Navigation: DIY Audio Projects / DIY Vacuum Tube Projects / Poddwatt: Class-A Stereo Push-Pull EL84 (6BQ5) Vacuum Tube Amplifier

5751 SRPP / EL84 (6BQ5) Push-Pull (class-A) Tube Amp

 Bruce Heran   USA Flag   To email Bruce, type out the email address.

Poddwatt: 5751 SRPP / Push-Pull EL84 (6BQ5) Tube Stereo Amplifier

This project is related to the other tube amplifier projects I have posted. It has more than a casual similarity to the first Push-Pull EL84 (6BQ5) Oddwatt project from nearly two years ago. Both tube amplifier projects use EL84 / 6BQ5 tubes and the earlier one used a ECC802S driver while this one uses 5751 driver tubes. The basic circuits are rather similar, but the new project has a number of differences. Most notably it is easier and less costly to build. The rather large output transformers used in the original project have been replaced by ones that are considerably less weighty and less costly. This alone makes for a $40 to $50 savings. A number of circuit refinements have been included as well. One area of particular note is the higher gain of this new tube amp. This was done to allow the direct connection (via headphone jack) of MP3 players to the tube amp. The earlier project with the ECC802S driver tube did not have sufficient gain for this. The end result is a great DIY project for those of you who need 5 to 7 watts of high quality sound. The amplifier is quite clean from 25 Hz to past 20kHz (see the measurements in Tables 1 and 2 below). There is plenty of bottom end, just not as much as the bigger amplifiers like the KT88 Odd Blocks which use much larger audio output transformers. I was very surprised how well this amplifier sounds. If you don't exceed the power limits it will provide sound quality that is easy to listen to with great detail and clarity.

5751 SRPP / EL84 (6BQ5) Push-Pull DIY Class-A Tube Amplifier

Photograph 1: Poddwatt Class-A Stereo EL84 (6BQ5) Vacuum Tube Amplifier

Warning: This project uses high voltages. Contact with such voltages can cause injury or death. If you are not familiar with high voltages and how to safely build high voltage equipment, then please do not attempt this project. I want everyone to enjoy music and not have anyone harmed in the process.


Power Supply

The amplifier uses a single power supply with four filter sections. The first three sections are in common with both channels. The final filter section is split into two halves so that each SRPP has its own filter. I do this so that there are minimal interactions between the two channels. The four sections start with fast recovery solid state rectifiers and a capacitive input. Three of the four sections are bypassed with polypropylene caps. The end result of this arrangement is an amplifier that is extremely quiet. I hate hum and noise and take special care to minimize them. I want to hear music, not buzz.

EL84 (6BQ5) Push-Pull Tube Amp Power Supply Schematic

Figure 1: EL84 (6BQ5) Poddwatt Power Supply Schematic (October 2009)


5751 SRPP Driver Stage

The driver stage is a fairly straight forward SRPP design using both sections of a new production Sovtek 5751 vacuum tube. For those who are not familiar with this tube, it is similar to a 12AX7. The 5751 is pin compatible with the 12AX7 but not directly interchangeable as some parameters are different (5751 gain is about 30% lower than the 12AX7). I tried other brands of 5751 vacuum tubes and the Sovtek brand had lower distortion than the others. The 5751 JAN Philips tubes used in the larger amps did not perform as well in these amps. The bias point for the SRPP is approximately 1.25 mA. This is sufficient to ensure linear response. SRPP drivers are particularly good as power tube drivers. I chose one here for simplicity, gain, linearity, drive capability, and power supply noise rejection. The gain of the 5751 SRPP driver stage is about 32 times or 30dB.

