The critical temperature of the listening room and why it should be controlled

[A.S.]

One of our members comments that she can hears a marked change in the bass quality of a Harbeth speaker over recent months and has spent time and money 'upgrading' her amp in the fruitless and unsuccessful attempt to restore the quality she prefers. I asked her to confirm the temperature in her listening room. She checked: at the speaker cones the temperature was 30 degrees Centigrade in the late evening, and doubtless much higher at mid day. This was a shock to me. Clearly the issue is not with the amp, or CD, or cables.

Asked how she could tolerate such incredible temperatures indoors and in a windless apartment she said that the air-conditioning has been turned off because it made some acoustic and electrical noises which disturbed her hifi listening.

All loudspeakers from all manufacturers are designed at and with a target operational temperature in mind. By convention, this is 20 degrees C, which is universally deemed to be the ideal temperature for human comfort: neither too cold nor too hot: just right. So what? Well, 30 degrees C is a +50% rise on the target operating temperature (which would be, say 18-20 degrees = 20 degrees +/- 10%) and this has a dramatic effect on the molecules within the drive units - specifically in the plastic cones and the rubber surround. It does not effect the steel magnet parts.

Rubber is an extremely temperature sensitive material. You know that from playing with BlueTak or chewing gum or a squash ball. When it heats up it become much more elastic and that increase in elasticity even over the 20>30 degree range has a marked effect on the bass unit's rubber surround. Its throw increases, control, damping and Q all degrade. So all the Thiele-Small parameters that have been carefully calculated to give a nice smooth, tight bass at 20 degrees C are completely screwed by the temperature rise. So, when the temperature is elevated as high as 30 degrees, the bass will be softer, flabby and there will be a lot more of it.

Strange things happen to the molecules inside the cone too at elevated temperatures. For all typical bass/mid plastic cone materials (including RADIAL) if the temperature is substantially elevated above target, the acoustic damping characteristics of the cone change - the 'air' dries out around instruments, and the sparkle is lost as the clarity in the presence band diminishes.

I would add, that this temperature sensitivity makes it virtually impossible to design loudspeakers in the UK during the winter when the ambient temperature is around 5 degrees C. At these low temperatures, the bass unit's rubber surround tightens.

For me, critical design and measurement is only viable in the UK from April-June and September-October. Attached is a picture of me setting up to do some testing in the big open space of the Village Hall. When I made the reservation for the day I had to insist that the heating was turned on at 7.30 am so that by the time I arrived (11am) the entire hall was up to 19.5 degrees C: perfect. Conversely, last December, with an outside temperature of 2 degrees C, I was obliged to make some measurements on a prototype speaker to be shown at the Jan'06 CES. The only available large space was the local dojo hall, which they kindly loaned to me. Even though the heating had been on all day, the temperature only reached 10 degrees C and as expected, my acoustic measurement of the systems true bass output was completely useless. See how critical temperature is? Before you can even think about measuring (or designing) there as so many critically important but seemingly inconsequential little details that you have to get right first! Grrrrr!

Rubber and plastics have completely reversible temperature characteristics, so once the temperature is returned to 20 degrees C, all will be well again.

Conclusion:

So taken together, the significantly elevated room temperature will .....

A. Give much more bass, less controlled and slower bass ...
B. Reduce clarity and detail in the presence band.

F.or critical listening, listen within a range of about 18-22 degrees and note that hifi demo rooms are normally set at this cool, shirt-sleeve temperature.

/

[Damian L.]

I know this is nit-picking and not particularly relevant, but I think it is slightly misleading to say that a 30?C is 50% higher than 20?C, since the 50% numerical difference only applies when the Celsius scale (which is rather arbitrary in in having the freezing point of water as its zero temperature) is used. For comparison of temperatures, it would be more meaningful to use abolute an temperature scale, e.g. the Kelvin scale. You would then be comparing 303K with 293K, which gives a much smaller percentage difference.

Not that that has any bearing on the change in rubber properties over the temperature range in question!

[A.S.]

... I think it is slightly misleading to say that a 30?C is 50% higher than 20?C, since the 50% numerical difference only applies when the Celsius scale ...

You are, of course, absolutely correct but I was trying to make this as simple as possible. I guess that's a byproduct of thinking in spl or voltage dB's where, for example, 6dB = +100% = a doubling and -6dB = a halving.

I considered adding that all physical materials have properties that lie on a continuum from being a solid, through being 'plastic', through being liquid finally to a gaseous state. We have recently seen in Indonesia that even granite, when heated sufficiently becomes molten lava (the plastic state), then a white hot gas.

Those engineers who design products we normally associate with operation in a wide temperature ranges - such as EADS or Boeing - are acutely aware that they have to select materials that are just at the right point on their solid-thru-gaseous path otherwise the structure will fail, and lives put at risk.

