Home » Sponsored » Pi Speakers » Speaker Voice Coil Cooling System Valve
Speaker Voice Coil Cooling System Valve [message #47206] Mon, 27 June 2005 17:26 Go to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

The prototype valve for the speaker voice coil cooling system is in my hands, ready for testing.

http://pispeakers.com/Cooling_Valve_Prototype.jpg
Voice Coil Cooling Valve

http://pispeakers.com/Cooling_Valve_Inserted.jpg
Cooling Valve Inserted in Speaker Magnet



The prototype is longer than production versions will be, and it is also slightly larger diameter. If it proves to work well, we'll streamline the device and make it smaller and easier to fit inside loudspeaker cabinets.


http://pispeakers.com/Cooling_Valve_Input.jpg
Voice Coil Cooling Valve Input

http://pispeakers.com/Cooling_Valve_Output.jpg
Voice Coil Cooling Valve Output



You can see the shape of the orifices inside the fitting in these photos. Unidirectional flow is created because each duct makes airflow turbulent in one direction, impeding flow in that direction. Airflow is laminar going one way and turbulent the other. That makes it much easier to flow one way, and by using two ducts, a warm air outlet and cool air inlet are formed. You would be surprised how effective this is. The difference in flow is immediately obvious, even when just putting your hand over the ducts and feeling gusts of air.

The next step is to connect pipes to the fitting and install an intercooler. There are a two pressures to balance: The volume of air inside the ducts and the volume of air behind the speaker cone. Each of these will contribute to cooling effectiveness and to system tuning. The rear chamber volume will be increased by a small amount due to the additional volume in the cooling system. This will reduce fb. Also, the volume of the cooling system should be large enough that flow is encouraged instead of pressure.

Re: Speaker Voice Coil Cooling System Valve [message #47211 is a reply to message #47206] Tue, 28 June 2005 13:14 Go to previous messageGo to next message
ToFo is currently offline  ToFo
Messages: 219
Registered: May 2009
Master
Hi Wayne,
that is really great looking. I will be excited to see what happens next. I have often scratched my head when simulating some of the long coil drivers, only to see that in my desired alignment (or any other in some cases) they can't reach xmax within thermal power limits. your package actually looks small enough to work with the small amount of air slug volume available from the dustcap back.

too cool,
Thomas

Re: Speaker Voice Coil Cooling System Valve [message #47212 is a reply to message #47211] Tue, 28 June 2005 15:46 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

I still remember when you and I discussed this almost a year ago. It's pretty exciting to be working on it, and I'm more and more surprised it hasn't already been done.


Re: Speaker Voice Coil Cooling System Valve - Initial Tests [message #47213 is a reply to message #47212] Tue, 28 June 2005 18:09 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

I am running a LAB12 with the valve installed. The signal is 40Hz, 40VRMS, which is approximately 400WRMS. I am running the driver open, so there is no pneumatic load to help the driver mechanically but it is not reaching its mechanical limits. I've even put 50VRMS into the speaker, which is 625WRMS. That exceeds both electrical and mechanical limits, but not violently.

I'm very excited about what we've done so far. We're really on to something here. There is obvious unidirectional flow. So that part is working. The heat generated by the speaker is tremendous, and my gut feel is that when the driver is used in a small sealed box such as the rear chamber of a basshorn, it must be superheating the air inside. I'm fairly confident that the majority of failures are thermal, provided it is used above the quarter-wave frequency, within the pass band of the basshorn. Whatever the case, we will certainly be reducing the heat inside by a lot.

Now the question is how much is a lot. We were going to use an infrared sensor to get some readings of the voice coil and magnet today, but we ran out of time. We also learned enough to realize we need to make a change first. Then we'll get some quantified measurements.

The thing we plan to change is the inlet. Right now, it's just a pipe. We plan to machine one from aluminum that has a lot of very small holes instead of one large one. The idea is to increase surface area. What we discovered today is that most of the voice coil heat is radiated onto the front plate and center pole instead of being convected into the air surrounding the coil. So we want to increase heat conduction from the center pole to the inlet, and use it as a heat exchanger from magnet to air going into the device. This will make total heat transfer better, and the system will be more effective. After we've made this change, we'll take heat measurements and make comparisons.


Re: Speaker Voice Coil Cooling System Valve - Initial Tests [message #47214 is a reply to message #47213] Tue, 28 June 2005 20:28 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

One other thing I forgot to mention. The venting system in the driver itself appears to introduce some unidirectional flow. The port tends to exit air more than it takes in. I expect there are turbulent regions that make the system tend to draw air in more around the spider and exit air more through the vent. The center pole gets extremely hot way down inside the motor, getting hotter as you get closer to the front plate. That's why we want to heat sink it to a metal duct that contains several small capillary-like passages. The metal duct will conduct heat away from the motor, and the capillary-like passages will have a large surface area to transfer this to the air.

If I were installing a cooling device like this in a loudspeaker with one of the cabinet walls very near the magnet, I might be tempted to make a large plate that acted as a heat sink, and connect it to a heat-conductive metal duct inserted deep inside the motor. This would allow air to pass through the duct, but would also heat sink to the plate. So for basshorns made this way, that might be a very attractive option. It's a similar system, but it uses the access panel as the heat sink.

Don't just place the heat sink near the speaker and hope convection will cool the motor. It won't. You'll need physical contact between the access plate and the vent, preferably via a metal duct that extends deep into the motor, and so has a lot of surface area contact with it. Also, I found that once a panel is placed within about 1/4" of the vent, it was shrouded so effectively that flow was reduced to almost nothing. So be careful that the vent isn't restricted and there is adequate flow.