5751 SRPP / EL84 (6BQ5) Tube Amp Schematic

Figure 2: 5751 SRPP / EL84 (6BQ5) Tube Amplifier Schematic (October 2009)


EL84 (6BQ5) SIPP Output Stage

The output stage is a self-inverting push pull (SIPP) design after the Compact Hi-Fi Power Amplifier by Melvin Leibowitz. The basic design has been around for a long time (50 years or so). The problem with early applications was that it was difficult to force it to operate in a linear manner. It can only function in a class-A mode with both tubes always conducting. The reason this circuit is attractive is that it eliminates the need for a phase splitter stage. It requires only a single-ended driver. The advent of solid state electronics provided a solution to the problem. A simple application of a common LM317 IC voltage regulator converts it to a very accurate constant current source (CCS). Using a CCS in the cathode circuit of the SIPP stage forces it to operate in class-A. What was a sort of a mediocre output stage thus becomes highly accurate and quite excellent. The voltage regulators can be somewhat sensetive in high voltage circuits so be sure to review my notes regarding The Care and Feeding of LM317 and LR8 Integrated Circuit Regulators, Particularly in Valve Circuits. At first glance it would appear that the output stage would not function well if at all. But if you study the circuit you can follow the signal paths. The first clue is that the current though the tube pair is constant. So if one tube gets a signal and tries to conduct more, the other must conduct less. The second clue is in how tubes are biased. This is the control voltage that regulates the current flow through the tube. It is determined by the difference between the grid and cathode. The more negative the grid is with respect to the cathode, the less current the tube will try to conduct. So clue three is the voltage across the CCS is not constant. When you put it all together here's what you get. If the driven tube gets a positive signal, it will try to conduct more than the idle level. When this happens, the cathode voltage goes down. This in turn increases the difference in voltage between the grid and cathode of the other tube. Remember that this tube's grid is essentially at ground potential and can not change. It is a see saw arrangement. If the driven tube gets a negative signal, it will try to turn off. This will then raise the voltage across the CCS and in doing so, make the second tube conduct more. The sum of the currents is always constant. Now some of you may be asking how can the circuit properly bias the tubes so they won't fry? It really is easy. The CCS provides the bias much like a cathode resistor would. The only unusual part of the circuit remaining is the tube balance control in the cathode circuit. It seems that regardless of how well matched the tubes are, they are not perfect. The 25 ohm variable resistor allows for a slight adjustment of the idle bias on the tubes. It provides a range of about 2-3 volts. This is usually enough to balance any typical pair of tubes. Badly mismatched ones and ones of different brands should not be used. It seems that accurate tube balance is key to the high level of performance of the design. Deviations as much as 2-3 mA between the tubes seem to have little effect. The accurate balance and subsequent quality output makes sense in a number of ways. First, it balances the dc current in the output transformer. Second, it allows both tubes to operate in the same portion of their accepted range. The closer they work together the better the cancellation effects on harmonic distortion will be. I have found that this particular circuit with the parts specified has less than 2% THD without tweaking or the use of negative feedback (NFB). I built in a small amount of NFB for two reasons. The first is that it reduces the mid band THD to the 0.1% range for small signals and under 0.5% for average levels. A more significant reason is that without NFB the amplifier will deliver a substantial amount of power out past 50 kHz. That is in the range of normal transformer resonance points and could result in oscillations and instability, most notably ringing. There are other ways to prevent this, (like larger grid stopper resistance), but for me the NFB works. New production Electro-Harmonix EL84EH vacuum tubes work well in this amplifier.

If the amplifier doesn't sound right or oscillates, then the NFB connections are probably backwards and applying positive feedback. The simplest way to resolve the situation is to reverse the speaker connections inside the amp. Just switch the ground and output. The problem can occur because of the connections to the primary side. Depending on which tube is connected to which primary winding, it will alter the phase of the output.


