Solid-state guitar combo amp: using speaker-forced airflow for cooling heatsink

[Thomeeque]

 Hi,

Is this "trick" used in some solid state combos? If yes, could you give me some examples, please?

I'd like to turn my closed speaker cabinet into simple LM3886 based combo (so far I'm using this cabinet with simple LM3886 head and I quite like it, I'm just bit tired by this head+cabinet configuration transportationwise ) and I'm looking for some inspiration.



I'd like to mount everything inside just by attaching it to the walls (no chassis or special separated section), I just wonder how to manage LM3886 cooling (I'd like to avoid heatsink mounted from the outside if possible). I'm aware of the fact, that since it's closed cabinet, I'll have to "open it" little bit (make few holes to it) of course , but less is better.

Thanks, T.

PS: sorry for this light off-topic, but I'm too used to this forum already

BTW: I will not disassemble existing head, it will stay as is, there will be new guts..

[R.G.]

The problem with opening a closed speaker cab is that
1) the speaker response in the bass changes - that's why closed, vs open, vs ported designs exist
2) it's very easy to get whistles, whooshes, and other noises from the air going through the port.

But the heatsinking would be fine. All it could mess up is the sound.   

Have you considered covering the back panel with a sheet of aluminum and bonding the LM3886 to it inside? That would be a visually smooth surface, and if thick enough, might give you enough area to cool the LM3886. I haven't done the numbers, so I don't know how much aluminum or how thick, but it would let you keep the existing setup.

The other alternative is a slam-dunk. Cut a hole in your back panel, cover it with a flat-back heat sink with short fins, and mount the amp to the sink inside the cab like many of the hifi subwoofer designs. You could even inset the heatsink into the panel cutout at some loss of heatsink efficiency.

[Nasse]

Electro-Harmonix Freedom Amp (70´s version) was very small closed box and most of heat generating parts were inside. Did google search and somebody claimed it was 55 watts. I bought one in the 70´s and the salesman told me it was 100 watts. He added "american watts" when he thought I dont trust his numbers.

There was extra jack or something and there was strong airflow above, narrow air beam, shaped like pencil, you sure could power something with it  

It had the transformer and pcb randomly screwed to the bottom plywood or chipboard, if I remember, that version I had.

[digi2t]

This smacks of perpetual motion 

[teemuk]

Is this "trick" used in some solid state combos?

Gibson (Chicago Musical Instruments) patented it in the late 1960's so most certainly it is. 

[Thomeeque]

 
Thanks, RG!

The problem with opening a closed speaker cab is that
1) the speaker response in the bass changes - that's why closed, vs open, vs ported designs exist

I see, I would not mind slight change in bass range (I cut lowest bass anyway already in amp)..

2) it's very easy to get whistles, whooshes, and other noises from the air going through the port.

Yep, I was aware of that, that's why I was looking for some existing combo amp with all these issues solved.

But the heatsinking would be fine. All it could mess up is the sound.    

Whistles, whooshes, and other noises I don't want to add, that's for sure

Have you considered covering the back panel with a sheet of aluminum and bonding the LM3886 to it inside? That would be a visually smooth surface, and if thick enough, might give you enough area to cool the LM3886. I haven't done the numbers, so I don't know how much aluminum or how thick, but it would let you keep the existing setup.

Hmm, I like that! I may even know where to get Al plate easily and cheap, but "numbers" are problem - heatsink calculations is something I have no experiences with yet - is there some good tutorial how to calculate it, please? I'll search for it myself, but hints are welcome

The other alternative is a slam-dunk.

Cut a hole in your back panel, cover it with a flat-back heat sink with short fins, and mount the amp to the sink inside the cab like many of the hifi subwoofer designs.

I'm trying to avoid any jags and/or sharp edges on outer surface..

You could even inset the heatsink into the panel cutout at some loss of heatsink efficiency.

..and this would probably mean relatively big hole in the back panel, which I'd have to obturate (wafer, tighten..? - I'm not sure, I'm using dictionary for this word) around the heatsink to avoid whistles & whooshes yet.. doable, but I like big plate idea more, now it just depends on what will the numbers say

Thanks a lot, T.