The effect of ambient temperature seems to been neglected in the humble monitor speaker. I don't recall seeing any particular mention of temperature in audio design books or product reviews when it is so easy to illustrate by measurement that ambient temperature is, and has always been a critical component of loudspeaker performance - let alone the effect temperature has on ones own body, hearing and brain, which we know, perform very differently in the heat and cold.

[jttlee]

Dear Alan,

Thanks for the insight.

How about humidity? In Hong Kong, it is normal to have relative humidity around 80-90% in summer. It is not affordable, or practical, to have air-conditioning turned on all day.

Will continuous exposure to high humidity damage a Harbeth speaker permanently? It worries me a lot because unfortunately my current pair of English-branded speakers (made in the UK as well) was permanently damaged by humidity. They actually developd a version 2 of the bass unit to cope with the humidity problem, which is quite common in South-East Asia.

Regards,
John

[A.S.]

Over the last 29 years since Harbeth was founded, the vast majority of our speakers have been sold to and are daily used in the Far East. I am not aware of any humidity issues in either the cabinets or the drive units.

You may not be aware - it is just another one of those details that seem inconsequential - but we veneer both the inside and outside of the cabinets to equalise the enormous bending power of real wood when it dries. As far as we are concerned, there are and have never been any issues of using Harbeths in the tropics, providing that the cabinets are out of the direct sunlight.

As for humidity and its effect on the drive units, I do not believe it to be an issue at all.

One thing that is a definite 'no-no' is foam surrounds. We have never used them as they're guaranteed to rot in the heat and humidity. We mustn't complain though, because we make large numbers of replacement drive units for one of our (British) competitors who did, atyplically, use foam for a drive unit surround - and now regret it!

However, there is one plus to using foam surrounds: it greatly diminishes the mass of the moving parts which in turn, increases the efficiency (and hence sensitivity, loudness) of the speaker. But gaining those extra dB's comes at an expensive long term price ....

[jttlee]

Alan, your attention and consideration of such minute details deserves much appreciation.

Based on past communications with several speaker manufacturers, I feel disappointed that designers don't always take environmental factors into consideration. IMHO, though sound quality is of utmost importance, temperature, humidity and normal home fittings all are factors to consider in the design process, especially for domestic speakers.

Regards,
John

[A.S.]

Alan, your attention and consideration of such minute detail ...

Thank you for your kind words. It does take time to contribute here and I do so because I am still as curious and interested in monitor speaker design as I was thirty five years ago as a teenager (don't the years flash past?) and I want to do my best to preserve my knowledge for the future - just in case someone should be remotely interested.

When one start out as a speaker designer the canvas is vast and blank and you see nothing but opportunities. As ones experience grows with time, the confines of reality - the engineering dead-ends, the aborted designs, the abandoned concepts accumulate, and you realise that huge constraints dominate all speaker design. That canvas is actually the size of a post card.

What matters is the harmonious working with those constraints, not pulling against them, on the basis that there is no such thing as a perfect speaker.

[A.S.]

Today was probably the hottest day of the year. My infra-red probe advised me that the temperature at the cone itself was 28 degrees C - stifling hot.

I measured various drive units to see how much their parameters had shifted from 20-20 degrees, my normal test temperature. Since magnetism, resistance and mass are not influenced by temperature, any variation would be mainly related to the softening of the rubber surround.

Harbeth surrounds are moulded in rubber from a precision compound, and in the case of the P3ES2 woofer, there was a 16% drop in surround stiffness, which is unlikely to be noticeable.

In the case of the B110 SP1228 used in the '11 ohm' LS3/5A, the surround is moulded onto the cone from a roll of commercial grade PVC sheeting, not rubber. (We perform this operation for KEF as it is common to several of their drivers). Surprisingly, under the same conditions the PVC surround becomes 24.6% less stiff - which is by any accounts a substantial change in characteristic.

[Douglas G A Murray]

Today was probably the hottest day of the year. My infra-red probe advised me that the temperature at the cone itself was 28 degrees C - stifling hot.

Very interesting thread!

And my admiration for Alan grows with with time.

Well, what do we do in Queensland, Australia, where for 6 months of the year the average daily max. is greater than 30 deg.C and it is a battle for domestic air conditioners to keep the temperature indoors to below 28-29 deg.C during the day. Without the air conditioners temperatures indoors are in the high 30's.

I have often wondered about the effect on my Super HL5s.

Douglas G A Murray

[Wolf]

I'm in the same position Douglas, I live in Los Angeles winter temps are in the 20 C range, but summer is average 30(which I find comfortable), and this last summer we had 2 months of at least 38 C and more. Quite uncomfortable but my electric bill would have been astronomical if I had AC runing all day. Usually it is just a week of hot weather, not that long, and then back down to 30C. I guess it is all a perspective of where you live. I know I couldn't live out in Palm Springs where it averages 38C and heads up from there in the summer. Some people just love it tho.