Tube Amplifier Construction

Construction of the amplifier is not complicated. I used an 8 by 12 inch (200 by 300 mm) chassis. The connections to the tube sockets are relatively few and should be completed first. Since I used 12VDC for heaters, I made two identical 6 volt sections. One for each channel and then wired them in series. I attached the heater positive voltage bias to the junction of the two groups. This is a critical aspect of the design. A disadvantage of the SRPP is that the plate of one section is connected to the cathode of the other. This results in a B+ voltage at the junction of about 1/2 the B+ voltage applied to the upper triode. One often overlooked specification of tubes is the breakdown voltage rating between the heaters and cathodes. In these triodes (and many others) the rating is around 100 volts. To protect the tubes two things must take place. First the heater circuit can not be grounded. Second a positive voltage is applied (about 80 in this case) to the heater circuit. There is virtually no current flow so large value resistors are used as a voltage divider tap on the B+ filter string. I have found that there can be noise in the heater circuits and a 1uF poly cap is used at the junction of the voltage divider to remove it. I discovered personally about 2 years ago what happens if you do not bias the heaters as covered above. Sparks, arcs and zap dead amp. Kind of neat if you go for those things. Me, I prefer to hear the music. The most complicated part of the amplifier is the power supply. With multiple sections and DC heaters, it will occupy a good chunk of the chassis. Do not skimp on power supply components. The power supply is half of the signal chain. Where do you suppose the power that goes to the speakers comes from? If the power supply is skimpy, or poorly designed it will have a drastic effect on the sound. Do it right, use good parts.

I use tinned copper wire in the size range of 12 to 22 gauge (see the American Wire Gauge (AWG) Table for reference) depending on the length and purpose of the wire. I use standard 60/40 solder. Yes I know it is not good for the environment, but I don't expect that the small amount I use will cause any significant damage. You can use silver wire and solder if you choose. I have not been able to determine that they make enough of a difference for me to hear. Plus my observations is that not all DIYers do a good job soldering with silver solder because of the higher temperature needed. Cold solder joints are bummers. Now let's get back to the wires. Route signal wires away from power wires and heater wires when possible. Cross such wires at right angles. I like to keep most wiring close to the chassis and run those that can be together in the corners. Use quality shielded wire for input connections. In this particular build I used a ground buss down the center of the amp. Several components are attached to it. The center of it is where the power supply and signal ground come together. It isn't a true star pattern, but is suitable for this design. Watch out for alternate ground paths as they will form loops and cause hum. The chassis should not be directly electrically connected to the signal nor power (B+) grounds. Use a X2 type cap (Rifa is one brand) between the chassis and signal / B+ grounds. The chassis should be connected to the third wire ground at the power entry filter (this can vary with some countries electrical codes so be sure to check your local code). This makes the chassis a safety device and if the amp fails it will blow the fuse or circuit breaker. It will also shield against ambient noise.

EL84 (6BQ5) Tube Amplifier Power Supply

Photograph 2: Tube Amplifier Power Supply

I tried to avoid using parts that are hard to obtain and with the exception of the power transformer the rest should be easy to get. The 25-Ohm 3-Watt rheostat (variable resistor) can be difficult to locate - it is available in the United States of America from Radio Shack (Catalog #: 271-0265, US$3.99 each), in Canada from The Source (Catalog Number: 2710265, CDN$6.99) and in Australia from Dick Smith Electronics (Product Code: R6905, AU$8.98). The prices quoted are from October 2009. If you know of European sources for these variable resistors lets us know and we will note them. The transformer is an OEM from Edcor for the Oddwatt tube amplifier kits, but other transformers with appropriate ratings (180V-0-180V @ 100mA + 12.6V @ 4A) will also work. Alternatively one could use a full wave bridge and an Edcor XPWR058 which is rated at 180V-0 @ 200mA + 12.6V (6.3-0-6.3) @ 4A ($47 from Edcor). You can try AC heaters and the results are likely to be good, but like I said at the beginning, I hate hum and noise and AC heaters are unlikely to produce an amp this quiet. There is no reason that you can't use 6VDC on the heaters either. It is just harder to get the amperage needed at 6 volts than it is at 12. If you want you can use a SMPS to get the heater power (like in the first EL84 Oddwatt tube amp and the KT88 Oddblock tube amps). They work well and are inexpensive. All other parts can be obtained from sources like Parts Express, MCM Electronics, and Tube Depot. If you can wait a few months and so desire, I plan to make a full kit of parts including a chassis available as well at OddwattAudio.com. No sales pitch intended, just an alternative.