[Thomeeque]

Electro-Harmonix Freedom Amp (70´s version) was very small closed box and most of heat generating parts were inside. Did google search and somebody claimed it was 55 watts. I bought one in the 70´s and the salesman told me it was 100 watts. He added "american watts" when he thought I dont trust his numbers.

There was extra jack or something and there was strong airflow above, narrow air beam, shaped like pencil, you sure could power something with it  

It had the transformer and pcb randomly screwed to the bottom plywood or chipboard, if I remember, that version I had.

Thanks, Nasse!

Do you mean this article? Either I get the scale wrong or even those 55 watts look pretty "american" to me The trannie looks like 10-15W max.

Do I see it correctly, that there is like inch gap at the bottom of back pannel?

T.

[Thomeeque]

Is this "trick" used in some solid state combos?

Gibson (Chicago Musical Instruments) patented it in the late 1960's so most certainly it is. 

Cool! Would you know about any particular (Gibson) combo using it? Thanks, T.

[Thomeeque]

 I wonder, if I would create two holes (let's say cca 3" diameter, rounded edges) at the bottom of back panel and place heatsink and PT like this:



- if I would get any airflow-noise issues?
- if it will cool well?
- what change I could expect sound(frequency-range)wise?

T.

Internal height is 434mm (17.09")

Edit: Or maybe just one centered hole for 3886 heatsink and PT somewhere not so far from this hole (PT will not produce that much heat, right?)..?

Edit2: Speaker is OEM version of Eminence Legend 1218.

[teemuk]

Cool! Would you know about any particular (Gibson) combo using it? Thanks, T.

I believe it would be the GTR600. But its overall design is not really close to what you are trying to achive.

First of all, the cabinet of that amp - or any such amp - can not really be "closed" because inherently the loudspeaker must push air to the heatsinks and the air must circulate so that fresh cool air can replace the heated air. A closed system will just store most of the heat.

So, you need to ask yourself: Is it a sealed cabinet which you're after, or a ported one? Two entirely different designs, the latter one being very picky about proper dimensioning in relation to loudspeaker's parameters.

A ported cab is much, much more easier to adapt into forced cooling because it inherently has air flowing in and out of the port. You dimension the port and cabinet volume properly and then place the heatsinks to critical spot, hoping they will not obstruct the airflow too much. Basically, you just line the heatsink's fins properly and maybe even try tunneling the damn thing so that the heatsink tunnel actually forms the port. Works well in quite a many tabletop speakers for computers and alike, why not even in a guitar amp.

It should be needless to mention, but a ported/ducted reflex system is extremely sensitive for both proper cabinet volume and proper port/duct volume. These need to closely match the loudspeaker and you should employ software like WinISD to calaculate some ballpark values for these dimensions based on loudspeaker's TS parameters. And likely after that build a few prototypes, ideally with a duct which's lenght you can fine tune. The duct itself should be ideally as smooth as possible to prevent the airflow from generating side noises, so implementing heatsinks to the system is always more or less a compromise.

The Gibson's design was not really trying to be a ported scheme and itself it is actually straghtforwardly simple: The speaker compartment is closed from all sides except from the top panel where you have few carefully placed slots letting air out and in. The amplifier's chassis sits atop this compartment and the chassis has holes in the bottom located under the heatsinks. These holes line up with the slots in the top panel of speaker compartment. Now, the pressure inside the speaker compartment will push air through this path. Cooling air will also try to flow in from same direction. This is where you easily get to those odd whistling noises because the entire flow path is obstructed as hell. If you going to have air flowing in and out of the cab, it must do so fluently, otherwise you ran into effects similar to blowing into a bottle or putting fingers into your mouth and whistling. You get the drill?

Also, the thing is not really a closed/sealed back design at all but a ported reflex system. And as such, poor as hell. For their defense, the designers never likely even thought what they created by using such configuration. It just seemed like a neat idea. So, the performance of the design is actually quite subpar both as an acoustic cabinet and as a forced cooling solution.

