It's kind of sad what tone he uses in his post. However, I am not very surprised.
Yes, I understand, but you'd need something quite different to the standard epoxy package to remotely make it worth your while.
Hard sell for anything outside of the (legit) high end space of instrumentation.
Nothing is happening, what's the problem? This piece of wood is sitting under my monitor and it looks the same as yesterday. More to come?
I don't know why you think a system that is set at a temperature below a changing ambient or pulling -q at Tset is not better than a system that needs to convect at some temperature above a moving ambient. Even if the case does not have the highest thermal conductivity, it still conducts!
Last edited:
Actually I am suprised that no one has tried to market a mineral oil filled amplifier. Freon and LN2 I can understand are a little hard to handle but oil is comparatively easy to contain (for a given value of easy).
Deep sea remotely operated vehicles have long used mineral oil to displace the air
for pressure sensitive components like electronics.
I meant in domestic audio with added oil of snake marketing. My bad for being lazy and skipping words...
I have used audio equipment that was oil filled. There are one or two spring reverbs that have an oil filled reverb tank. The obvious reason is to dampen the movement of the spring and make it not so "boingy". Works too.
Cheers
Alan
Cheers
Alan
Hi CC,
-Chris
True, but there will be a gradient involved between the heat producing part and the point where you are drawing the heat out. You're further ahead allowing the case to reach a thermal equilibrium on its own. Just don't run the stuff too hot.
-Chris
well, isn't most of the thermal affects caused by the output stage? Find one of those old IC/opamps which has a pin/leg at the ops input or Vas out node. Measure affects of temp with and without loaded ops. and maybe have separate discrete ops with same loading but not creating internal temp gradients.
-THx-RNMarsh
-THx-RNMarsh
Hi Richard,
Absolutely!!
But the effects of temperature gradients across the diff pair or multiple current sources were studied long and hard. These days the effects are minimal. Still, why suffer with any when all you have to do is use the chip intelligently. I'd bow to Scott on this topic, he's the expert at this little party.
-Chris
Absolutely!!
But the effects of temperature gradients across the diff pair or multiple current sources were studied long and hard. These days the effects are minimal. Still, why suffer with any when all you have to do is use the chip intelligently. I'd bow to Scott on this topic, he's the expert at this little party.
-Chris
A few years ago some PC overclockers were using 3M Fluorinert for this purpose. It's impractical though, and not as good as LN2 for extreme stuff obviously, so regular water cooling with machined blocks that encompass the switching regulators are now common on the very high-end.
There are a lot these all-in-one sealed "water" cooling systems that are popular now. They can be pretty effective but who wants to hear a pump.
Not exactly a perfect test but you can just use a unity-gain buffer/amplifier in the loop like the LME49600.
Yeah, I had been meaning to mention the effects of the extreme overclockers give some information about what you'd gain with all that cooling. Bare chip or bust.
Richard: depending on the opamp, sometimes the standing current in the input pair is a significant portion of the power/thermal budget.
Modern opmps measure vanishingly low steady state distortions.
Are thermal behaviors a subjective differentiator ?.
Dan.
You might have something important there, Max. Steady state is not going to tell us much in this case, but thermal transients 'might' be a real cause for concern.
Of course, this was especially true back in the uA741 days, and designers have known about the problem since Solomon published his paper on this and other linear IC problems back in the early 1970's.
However, contrary to what Scott Wurcer says, I doubt that the problem of thermal feedback has been completely eliminated in EVERY linear IC that is used typically in audio, and many times, for various reasons, the linear IC's can be excessively loaded down by feedback and load resistors more often than you would think.
Of course, this was especially true back in the uA741 days, and designers have known about the problem since Solomon published his paper on this and other linear IC problems back in the early 1970's.
However, contrary to what Scott Wurcer says, I doubt that the problem of thermal feedback has been completely eliminated in EVERY linear IC that is used typically in audio, and many times, for various reasons, the linear IC's can be excessively loaded down by feedback and load resistors more often than you would think.
Hi Dan,
Subjective? No, I wouldn't think.
Remember that industry uses most of the devices and the audio industry represents a very small, but high maintenance market. So effects caused by thermal variations are judged by how they affect measured performance.
Sorry Dan, you will always be at the mercy of some dude with a meter between you and any devices you are interested in. If that Dude decides that certain parameters are more important than ones that affect sound, guess what? You're getting the device that satisfies those measurements. That's in general covering the entire industry. Even diodes. They decide that a certain characteristic isn't used enough for sales volumes, that part hits EOL, or End Of Life and is discontinued. This is what is happening at TI with their audio product line. As long as high volume manufacturers use those parts, they will be available to us. That's it, that's all.
-Chris
Subjective? No, I wouldn't think.
Remember that industry uses most of the devices and the audio industry represents a very small, but high maintenance market. So effects caused by thermal variations are judged by how they affect measured performance.
Sorry Dan, you will always be at the mercy of some dude with a meter between you and any devices you are interested in. If that Dude decides that certain parameters are more important than ones that affect sound, guess what? You're getting the device that satisfies those measurements. That's in general covering the entire industry. Even diodes. They decide that a certain characteristic isn't used enough for sales volumes, that part hits EOL, or End Of Life and is discontinued. This is what is happening at TI with their audio product line. As long as high volume manufacturers use those parts, they will be available to us. That's it, that's all.
-Chris
- Status
- Not open for further replies.