5751 SRPP / EL84 (6BQ5) Push-Pull DIY Class-A Tube Amp

Photograph 3: Underside View of Tube Amplifier

For additional tube amplifier design and construction tips, see my design and construction tips and suggestions for vacuum tube amplifiers. Also, I have previously posted some suggestions for a tube amplifier wiring color code that you can use during construction to help make debugging easier. Finally, please be sure to also see my tips about grounding and shielding for your DIY audio projects.


Tube Amp Testing and Measurements

The amplifier sounds and tests well. With a single output tap that is nominally 6 ohms, the amps work pretty much the same on either 4 or 8 ohm speakers. Based on the output voltage the amps will deliver around 7.5 watts RMS into 4 ohms and about 5 watts into 8 ohms. The actual volume difference between the two levels is small. The amplifier measurements are in Tables 1 and 2 below. There are no real areas of concern with the performance. The primary limiting factor is the audio output transformers. They were chosen as a good blend of performance and cost. They perform well in the low bass region to about 25 Hz and then the distortion starts to rise a bit below that frequency. On the top end they are excellent and the amp can reach more than double the top of the audio band (~40kHz). Since most speaker systems don't respond much below 40 Hz, the bass distortion may be a moot point. My main systems are pair of upgraded Altec Lansing A7-500 that do go down cleanly to 20 Hz. I could not detect the distortion when played though these using music that had significant content below 30 Hz.

The tube compliment during testing was Sovtek 5751driver tubes and Electro-Harmonix EL84EH output tubes. The test setup consisted of:


 Frequency (Hz)   Output   Distortion (%) 
25
30
40
50
100
200
1000
2000
10000
20000
 1VRMS / 0.125W 
1VRMS / 0.125W
1VRMS / 0.125W
1VRMS / 0.125W
1VRMS / 0.125W
1VRMS / 0.125W
1VRMS / 0.125W
1VRMS / 0.125W
1VRMS / 0.125W
1VRMS / 0.125W
0.50
0.50
0.33
0.18
0.11
0.08
0.06
0.09
0.06
0.07
25
30
40
50
100
200
1000
2000
10000
20000
 2.83VRMS / 1W 
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
1.25
0.65
0.36
0.30
0.18
0.20
0.18
0.19
0.20
0.22

Table 1: Distortion Performance of Poddwatt Amplifier


 Frequency (Hz)   Output   Level (dB) 
10.9
17.9
25
1000
20000
33600
45800
78700
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
2.83VRMS / 1W
-1.0 (see Note 1)
-0.5 (see Note 1)
-0.1
 0 (Reference Level) 
-0.1
-0.5
-1.0
-3.0

Table 2: Frequency Response relative to 1000Hz

Note 1: Output here is distorted. Smooths out by 25Hz.

The maximum output is 5.8VRMS into both 4 and 8 ohms - about 4W into 8 ohms and 7W into 4 ohms.


Listening Comments

I like lots of different types of music. Everything from hard rock to classical. Three areas are my favorites though, string instruments, female vocals and stuff with a solid bottom end. To listen to this amp, I just swapped it into my main system. That is with the previously mentioned Altec Lansing A7-500, a highly modified Velleman tube preamp, a modified Marantz CD player, and a Dual 701 direct drive turntable with Grado Silver Prestige cartridge. This was done in the listening room I designed and built about 18 months ago (BTW a great thing to do if you can swing it). Now how did it sound? Well it isnít going to make me get rid of my tube mono-blocks (KT88 version of the Odd Blocks), but it sounded awfully good. Very clean, detailed mids and highs. It has a great sound stage. Bass was solid and had a lot of muscle, any low end distortion was not obvious. Some examples: Goldfrapp - Number 1, I didn't think it could make that much sound and reach that low. Wow! Madonna - Vogue, the holographics were first rate. Celtic Ladies - Danny Boy, she was right in the room, well centered. Bronn Journey - several tracks, nice bite to the harp strings. Stevie Nicks / Fleetwood Mac - several tracks, raised the neck hairs. So how would I rate the amp? Well I'll put it this way, it will be my back up amp in the main system for those occasions when I don't want to hear things at huge levels or heat the house (the Oddblocks each dissipate over 140 watts just idling).

