|Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting|
This vacuum tube power amp design is the right answer for this author’s cherished speakers prior to engraving the front panel.
The idea to build a low-power, high-quality power amp came after I had built the small French towers. These two little towers were so good in relation to their cost, dimension, and look that they deserved to be driven by a dedicated hi-fi amp. But, what scheme? Which technology? The answer came rather quickly: A classic 1950s schematic with modern modifications, where justified, and, of course, vacuum tubes.
After reviewing several designs with different output power and features, I elected to start, as a base, with the one described in the “RCA receiving tube manual” page 692 (a push-pull capable of 15W). From that base I made the following, substantial modifications:
1. Employment of ELS4s as power tubes in an ultra/mean reduced load mode, instead of 6973s (lower cost and readily available).
2. No more than 8 - 10W crisp and clean.
3. A “no compromise” solid-state power supply.
4. Addition of a passive tone control circuit.
5. Selections of inputs through relays as opposed to a rotary switch.
6. Addition of a power indicator meter (just for fan and to give it an esoteric look).
The end result is shown in Photo 1.
The final version of the schematic is shown in Fig. 1. The input stage is a V1 cathode follower (CF) with no gain. It is located close to the input relay board on the back of the chassis and transfers selected signal to the treble and bass controls located at the front panel. This arrangement allows the signal to travel across the whole amplifier in a low impedance line, reducing potential hum pickup and high-frequency loss. In addition, the CF feeds the rec-out sockets without significant load to the circuit.
The input selection arrangement uses a four-position pushbutton switch to drive the proper relay. Selected input is loaded by the value of RL, which I chose to be 100k-ohm. From the tone controls, the signal goes to a V2 amplifier stage, which is necessary to restore the loss of the tone control circuits. These two stages, including attenuation introduced by the tone controls, have an overall gain of about 2.5.
The driver stage is a pentode/triode V3, 7199 and provides proper amplification and phase inversion to drive the push-pull stage with an undistorted 22Vpp signal. A 100k-ohm pot with a 10k loudness tap provides the volume control. The loudness can be inserted by means of a push-button switch which, in turn, activates a relay.
The cathode of the 7199 pentode section receives a 15dB negative feed back from the output transformer secondary winding. The amplifier has proven to be very stable without adding a capacitor in parallel to R19. I made a considerable variation (with respect to the RCA manual schematic) in the final stage.
The concept of this modification (as described in the 1959 Philips manual) is based on the fact that the average signal value of music and speech is consider ably lower than the maximum output signal measured with a constant sine-wave. Therefore, when the output stage is polarized via a high cathode resistor bypassed by a capacitor, the bias voltage remains relatively constant even if the signal peaks to its maximum level, thus emulating a fixed grid bias condition. The advantage of this solution, however, is that the cathode polarization produces a lower distortion with respect to a fixed bias scheme (Table 1), when the stage is driven toward high power level by peak of music or speech signals.
The stage employs a push-pull ELS4 with 440-ohm R24, R25, ½ P4 cathode resistors (390 + 50) per tube bypassed by a 100uF capacitor. This solution allows the tubes to work with a reduced load. The EL84 total idle current is, in fact, 25/27mA per tube instead of the usual 40mA, dissipating 30% less power and increasing the tube’s life considerably. The disadvantage, however (theoretically), is that the stage cannot be driven to full power using a continuous sine wave signal, the distortion will be excessive because it is designed to handle peaks of power.
A 100-ohm pot on the cathode circuit allows for a DC balance of EL84 tubes. Should you desire to establish standard operating conditions, replace the R24 - 25, 390 ohm resistors with 2 x 220-ohm. You may prefer to leave the bypass cap or take them off to add some sort of feed back.
From the 16-ohm tap, the signal through R26, R27 feeds a meter to monitor the output power. The resistor value is chosen to obtain a full scale reading at maximum output.
As I stated, the power supply (Photo 3) is a no-compromise circuit with CLC filters all over. There is absolutely no hum at the speakers even if you put your ear inside them. The power stages get their voltage through two separate CLC cells (Z2, Z3), while the input stage and driver sections get their power through a third CLC cell (Z1). A ±12V fixed regulator feeds the heater of the first two stages and the various LEDs used to backlight the pushbutton switches. There appears to be a lot of redundancy in this circuit, and in fact there is. How ever, this is so the final stage has plenty of current when needed.
The whole amp is built with point-to- point wiring with the exception of the relays input board and 12V cc filament supply, which were made from commercial kits. Photos 2 and 4 show the component layout and all the related wiring. Even if this is to be considered a small and relatively simple power amp, it must be built with great care. Good components throughout are essential to achieve quality results.
MECHANICS AND CONTROLS
The chassis is an old Wavetek Sweep-Marker no longer in use, modified as required to fit the new mechanics. The front panel accommodates the tone and volume controls, a pushbutton input selection, and the on-off switch, I’ve also added a power meter for reference; it provides an indication of relative power in use. The back panel (Photo 5) contains the power cord, fuse, speaker output, and all the RCA type input/output sockets.
I’ve already commented on the performance of the French speakers, but the results, when driven with this little 10W Gamma, are remarkable (Photo 6). The employment of the tone controls allows you to tailor the sound to your pleasure— and that is exactly how you should enjoy reproduced music.
I made performance measurements using a lower cathode resistor (220 + 50
ohm) with a continuous sine wave signal. At 8W, 1kHz, THD is well below 1%,
while the HP 3585A spectrum analyzer shows a rapid monotonic decrease of harmonic.
Below 100Hz distortion in
Under reduced load conditions, over all performance should be even better. You’ll like it.
PERFORMANCE MEASUREMENTS TEST SET