Audio Projects – 50 W Stereo Amplifier

This page includes supporting information for the 50 W stereo amplifier project described Chapters 11 and 12 in The TAB Guide to Vacuum Tube Audio, along with commentary from the author. This page picks up where the book left off.

Shown on the right is the second-generation 40 W stereo amplifier described below.

50 W stereo amplifier

Note 1

The 50 W stereo power amplifier described in Chapter 12 of the book used hand-wired terminal strip construction. After the book went to press, the author built another version of the 50 W stereo power amplifier using the PWB designs described in Chapter 8 for the power supply and in Chapter 11 for the amplifier.

For the PWB installation using the board described in Chapter 11 and tube sockets specified in Table 11.7, it is necessary to use two types of standoffs to mount the sockets to the PWB. The 0.75-inch standoffs used in previous projects can be used for mounting V1 (the 7199 input/phase-splitter tube), but the 9-pin Novar sockets for the output tubes should be mounted on shorter standoffs (0.5-in). The same mounting considerations apply to the 5BC3 socket on the power supply PWB. When using the 0.5-inch standoffs on the 9-pin Novar sockets, it is necessary to pad the other standoffs to match the finished level of the Novar sockets. Two #4 flat washers are sufficient. It should be possible to add the washers and not exceed the minimum clearance limits of the PWB traces.

Construction of the power supply PWB is essentially identical to the design described in Chapter 8 of the book, with a few exceptions. Tubes V1 and V2 (5BC4 and 6080) were mounted directly on the chassis and connected with short pieces of hookup wire to the PWB socket pins. This was done in order to accommodate the large power transformer and choke used with the 50 W amplifier. The tubes will fit with the PWB design, but only with limited space between V1/V2 and the choke. To facilitate good cooling, a generous spacing margin was provided. For applications of the power supply PWB where all tube sockets are attached to the board (as designed), longer standoffs of approximately 1-inch are recommended to provide for adequate wiring space. Because the power supply circuits are essentially immune to noise concerns, the extra spacing has no material impact on performance of the circuits.

Note 2

The 50 W stereo power amplifier consumes all available space on the chassis. As a practical matter, this is about as big an amplifier as can built in a reasonably sized chassis. Beyond this point, dual mono block amplifiers may be the better approach. This amplifier is also quite heavy. The handles included on each side of the chassis are for more than just appearance. They are quite useful for moving the unit. The amplifier also generates a fair amount of heat owing to the beam power output tubes, rectifier, and series regulator tube. Normal convection cooling is adequate so long as airflow around the amplifier is not restricted by other devices, shelving, etc.

Because of the large number of interconnecting wires needed for this amplifier, expanded sleeving was used to organize the cabling. This tended to simplify cable routing and provided for a cleaner appearance. Wherever possible, cables were dressed either tight to the chassis, or well above the chassis so as to stay as far as possible away from signal-carrying lines and rest against the bottom cover plate, which further improved shielding. Because of the tight spacing of component elements, particularly on the power supply board, the order in which components and interconnected wiring were installed were important considerations.

Note 3

After building the 50 W stereo amplifier using PWBs, I had a difficult time completing the proof of performance while achieving the numbers expected. I was stumped by a distortion problem. The left channel amplifier performed very well—as expected—in all parameters (frequency response, noise, THD, and IMD). The right channel amplifier provided essentially identical performance in frequency response and noise, but THD and IMD were really bad. For example, at 20 kHz just below clipping, THD on the left channel was 1.6% but on the right channel it was over 6%. Both channels are identical; same design, same PWB layout, same parts. I tried switching tube sets, but the problem stayed with the sockets. In any audio amplifier, there are a few levers the builder can pull to make things look good, the easiest being the bias adjustment. But, no amount of bias setting on this amplifier would correct for 6% distortion (without the plates of the output tubes simply melting).