Good listening, Bruce


 UPDATE  - March 2011
The PoddWatt EL84 / 6BQ5 amplifier is now avalable as a kit from OddWatt Audio. The PoddWatt Series is an integrated stereo amplifier kit that comes finished chassis. For more information, see the PoddWatt Series tube amplifier kits.

 UPDATE  - 30 September 2012
Bruce has updated the OddWatt self-inverting push-pull EL84 amplifiers once again. The latest version is constructed as mono blocks and again uses a 5751 SRPP driver stage. The mono block amplifers deliver more power and improved performance. For full details see the Mini Block ultra-linear class-A push-pull EL84 valve amplifier project page.

 UPDATE  - 8 October 2012
The Poddwatt is one of my favorite projects and is a serious over achiever. Since the project was first posted I have made a few updates that raise it to even higher levels. The changes involve custom designed output transformers (Edcor EMO750, now available on an individual basis), a custom wound power transformer (EMO719 from Edcor also available for individual sale) improved power supply, changes that improve linearity and slight increases in output power. There are revisions to the grid circuits of the power tubes and the use of LR8 solid state regulators in the driver stage and in the heater reference circuit. The heaters are now powered by AC although DC could be used. I found no difference in the level of hum and noise between the two types. A number of small components have been changed as well. The result is an amplifier that has exceptional performance. I have been told a number of times it sounds more like a SET than many SETs do. I personally can't comment on that as I don't design (yet) or own any SETs. It is quiet, wideband and low distortion. It is also uncomplicated to build and there are now no difficult to obtain components. Earlier Poddwatts can be retrofitted with the latest modifications without excessive difficulty. Even the new output transformers are close in foot print to the originals and usually can be mounted in place of the originals. The attached schematics show the changes through October 2012.

5751 SRPP / EL84 (6BQ5) Valve Amplifier Schematic

Figure 3: PoddWatt 5751 SRPP / EL84 (6BQ5) Valve Amplifier Schematic (October 2012)

EL84 (6BQ5) Push-Pull Valve Amplifier Power Supply Schematic

Figure 4: Poddwatt Power Supply Schematic (October 2012)

If a pair of mono block amplifiers is more suited to your system then my project the Mini Block ultra-linear class-A push-pull EL84 valve amplifiers are the way to go. There are similar to a half of a Poddwatt. The use of a larger power transformer in them provides a slight increase in output. The increase is from about 5 watts RMS in a Poddwatt to 5.8 watts RMS in the mono blocks. The difference is not audible, but the mono blocks seem to sound more powerful than they really are. They drive my Martin Logan Vista ESL speakers to quite satisfactory levels. They will not replace the KT88 and KT120 mono blocks I have when I really get into seriously loud music, but for jazz, stringed instruments and the kind of listening I do quite often they are excellent.

For those of you interested in things like tube life and long term performance I finally retired the original Poddwatt shown in the project. It ran for about 8.5 hours, five days a week for nearly four years in my office. I figure that was somewhat over 2200 hours. The original tubes are now in the updated Poddwatt I use as a demonstrator at audio shows. I expect them to go at least that many more hours and in all probability reach 6000 to 8000 hours if I have it that long. Conservative designs really pay off in this area.

The latest information about the PoddWatt stereo amplifier project is available on the PoddWatt EL84 Push-Pull Stereo Tube Amp - Support Thread in the DIY Audio Projects Forum. Feel free to use the support thread to ask questions or share your comments about this project.