Heat buildup in sealed boxes

H

Helge A. Bentsen

Senior
Jan 19, 2011
1,016
63
48
43
Oslo, Norway, Norway
drbentsen.no
I've always heard that sealed boxes will have too much heat buildup inside so they can't be used in pro audio(subs), but I've never seen any data on it.

Is this a known issue?
Is there a way to calculate the temperature rise?

I have a sealed box on the drawing board in two editions, one with the driver mounted with the magnet inside the box and one with reversed mounting, and I can't decide between them on this issue.
 
Re: Heat buildup in sealed boxes

You can calculate the temperature rise if you know the thermal resistance between the driver and the ambient environment. In practice, it's probably easier just to measure the temperature inside and outside the enclosure during testing, especially if you are planning on building a prototype anyway.

My guess is that the ambient temperature inside the enclosure isn't significantly higher in a sealed box than in a similar ported box, but I don't have data to back that up.
 
Re: Heat buildup in sealed boxes

You could also measure the increase in DC resistance of the voice coil then back-calculate using the thermal coefficient of resistance - I think off the top of my head it adds about 0.004x per Celsuis for copper but that should be relatively easily verifiable (assuming conventional voice coil material of course).
Not sure how accurate your multimeter would have to be to get meaningful data for any smaller changes however.
HTH,
David.
 
Re: Heat buildup in sealed boxes

The large majority of the power dissipated by a speaker driver goes into heat rather than acoustic energy, so I would guess you could approximate the heat radiated into the box as the wattage the speaker is drawing, knowing you are only overestimating slightly. That suggests that there would be substantial heat buildup in a typical live sound application with a thousand watts or more powering a sealed sub all night. Power compression even with reflex ported subs is a common problem. I think power compression rather than burned voice coils will be your main enemy, and I would predict this could happen a lot sooner with the magnet inside a sealed box. How much the driver voice coil heats up depends on wattage it draws, voice coil / magnet design, volume of the box, efficiency of production of sound energy vs thermal energy, thermal conductivity of the metal speaker frame, stuffing of the box, porting or not, proximity of the magnet to the ports, and other factors. Most of these factors would be difficult to model and predict. Empirically measuring the heat in a prototype would seem to be the most practical way.

I built a sealed 12" sub for my recording studio monitors but I never use more than a quarter or half of the max power handling of 300 watts RMS. Even so I have considered building sealed subs for live sound use (very small gigs) and thought a lot about putting the magnet on the outside where airflow will ventilate and cool it. As I remember from my BassBox Pro simulations, the box volume isn't critical so you may want to just go ahead and build a prototype of the magnet-outside design, mount the driver with the magnet inside, and actually measure the magnet temp rise inside the box at full SPL. If it is insignificant, you can build the rest of your boxes with the magnet inside. If it gets too hot, build the magnet outside design. Since power compression will likely be the predominant SPL-limiting factor, I personally would bet on the magnet-outside design for your first prototype.

If you have a bass reflex sub with a similar wattage driver, you might want to just put a temp probe on the magnet, stuff some thermal insulation in the ports and crank it up to just below the power handling rating of the driver.

Or you could wait for Art Welter or someone else who actually knows from experience what they are talking about to comment!
 
Re: Heat buildup in sealed boxes

Helge Dr. Bentsen said:
I've always heard that sealed boxes will have too much heat buildup inside so they can't be used in pro audio(subs), but I've never seen any data on it.

Is this a known issue?
Is there a way to calculate the temperature rise?

I have a sealed box on the drawing board in two editions, one with the driver mounted with the magnet inside the box and one with reversed mounting, and I can't decide between them on this issue.
Click to expand...

You need to think in terms of energy.

How hot does the inside of the box need to be compared to the outside before it can transfer energy at the rate equal to the amount being dissipated by the voice-coil. The greater the difference the great the transfer rate.

I would assume almost all of the energy you put in to the driver turns to heat. Music’s average to peak ratio means about 125 to 250 watts of heat for a 1000W driver at full power.

Here are some very simple examples of the type of calculations needed.
http://formulas.tutorvista.com/physics/heat-transfer-formula.html

The k for wood is about 0.15 W/mK and paper (the speaker cone) 0.05 W/mK ... but it varies.

http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

For more accurate calculations you need this type of approach -

http://windows.lbl.gov/software/therm/therm.html
 
Last edited:
Re: Heat buildup in sealed boxes

Thanks for the info :)

Four of my old labsubs didn't have aluminium covers, just wood covers. Didn't experience any issues with them, but I didn't do any dubstep-festivals either.
Based on experience, a sealed box with the magnet inside should work, I guess the question is for how long.
 
Re: Heat buildup in sealed boxes

As usual what Peter said....

If you do a patent search you find a large number of patents dealing with the process of getting heat out of the box.

These techniques vary and include trick drivers that use cone motion to pump hot air out. I am familiar with one patent (a friend of mine got) for speaker ports that bias the direction of air flow kind of like a diode, so that the normal air movement inside a port actually pumps cold air into one port and hot air out of the other. More simple a high and low pair of ports will enjoy some convection flows as warm air escapes out the top port to be replaced by cool air coming into the lower port.