Under chassis view of 50 W stereo amplifier

The underside of the chassis is shown on the left. After checking solder connections, component values, and ground connections for the nth time, I noticed that the channels are not completely identical. Because of the way the circuit boards lay out, the connection from the RC feedback circuit to the board is short on one side (about 2 inches) and longer on the other side (about 6 inches). For the left channel (the “long” one) I used a short piece of shielded cable to connect to the board. On the right channel, I just installed a 2-inch piece of hookup wire. On a hunch (I had tried everything else), I replaced the 2-inch piece of hookup wire with a 6 inch length of shielded cable. And (you guessed it) distortion at 20 kHz just below clipping dropped from 6.5% to 1.5%. It seems there was just enough capacitance in the cable to influence the feedback circuit in a good way. I had built this amplifier before in a hand-wired (terminal strip) variation and did not have this problem. But, I had also placed the feedback RC components next to the speaker terminals and ran shielded cable (about 14 inches worth) to the input pentode cathode circuit. In the attached photo, you can see the RC components on each side of the chassis wired to terminal strips near the front panel.

So, the fix was to relocate the RC feedback components to the back panel of the chassis (next to the speaker terminals) and run shielded cable up to the front to the PWBs. Problem solved. Carrying the experiment a little further, I measured the capacitance of the shielded cable from the feedback circuit to the PWB terminals and then padded out the capacitor to compensate (about 30 pF). It didn't make any noticeable difference in measured performance, but I felt better. One more suggestion: keep the lengths of the shielded cable the same for each channel.

Note 4

In Chapter 11 of the book, the option was discussed of substituting a 6U8A pentode/triode for the 7199 originally specified. As discussed elsewhere on this site, the 7199 is becoming difficult to find (and expensive). The option of a substitute, therefore, has certain advantages. As documented in the book, the initial effort to substitute a 6U8A for the 7199 was not successful. Oscillations were observed at certain settings of the volume control, and as such use of the tube was not recommended. The implementation described in Chapter 11 was a hand-wired terminal strip design, and I wondered what the performance would be with the PWB-based layout. Returning to the 6U8A option, tests were made on the amplifier using the plug-in adapter detailed for the 25 W stereo amplifier. The results were quite positive.

The oscillations observed with the hand-wired implementation were not present with the 6U8A for the PWB implementation. Bench tests indicated essentially identical performance relative to the 7199. Tests were also made of the 6GH8A as a substitute for the 7199. Here again, bench tests showed no significant differences relative to the 7199, with the exception of THD at full power (25 W) where distortion at 10 kHz and to a lesser extent 20 kHz were about 1 percent higher than measured with either the 7199 or the 6U8A. Only a single 6GH8A was tried, so I cannot be certain this issue rests with the type of device or simply the device on hand.

Looking more closely at the oscillation issue initially observed, the first step was to examine the interelectrode capacitances of the two tubes, recorded in this table. Focusing on the pentode stage, the values for the tubes are closely matched. The oscillation did not occur at the minimum volume setting; not much of a surprise since the input is at near ground potential. The oscillation also did not occur at the maximum volume setting, arguably because that the capacitance of the cable to the back panel RCA input jack loaded the control grid of the pentode sufficiently to prevent the oscillation from starting up.

Testing this theory, with the volume control at an "oscillating" position, a 3.3 pF ceramic capacitor was connected between the 6U8A control grid pin and ground. The oscillation disappeared. It would appear that the adapter socket used previously exhibited just enough capacitance to prevent the oscillation. So, problem solved with a 3.3 pF capacitor between the pentode control grid and ground.

Looking at the oscillation problem from a different perspective, I was able to eliminate the oscillations and eliminate the shunt capacitor by changing the volume control potentiometer from 1 megohm to 100 kilohm. Since most modern input sources are well below 100 k, this is a simple and effective fix.

50 W stereo amplifier on bench

The photo on the left shows the 50 W amplifier on the bench undergoing performance testing. The 6U8A adapter socket is shown on the right.

The next step in the evaluation was listening tests with various types of program material. Results were mixed. A difference in the reproduced sound was observed, with a preference for the 7199. Having said that, the 6U8A worked well and reliably.