Secondly, the heatsinks of that amp weren't entirely relying on forced cooling to begin with. In fact, the chassis compartment of that amp was pretty much open from the back and simply covered by a grille mesh. Third, this amp had quite hefty heatsinking. So basically,it still had decently sized heatsinks sitting in open air. The airflow from the speaker compartment might have done some extra but overall the safe opeartion within proper temperature limits didn't even depend on it.



Overall, I don't understand why you can't simply build something in this vein. Chassis in its own compartment and heatsinks covered by a grille mesh so you don't tear yourself into them or damage them when the amp gets transported. Notice also the cabinet shape: It also tries to achieve this same function; the heatsinks or even the protective grille are not portruding outside. This is a much more effective solution than trying to force cool with loudspeaker (especially if the cab ideally should be entirely sealed) and its also much better solution for plain convective cooling than "plate amps" or heatsinks inside a moderately sealed enclosure.

[Thomeeque]

 Thanks a lot teemuk for this very informative post!

So I'll be careful with "opening" my cabinet

I'll try to do heatsink calculation and go with Al plate attached to the back from outside with LM3886 mounted from inside thru (sealed) hole as proposed by R.G. - it looks like less invasive way, where cabinet still stays closed.

Hm, when I think about it - PT will not be cooled this way - does it matter?

Overall, I don't understand why you can't simply build something in this vein..

I'm not building it from the scratch, cabinet already exists (you could see it in the first post if my homepage was not down for some reason today) - plus it is firmly glued together and covered by (again glued) carpet - I don't want to mess with and ruin it to much. And this should be relatively quick mod - I don't want to spend months with it. If I'd want to, I would probably rather start from the scratch.

Thanks, T.

[R.G.]

Hm, when I think about it - PT will not be cooled this way - does it matter?

Mount the PT on the plate too.

I did think of one more issue with speaker-duct airflow cooling.

If you try to breath through a tube underwater, it works great for short tubes, but as the length of the tube increases, the volume of air in the tube itself also increases, and at some point the volume of air in the tube itself approaches the amount of air you inhale and expell. When that happens, the air in the tube is simply rebreathed over and over and you soon suffocate because you're reusing the same air and it's filled with CO2 and depleted of oxygen.

The air in a speaker port only moves so far, then returns the other direction. It doesn't so much flow as vibrate back and forth (oh! it's sound vibrations!) so if the port is longer than the back and forth travel of the air, then the air stays in the port and moves back and forth, like the air in the breathing tube. So the port has to be very short, likewise the heat sink, otherwise it mostly heats the air in the port, and does not exchange for cooler air. Some will be exchanged at the ends of the travel with each stroke, but far less than the amount of air motion would make you think.  And air is a good thermal insulator, so the part of the heat sink covered with non-exchanged air might as well not exist as far as the heat dissipation it provides.

The plate sink seems better and better.

[R.G.]

At the risk of making this more complicated, you might look here:

http://apexjr.com/amps.html

They stock an assortment of amplifiers on metal plates, complete. You might have to modify the circuit to UN-lowpass them. But the mechanical part is already solved and complete.

I ... hate... the mechanical part of construction. It's always difficult and time consuming.

[Nasse]

Yep, that plate like design might be good if I ever re-create teh amp I once had. I did few google pic searching for Matthews Freedom Amp (many versions but I had that black three knob version without phaser or batteries) and it must have been done so, that power transistors are on that same metal sheet as pots and jacks, like many small solid state combos. But power transformer was on the bottom, and rectifier perhaps. It might have been other transformer imported to Europe (220 volts ac) that us versions

It must have been just local pressure over that empty jack hole. Loudspeaker paper cone is relatively thin and does not isolate much, loudspeaker metal frame and magnet have some thermal inertia, like the air inside and wooden cabinet walls. The box was very closed, only hole was that extra jack, the pic on australian link, the back part is not original. So I believe it could be done without any forced airflow or something, if done properly. Average temperature inside should watched. Over 80 degrees celsius long time and wooden parts might catch fire.

http://www.thomann.de/fi/the_tamp_pm40c_endstufenmodul.htm Self made amp is of course self serviceable

[Thomeeque]

 Hi,

for those who might be interested about following developement, here is a brief report (including some porn )



Thanks again for all your input!

Cheers, T.