - it is why we concludes (I quote, assuming the risk it can be badly interpreted by those who did not read the whole text)


But when I listen to music through a pair of Lamm ML2.2s, I can’t help but question whether nearly a century of conventional wisdom about what makes an amplifier “good” has led us down the wrong path.

After all this disagreement, I can say we find ourselves in agreement with him in the fundamental of the editorial!

 

[fas42]

Just a quick correction: you presumably meant "doubling the power when halving the load" (not doubling the load).

I wondered if someone would query that, when I wrote it: yes, the load is halved in terms of the value of its impedance, but the load with respect to the amplifier is doubled, it has to work twice as hard, so to speak. I used that nomenclature to emphasise in a, non-technically correct, way what the issue is from the point of view of the amplifier.

Frank

 

[Gregadd]

I am currently without the article or thesaurus /dictionary. Suffice it to say we are in agreement on the important point.. The SET is not a niche product, but a serious device capable of bringing us closer to the real thing, when used propeorly.

 

[fas42]

Really?? Why worry about operating temps?? You want them to run cool. My amps barely ever get noticably warm except when I am running them harder than normal. The distortion figures always seem to be lower at lower wattages so why would they be worse when run at lower levels.

Rob

Actually, Jack is making a good point: It is not so much the temperature, per se, but that the electrical parts have been sized to be able to deal with running at high powers. Which means, that if running at light loads, that the internal structures of these parts haven't reached an equilibrium with respect to all sorts of parameters.

My suggestion when you want to do serious listening, and soon, is to thrash the system hard beforehand; driving R&R at realistic volumes, say: don't have to be in the room at the same time, of course :b! If the system is in good order, then you should be able to put on solo violin, say, and it will be sweet and true ...

DACs have this problem badly, there is a lot of capacitor coupling in some implementations, including mine, so to "condition" those capacitors I run a maximum level, high frequency, test signal through the circuitry for an hour from cold. The difference in SQ if I don't do that is quite dramatic ...

Frank

 

[Robh3606]

Actually, Jack is making a good point: It is not so much the temperature, per se, but that the electrical parts have been sized to be able to deal with running at high powers. Which means, that if running at light loads, that the internal structures of these parts haven't reached an equilibrium with respect to all sorts of parameters.

Hello Frank

I don't know what you do for a living but I work in a company that packages raw die. I have also worked in facilities that have die probe capability in wafer form before the transistors, diodes and IC get cut out of the original wafer they are fabricated in. The actual working parts of a transistor/IC/diodes all the active devices get up to temp long before you can feel any heat in the heat sinks. For many power devices depending on how they are biased the actual warm up is much quicker than you would think.

Actual die sizes for diode and transistors can be as small as .020x.020. How quick do you think it takes a piece of silicon that size to reach equilibrium?? You know what the real answer is NEVER! Know why?? It's a slave the signal being feed to it. With a dynamic signal they never reach a constant "optimum temp" simply because each piece of music has unique power demands. The actual temp of the die changes all the time depending on signal and load.

When you do the die probe in wafer form you use very low power because the die are not heat sinked and that is when all the die selection happens as far as the parts meeting their small signal electrical requirements. Parts that fail are inked out and discarded during the dicing operation when the wafer is cut so the individual die can be separated. They then get packaged and screened for their high power testing.

Just because you don't feel a warm heat sink in no way shape or form is an indication that the die are not at a "working temperature". You have to remember the heat sinks are designed to keep the die safe at max output not at light working loads. At light loads you should feel next to no heat at all unless you are talking about a class A amp.

Rob

 

[JackD201]

First off, I'm in the tropics so I'm airconditioned. Cold to begin with. Secondly transistors change transconduction with temperature. Heatsinks, fans even active temp/biasing circuits are designed to keep these in an optimal range. Not much talk about amps running cold because the amps don't blow up run cold but do when overheated. This doesn't change the fact that some amps do not perform as they should when they are below the desired operating temp range. If distortion is defined as a deviation from the input signal, that's what you get when the amp is unstable, hot or cold.

With a light load, the amp will get there eventually in an airconditioned room. It just takes longer. My solution is to run them for about 30 mins with the airconditioner off.

 

[fas42]

Hello Frank

I don't know what you do for a living but I work in a company that packages raw die. I have also worked in facilities that have die probe capability in wafer form before the transistors, diodes and IC get cut out of the original wafer they are fabricated in. The actual working parts of a transistor/IC/diodes all the active devices get up to temp long before you can feel any heat in the heat sinks. For many power devices depending on how they are biased the actual warm up is much quicker than you would think.

Rob, did EE at university but never did it for a living: computer systems design and building was my game. So the electronics side has always been a hobby adjunct ...

Agreed, the die may stabilise but that's the least of my concerns -- I'm talking about every single component, not just the actives. Capacitors are notorious, especially electrolytics; the power transformer needs to come up to full temperature, the insulation in many parts of the circuit may take a while to settle in terms of its characteristics. The whole exercise is equivalent to taking a racing car out for a warm up lap: it's not just engine, but the gearbox,suspension, brakes, tyres, etc, all have to get in their "normal" state. Some of these stabilising "effects" may be very subtle, but on a highly tuned system they could mean the difference between delight and disappointment.

Frank