The 6U8A is a viable replacement for the 7199, although probably not the first choice. Critical listening tests were not conducted on the 6GH8A.

.6U8A tube adaptor

Note 5

The bill of materials for the 50 W stereo amplifier calls for an 8 A circuit breaker for CB1. As discussed in the Regulated Power Supply section, a lower-rated device is quite adequate for this application. Suggested part numbers for a 4 A and 5 A device are listed in the note. With an input line voltage of 117 V, the current draw is about 1.5 A with no signal input and no load connected to the auxiliary power supply socket.

Note 6

A new version of the 50 W amplifier has been completed, using the basic design and components described in Chapter 12 of the book. The major change in the new unit is the elimination of the power supply regulator circuit, which is not strictly necessary for operation of the amplifier. This change reduced the bill of materials cost by several hundred dollars, reduced the weight of the amplifier by about 8 pounds, and reduced the heat produced during operation.

Second generation 50 W amplifier

The photo on the left shows the completed amplifier. This implementation is described as a "40 W Stereo Amplifier" in recognition of the difficulty of getting a full 25 W power output from each channel when using the ultra-linear operating mode (output transformer screen taps) with acceptable distortion levels. At optimal bias settings, the amplifier is below 1% THD and IMD at 20 W output. Full-power frequency response is 10 Hz to 35 kHz (within 1 dB).

As mentioned, the power supply was modified from the previous version to eliminate the regulator circuit. Also, the 5BC3 rectifier tube was replaced with a 5U4, which is more readily available.

As with the other power amplifiers, this unit includes a decorative acrylic cover, which serves to protect the tubes (and protect users from the hot surfaces of the tubes).

Chassis view of second generation 50 W amplifier

A total of five printed wiring boards are used in this implementation. The new power supply board incorporates connection points for all primary circuits, which simplifies construction considerably. A new amplifier board (two are used—one for the right channel and one for the left channel) was designed for the 6U8A input tube. Two speaker terminal PWBs are used, one for each channel; they serve to terminate all secondary leads from the output transformers.

Expandable nylon sleeving is used for cable management, in particular the primary and secondary transformer leads. Quick-disconnect terminals are used on all boards. This permits a PWB to be removed for service without the necessity of unsoldering wires.

As with the previous implementation, an auxiliary power connector is provided to drive the stereo preamp described in Chapter 12 of the book.

Performance of the amplifier is quite good—essentially identical to the 50 W unit.

Note 7

While working to optimize the performance of the second generation amplifier, some adjustments were made to the feedback circuit that have proved to be helpful in reducing THD and IMD, particularly at high power levels. The original values for the resistor/capacitor feedback devices (R16/C6 in Figure 11.3 of the book) were 270 ohms and 0.01 microfarad, respectively. With the benefit of additional consideration, for circuits using the 6U8A tube for the input pentode/triode, it is suggested that R16 be changed to 200 ohms 0.5 W, and C6 be changed to 0.015 microfarad 200 V or so. The following parts are recommended:

• Vishay Specialty Capacitors, 0.015 microfarad, 200 WVDC, part #715P15352JD3, Allied stock #70079360
• Dale/Vishay, 200 Ohm 0.5 W, part #RN65D2000FB14, Allied stock #70200273

It should be noted that with these values, the amplifier sensitivity is reduced somewhat—on the order of 0.1 V rms.

Note 8

Based on the good experience with the automatic protection board for the Dorm Amp (see the New Projects Page) and the second-generation 20 W stereo amplifier, I retrofitted the 40 W stereo amplifier with the Auto-Protect board. As described on the New Projects page, this circuit board includes several system-protection features: automatic shutoff, over-temperature shutdown, and B+ under-voltage shutdown. Because of the physical layout of the 40 stereo amplifier, it was necessary to design a new circuit board in order to yield a PWB that would physically fit in the amplifier. The end result is shown below.

