Cooling of amplifiers

I use an older Heathkit SB-200 output stage. This forty-year-old PA is equipped with two glass triodes 572b and works really well. However, years of operation have resulted in increased fan noise. It is a 120V motor with a metal propeller, which swirls the air in the tube space. Not just noise, but also vibrations, which the fan causes, they are unpleasant. In addition, this method of air flow is not the most advantageous - the air mainly swirls in the space of the tubes. You can find an excellently written article on tube power stage cooling at www.ok1kvk.cz

Harbach Electronics offers a new fan, but its price including postage is high. The use of axial fans would require major mechanical modifications. However, they are available at GM Electronic exhaust turbines aside, so-called. blower SUNON. This allows for sub-chassis mounting along with the air tunnel. This will ensure the flow of air through the holes in the chassis into the space of the tubes and then the outflow through the top cover. The team, that the air does not swirl, but it goes around the tubes, the cooling efficiency will increase.
The SB-200 must first be supplemented with a 12V/300mA power source. A suitable space will be created after replacing the old electrolytic capacitors in the anode source. New types of capacitors have smaller dimensions. A small mains transformer with rectifier and filter capacitor can be placed above them. Another option is to place the switch source in the subchassis space next to the turbine.

The source can also be used for the soft-start circuit, resp. powering other accessories in the final stage. The air tunnel is actually a box into which the outlet from the turbine opens. The turbine blows air into the box. An increase in pressure causes the air to be forced out in the only possible direction – into the space of the tubes. The box must therefore ensure a tight connection with the turbine and the chassis, so that the air cannot escape elsewhere.

In amateur conditions, it is probably easiest to solder a cuprextite box. It is then screwed onto the chassis. A layer of silicone was applied to the contact surfaces. Silicone also seals the passage of the turbine - air tunnel.

It is advantageous to add turbine speed regulation. It's useless, so that in a state of rest (unlocked, just a glow) the turbine was running at full speed. It is possible to use various continuous electronic regulators with comparators. Another solution is bimetal switch. The contacts of the switch are bridged with a resistor of that value (cca.10ohm), so that the turbine rotates slowly. The switch is attached to the chassis. When the temperature rises, the switch closes and the turbine spins at full speed.

The efficiency of cooling in the described way is noticeably higher. In the case of PA with ceramic tubes, the use of a turbine is even a must.