The sealed box works against exchanging hot internal air with cooler outside air. Still air is a pretty lousy conductor of heat.

For limited duty cycle and time most designs can survive for the time it takes to heat up their thermal mass. For 24x7 design with high power output a sealed box requires some extra consideration IMO. While just like amps, few speakers are 24x7 at full power so any high duty cycle design requires consideration.

JR
 
Re: Heat buildup in sealed boxes

Helge Dr. Bentsen said:
Thanks for the info :)

Four of my old labsubs didn't have aluminium covers, just wood covers. Didn't experience any issues with them, but I didn't do any dubstep-festivals either.
Based on experience, a sealed box with the magnet inside should work, I guess the question is for how long.
Click to expand...
I did some testing on one of our high output subs that utilize a normal type horn, and the original plan was to use thick aluminum panels for the access hatches.

I had the regular aluminum panel on one driver and a braced wood panel on the other driver. Both were exposed to the surrounding air equally.

So I put thermal probes on the magnets of each of the drivers and brought them outside so I could measure each of the drivers.

I ran pink noise at full power for 2.5 hours-monitoring the temp from time to time.

After they started "warming up", the aluminum panels were between 2 to 3° C cooler.

After 2.5 hours the temp had gotten to around 80° C. The limit for the drivers is 100 C.

As they were cooling off I also monitored the temp. The wood panel driver was cooling off faster than the aluminum panel driver.

As a result of this test-we changed to the wood panel-simply because it costs less to manufacturer and we did not feel that 2° C was worth it.

This was INSIDE. When outside the aluminum panel could conduct the suns heat and bring it INTO the cabinet.

Also the panels would not be exposed to the outside air all the time. In some setups they would be covered-which would reduce the heat transfer.
 
Re: Heat buildup in sealed boxes

Helge Dr. Bentsen said:
I'm considering buying a cheap thermometer with a remote probe, tape the probe on the magnet and play 50hz for a couple of hours. Poor-mans data sampling, but it might be good enough.
Click to expand...
Sine waves have a crest factor of only 3 dB, while pink noise has a crest factor of 12 dB, and the usual test signal is compressed pink noise with 6 dB crest factor.

A sealed box will increase the temperature compared to a ported box, here is an actual temperature comparison of 20-2000 Hz pink noise taken from the white paper "Maximizing Performance From Loudspeaker Ports" which is chock full of useful information:
 

Attachments

  • Maximizing Performance From Loudspeaker Ports.png
    Maximizing Performance From Loudspeaker Ports.png
    140 KB · Views: 16
Re: Heat buildup in sealed boxes

Art Welter said:
Sine waves have a crest factor of only 3 dB, while pink noise has a crest factor of 12 dB, and the usual test signal is compressed pink noise with 6 dB crest factor.
Click to expand...

Just to be pedantic, truly random pink noise in a system with infinite bandwidth will have infinite crest factor. However, by the time you cram pink noise through a band-limited system it ends up having about 12dB crest factor just like essentially all music signals.
 
Re: Heat buildup in sealed boxes

Bennett Prescott said:
Just to be pedantic, truly random pink noise in a system with infinite bandwidth will have infinite crest factor. However, by the time you cram pink noise through a band-limited system it ends up having about 12dB crest factor just like essentially all music signals.
Click to expand...

Wow, thanks for pointing that out ;-) The more important point is that, pink noise will have a power spectral density closer to that of music, or at least that is what some would argue.
 
Last edited:
Re: Heat buildup in sealed boxes

It can be more like music and still not be like all music.

Using shaped (filtered. limited, whatever) noise for power testing speakers is still more realistic and should be considered for amplifier power testing.

Similarly rated amps might be easier to match, apples to apples, with speakers.

or not...

JR
 
Re: Heat buildup in sealed boxes

John Roberts said:
It can be more like music and still not be like all music.

Using shaped (filtered. limited, whatever) noise for power testing speakers is still more realistic and should be considered for amplifier power testing.

Similarly rated amps might be easier to match, apples to apples, with speakers.

or not...

JR
Click to expand...

Well we are talking about the long term heating of the magnet structure, not really testing the amps or speakers, etc. That was all I was getting at. In a lot of tests it might be important to talk about this in terms of 'crest factor', but I would argue not so much in this case.
 
Last edited:
Re: Heat buildup in sealed boxes

Mark DeArman said:
Well we are talking about the long term heating of the magnet structure, not really testing the amps or speakers, etc. That was all I was getting at. In a lot of tests it might be important to talk about this in terms of 'crest factor', but I would argue no so much in this case.
Click to expand...

You can talk about what you were talking about and I will talk about what I am. (I already commented on the thread topic).

Amplifiers have short term and long term heating issues too.

Matching amps to speakers seems to be a chronic problem of wide general interest, because of typical test methods that don't simply relate to each other.

JR
 
Re: Heat buildup in sealed boxes

Mark DeArman said:
Wow, thanks for pointing that out ;-) The more important point is that, pink noise will have a power spectral density closer to that of music, or at least that is what some would argue.
Click to expand...

Hey Mark, I prefer EIA RS-426-B's power spectrum to represent the long term power spectrum of music... but we're obviously on the same page here.

EIA RS-426-B.png
 
Top Bottom