[steve]

Very hot here in Israel too, also the A/C is turned off if we are out at work otherwise it would cost a small fortune. So I think the speakers ought to be built to tolerate rooms which for Allen in UK may seem "unbearably hot"

Steve Rogers

[A.S.]

Loudspeakers in common with all other mechanical devices are to a greater or lesser degree, temperature sensitive. It's the soft, rubbery components that have a performance keyed to temperature whilst the harder components (such as steel parts in the magnet) are not at room temperatures.

The rubber blend used in the British-moulded cone surrounds that we use in our RADIAL drivers is blended and compounded by rubber engineers and scientists here in the UK who really know all there is to know about rubber. They should - they've been in business for seventy five years. Most rubber surrounds in hifi speakers are made in mysterious factories in far east and we question whether those sources really understand the technicalities of temperature performance and acoustics or whether they are driven solely by lowest cost. Take a look at a RADIAL driver and you can see how beautifully moulded the surround is.

We would not want to use them - they are not repeatable, nor do those companies have the engineering staff that can tune the material to give us the acoustic properties we need. To offer that customer support, the engineers must have a deep and wide knowledge accumulated over a generation as to what chemicals in what proportion, moulded in which way give certain hard to define acoustic characteristics.

You do not need to be concerned about temperature v. acoustic performance under conditions which you could reasonably tolerate indoors. That said, rubber being flexible, you would expect the bass unit to be more floppy at, say, 30 degrees C compared to 15 degrees C - and that is so.

Because we source the humble, anonymous surround from a UK supplier who actually makes it here, I believe we have far better control over this critical part than most other speaker manufacturers.

Oh - and by the way, note that the previous post was from Israel (a good market for Harbeth) - an order crossed my desk today from BBC Israel for five pairs of M20.

[kittykat]

...Thiele-Small parameters... /

There is a quite an interesting and recent interview with Mr. Richard Small at this link...

http://www.avhub.com.au/Features.asp...&FeatureID=109

[hifi_dave]

My speaker store room is unheated, so during the winter months I bring in all speakers I might need for a demo, into the demo room to warm up prior to the dem. I also run them with pink noise for a few minutes to further warm them up.

[kittykat]

It's finally getting cooler in the Southern Hemisphere and reaching optimum listening temperature. hee hee. Its about 20-21 degree C inside the house (without heating) where i am and this cooler weather will really only last for another 3-4 months.

[markus sauer]

A pertinent quote:

Personally, I think that of all the conventional driver components, suspensions are the one still in a primitive stage. I am amazed that designers are still content to use bits of paper and fabric with properties that vary wildly with the environment. It would be nice to have a breakthrough here that would give us at least the same kind of environmental stability that magnets, coils and cones achieve.

Source: http://www.avhub.com.au/Features.asp...&FeatureID=109

[PK_fin]

It's finally getting cooler in the Southern Hemisphere and reaching optimum listening temperature. hee hee. Its about 20-21 degree C inside the house (without heating) where i am and this cooler weather will really only last for another 3-4 months.

Well, here in north it is finally getting warm. Sometimes even up to 20-21 degrees C outside (14 at the moment). And this warm will only last for about two months.

[Astroalex]

Dear Friends, I have owned Harbeth M30 for a couple of months or maybe less. These loudspeakers are driven by an YBA Passion Intègré 300; CD player is a Cayin CDT-23 and cables and interconnects are Van den Hul. The sound is wonderful.

Here in Rome in this terrible summer, the temperature of the place I live and paint (I am a painter) is 30° C. I cannot stand air conditioning and so I just try to survive. I must say that I succeed in doing it quite well, but after reading the interesting notes by Alan I am a bit worried. Could this terrible hot damage my speakers for good? I can tolerate a less accurate performance of the whole system with "softer and flabby basses" (even if honestly in my case it is not so dramatic, maybe my Harbeth speakers had already to get used to these latitudes, it is for them a serious issue of survival), but I couldn't bear any serious and permanent damage to the speakers.

Thank you.
Best regards

[STHLS5]

Hi, Alan mentioned this:-

....Rubber and plastics have completely reversible temperature characteristics, so once the temperature is returned to 20 degrees C, all will be well again.
...

I wouldn't worry too much about it. If that is the case then there would no speakers can survive in my tropical country. I am not sure if you can find anything in the manual. The link http://www.harbeth.co.uk/uk/uploadfo...rguidancee.pdf seemed broken.

ST
167

[Astroalex]

THANK YOU STHLS5. And anyway we shouldn't care about our stereo systems more than we do about our own bodies, more than we do about ourselves.. But maybe we consider them a part, an important part of ourselves!. And this is wrong.
Ciao