Re: Speaker Voice Coil Cooling System Valve - Initial Tests [message #47216 is a reply to message #47213] Wed, 29 June 2005 09:55 Go to previous messageGo to next message
ToFo is currently offline  ToFo
Messages: 219
Registered: May 2009
Master
Wow, internal heatsink, crazy! As a quiet pc enthusiast I have considerable experiences to share in a couple of areas. You may already know these, but it can't hurt, so... Steel and iron are bloody awful as a heat sink material. About 1/4 of the efficiency of aluminum or copper. You will never get great cooling from a speaker pole, but you can make it better. A great thermal interface for your aluminum capilary or fluted pipe might do it. While we may never escape thermal runaway, we only need to significantly raise the threshold to be a success. My two cents is to go with the best thermal paste you can get. Using better paste will give you a 3-5 degree celcius advantage on a cpu. Most dismiss this seemingly tiny "advantage", but my experiments show that on otherwise identical systems, you would have to use twice the airflow to match the effect of the better paste. Try your best to get a really nice fit on the sink in the pole. Paste manufacturers say the thinner the paste the better the transfer, so close fit is it.

Thomas

Re: Speaker Voice Coil Cooling System Valve - Initial Tests [message #47221 is a reply to message #47216] Wed, 29 June 2005 12:28 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

I was thinking along the same lines. The conductive paste material between transistor and heat sink comes to mind. We'll use a pressed-fit aluminum vent and drill it with several capillary holes. It fits snugly in the center pole, so conductive heat transfer is good. The one we already have worked very well to introduce unidirectional airflow, and now we want to augment this with improved thermal conductivity from the center pole. I'm extremely encouraged with our progress so far.


Re: Speaker Voice Coil Cooling System - Heat Sink [message #47231 is a reply to message #47221] Thu, 30 June 2005 11:59 Go to previous messageGo to next message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

Since the motor core becomes extremely hot from induction and radiation, one could sink the heat into a device like that shown below.

http://www.pispeakers.com/Speaker_Heat_Sink.jpg


It's a simple device, just two parts.

One part is a large aluminum panel that attaches to the cabinet. If the system is a basshorn, this can double as the access panel. It is just a large sheet of aluminum, thick enough to prevent vibration or panel flex, since it will be used as part of the loudspeaker cabinet. Grooves may be cut in its surface to increase surface area. That's the key, because it will function as a heat exchanger, as a radiator of heat.

The second part is an aluminum rod, of diameter that fits snugly in the speaker vent hole. On one end, it is drilled with a threaded hole to accept a bolt, so it can be attached to the radiating panel. On the other end, it is drilled with a larger diameter hole, to act as the duct for cooling air. Holes are also cross-drilled along its shaft so that cooling vent air can pass through it.

http://www.pispeakers.com/loudspeaker_with_heat_sink.gif


The duct should fit snugly. When completely inserted, the internal end should be approximately flush with the front pole piece. That way the aluminum tube touches as much of the pole piece as possible, for most effective heat transfer. But it is not longer, so that it doesn't interfere with the cone or dust cap at high excursion.

These parts are simple and can be machined at any competent machine shop. Assembly is straight-forward. Put some heat-sinking compound on the open end of the tube and slide it into the speaker magnet vent. Use the same compound that is used on power transistors when mounting on a heat sink. Put some heat sink compound on the threaded end of the tube also, and attach it to the radiating panel with a bolt. Tighten the panel down and you're done.

http://www.pispeakers.com/Heat_Sink_Schema.gif


This kind of system can be used on basshorns with access panels near the speaker magnet. It can also be used in standard direct-radiating cabinets, by making a version with a longer rod that reaches from the cabinet to the magnet.

Speakers with exposed magnets, such as dipoles, isobaric and push-pull configurations may benefit from ducted arrangements since they serve to muffle vent noises in addition to assiting in cooling. But if the magnet isn't exposed, then this simple heat sink radiator might be more attractive.

Re: Speaker Voice Coil Cooling System - Heat Sink [message #47236 is a reply to message #47231] Thu, 30 June 2005 22:20 Go to previous messageGo to next message
ToFo is currently offline  ToFo
Messages: 219
Registered: May 2009
Master
I like this concept. I am not a master of heat, but I read from what I consider a good source, that beyond a certain short distance, heat conduction falls off rapidly. From what I remember it seemed like the inverse square law with shorter dimesions.(suposed to be the reason for heat pipes) Is this true? Would that force you to have a really short pipe? would a more heat conductive metal like copper help out? Maybe copper to move the heat and aluminum to get it into the air.(also read that copper is great for sucking up heat and crappy at disipating it into air. didn't understand exactly why)

Food for though (or not, I am not sure)
Thomas

Re: Speaker Voice Coil Cooling System - Heat Sink [message #47237 is a reply to message #47236] Thu, 30 June 2005 22:49 Go to previous messageGo to previous message
Wayne Parham is currently offline  Wayne Parham
Messages: 18670
Registered: January 2001
Illuminati (33rd Degree)

The M.E. would be the one to answer that, but one thing is certain. Aluminum is a better conductor of heat than air.


Previous Topic: tuning the Alpha 10"
Next Topic: A few pics to share
Goto Forum:
  


Current Time: Thu Mar 28 14:46:23 CDT 2024

Sponsoring Organizations

DIY Audio Projects
DIY Audio Projects
OddWatt Audio
OddWatt Audio
Pi Speakers
Pi Speakers
Prosound Shootout
Prosound Shootout
Smith & Larson Audio
Smith & Larson Audio
Tubes For Amps
TubesForAmps.com

Lone Star Audiofest