Chassis view with auto-protection option installed

The PWB measures 2.95-inches by 12-inches and mounts on the inside of the chassis adjacent to the left channel amplifier board. The PWB before installation is shown in the photos on the right. The finished unit with all connections made is shown on the left. Quick-disconnect terminals are used for all connections to simplify installation and maintenance.

Auto-protect board top

The automatic power-off feature can be enabled or disabled using a small switch placed behind the front panel on the right side of the chassis. In addition, an auto-off timer reset button is provided.

Auto-protect board foil side

Note 9

The cost of building any project is an important consideration for the audio enthusiast and hobbyist. The focus of the first-generation 50 W stereo amplifier was to optimize the various elements that went into the project. Cost was a secondary consideration. For the second-generation 40 W unit described above, an effort was made to accurately determine the bill of material (BOM) costs and the time needed to build the amplifier.

The total BOM (including shipping) for the second-generation 40 W stereo amplifier was $2,250. The major cost centers included:
• Electronic components = $540
• Transformers = $530
• Printed wiring boards (PWB) = $650
• Plexiglas cover = $125
• Tubes = $165
• Decals = $75
• Front panel = $100

Builders could eliminate the PWB cost by using hand-wired terminal strip construction. Note that three PWB designs are used in this amplifier—one for each channel, one for the power supply, and one for each speaker output circuit, for a total of five boards. Other potential areas of cost reductions include the back and top-side decals, and the Plexiglas cover.

The total time required to build the second-generation stereo amplifier—from ordering parts to completing performance measurements—was approximately 22 hours.

The BOM cost for the optional auto-protect board was $686. The cost breakdown was roughly 50% for parts and 50% for the PWB. Construction time was about 2 hours.

Note 10 (January 2017)

Continued work to improve the 40 W stereo amplifier has resulted in an updated second generation amplifier. Detailed step-by-step documentation is provided in a 202 page User and Assembly Manual that is available for download. If you are familiar with the Heathkit assembly manuals of the past, the approach taken with this new manual should seem comfortable. For those who want a printed version, it is available for purchase on Lulu—a web site for specialized, print-on-demand documents. (There is charge of $14.95 for the printed manual; see "40 W Stereo Power Amplifier: User and Assembly Manual." on the Lulu web site.)

In addition, a ZIP file is available for download that will assist in building this amplifier. The ZIP file contains the following individual files:

• Schematic diagrams of the amplifier, power supply, and auto-protect circuits as Acrobat (".pdf") files
• Bill of materials for the amplifier and auto-protect circuits as an Excel file
• Printed wiring board layout files for the amplifier, speaker terminal, power supply, and auto-protect boards as ".pcb" ( ExpressPCB) files
• Front panel layout as a ".fdp" ( Front Panel Express) file
• Chassis layout, bottom view, as an Acrobat file
• Chassis layout, top view, as an Acrobat file
• Drill pattern for the chassis as an Acrobat file
• Layout for the acrylic trim piece as an Acrobat file

Note that the files above are provided as-is. Every effort has been made to make sure they are complete and accurate, but no warranties are expressed or implied. Builders are encouraged to double-check the information contained in the above files prior to proceeding. For the front panel layout, users can customize the text as desired; e.g., "Built by John for Mary, January 2017." All files are provided for personal use only. No further distribution is allowed.

The current version of the 40 W amplifier builds upon the previous versions, adding some new features and simplifying construction. A description of the latest design can be found in the User and Assembly Manual, and for the sake of simplicity will not be repeated here. The estimated cost to build the amplifier is $4,250 for parts, and about 25 hours assembly time. As noted previously, cost reductions are possible, depending on the preferences of the builder. For example, the BOM cost can be cut almost in half by eliminating the add-on circuits (preamp power supply, auto-protect board, and auto-off board), which then places the BOM cost in the same range as detailed in Note 9 above.

Considering the price tag, this project may be outside the range of some builders. However, you may find that elements of the available downloads will be useful in other projects, or wish to adapt some of the circuits for your own projects. The available documents are, thus, offered as a starting point for future efforts.