The only problem discovered is that a cast aluminum box of about 4" x 4" x 5" can not dissipate 100 watts at temperatures below the level of burnt fingers. With 25 to 30 watts, it gets comfortably warm. The problem is, "How to dump 100 watts without a fan". I believed he wanted a power dissipator that would fit comfortably behind a guitar amplifier, kind of in the speaker area. Right now I am imagining a 19 inch rack mount box with lots of fins for radiation and convection, but I'm not going to do it because the customer is happy with this heat compromise, the performance, and the price.

 

I'm not going to do it because the customer is happy with this heat compromise, the performance, and the price.

in simple words: "if it ain't broke, I ain't fixin' it" ...

 

The problem is, "How to dump 100 watts without a fan".

The right (small and quiet) fan could help a lot. Why not use one?

 

How about using the air pumped by the speaker cone? That's what is used in amp speaker combos.

 

A speaker does not pump air. Instead it vibrates air back and forth a very small amount.
The port on a bass-reflex enclosure makes a puff of air with each drum beat. The puffs are not continuous air flow like from a fan.

 

The right (small and quiet) fan could help a lot. Why not use one?

Entirely doable with a wall wart, a computer fan, and a lot of holes, but right now, we are stuck at, "The customer is so happy with it that he refuses to send it back for modifications."

I think I said early in this thread that I have never done anything for this guy that was perfect on the first try. I'm keeping the tradition alive.
If he sent it back, I would change the attenuation range so he doesn't have the volume set on, "2" and add a fan if he wanted one.

 

How about using the air pumped by the speaker cone? That's what is used in amp speaker combos.

There are tiny piezo fans which wag a fin at a transistor, but I don't believe you can scale that up to 100 watts with guitar driven speaker movement. I can do air flow calculations. This would need about 6.5 CFM. Ten would be better, and very achievable with a computer fan.

 

There are tiny piezo fans which wag a fin at a transistor, but I don't believe you can scale that up to 100 watts with guitar driven speaker movement. I can do air flow calculations. This would need about 6.5 CFM. Ten would be better, and very achievable with a computer fan.

Wouldn't the fan's noise be an issue? Like, if the device is used for recording sessions?

 

Wouldn't the fan's noise be an issue? Like, if the device is used for recording sessions?

I don't think so. I'm sitting 2 feet from a computer with an exposed fan right now, and I can barely hear it. Besides, recording sessions are generally done VERY LOUD or hard wired from an instrument (like a keyboard) to the mixing board. This Power Soak is designed to hang behind the amplifier. That would also attenuate it's noise.

Meanwhile, I just posted the air requirements. 6.5 CFM to keep the dT to 50F. I think you can do that with a 2.5 inch fan.

 

Meanwhile, I just posted the air requirements. 6.5 CFM to keep the dT to 50F. I think you can do that with a 2.5 inch fan.

Or maybe just convert the shell to fins for passive cooling. I’m picturing one of those higher power inverters you can find. The entire outer shell is aluminum ribbing. It might get hotter inside than with a fan but no user burns. Extra rugged too.

 

I figured out how to deal with the suggestion of AG. If you start with the load resistance of the speaker in parallel with a nice, low value damping resistor, the math almost does itself. How many decibels do you want for minimum attenuation? The input resistor compared to the load and damping resistor in parallel. How much is the maximum attenuation? Presto! The value of the pot arrives. Can't buy a pot that size? Add a resistor across the pot or from the input resistor to the wiper.

I feel that AG could have been a lot more helpful.

As for the nice, low value damping resistor...a transistor amplifier has a damping factor of less than one ohm. I might guess that a guitar amplifier has a damping factor consistent with the impedance of the output winding of output transformer, 8 ohms. That can't be true because of the magnetic coupling back to the vacuum tubes. The plates of the vacuum tubes are high impedance, but the path back through the output transformer primary includes the power supply capacitors. What is that impedance? Apparently above my pay grade, but maybe AG could help. I am attaching a schematic of a Fender Super Reverb for his convenience.

 

For an audio amplifier, the impedance of power supply capacitors is always a dead short at audio frequencies.
The Fender amplifier has negative feedback (820 ohms) from the output to an input stage which reduces the output impedance.

 

I feel that AG could have been a lot more helpful.

Should have told him your from India.

 

For an audio amplifier, the impedance of power supply capacitors is always a dead short at audio frequencies.

You can't put a dead short into a Power Soak for a damping resistor. I believe the correct answer is 8 ohms or less, but how much less? It depends on the inductance of the primary of the output transformer, right? Or does it all depend on what resistance will damp the speaker sufficiently below 2 KHz and the amplifier isn't all that important because all modern amplifiers have a much lower impedance than needed? I don't have any push-pull guitar amplifiers here right now. The next time one comes in for a repair, I will do a frequency sweep on it to find the impedance...unless you have an approximation right now. Will it be the DC ohms of the windings? Will the windings have enough inductance to be significant from zero Hz to 2 KHz? Is the answer really about the speaker impedance?

 

The very low impedance of a power supply capacitor is not parallel with the load, instead it is parallel to the power supply to keep the voltage steady.
A speaker resonates at 30Hz to 150Hz which is far from 2kHz.

 

The very low impedance of a power supply capacitor is not parallel with the load, instead it is parallel to the power supply to keep the voltage steady.
A speaker resonates at 30Hz to 150Hz which is far from 2kHz.

Your answer is not a number, an equation, or a concept I can use to find a number. The nicest thing I can say is, you didn't rant about how much musicians don't know about sound. I thank you for that.

 

I have repaired, restored and rebuild tube guitar amps for quite a while. I know how difficult it is to get the distortion and sustain required by musicians at acceptable volume levels. The best ways I have found are either having them use a smaller amp (5-10 watts) that will distort but not make your ears bleed or by installing a post phase inverter master volume control. That way you get the distortion from the entire preamp and phase inverter section without driving the power tubes into additional distortion. The power soaks do work but they make nice little heaters too! Just remember to keep the speaker load somewhat constant on a tube amp and match the amp to the load or you could be replacing output transformers which 'aint cheap!