The 6CY7 Amp

6CY7 Amp

An Exercise in Reuse

This amplifier came about due to an unexpected turn of events. Those with sharp eyes will recognize that the chassis for this amp is actually from the “Recovery” headphone amplifier. This is what happened. Shortly after I finished the Recovery amp, I started work on my 6V6 Lacewood amplifier. I used the Recovery amp for a while, however when the 6V6 Lacewood amp was completed, it quickly became my default amplifier for most of my music listening. By this time I already had the “Rebuild” headphone amplifier for my desk at work and I had largely stopped listening with headphones at home. This meant that the Recovery amp was put aside.

Then recently I decided I could not simply let this unit sit unused. So I began to look for a suitable project. At first I was thinking some type of preamp until I realized that the power transformer used for the Recovery amp could be reused for a small stereo amp. It is an older Hammond 269EX with a 190-0-190 @ 65mA high voltage winding and a 6.3v @ 2.5A filament winding. As it turned out, this is exactly what was needed to build a circuit using the 6CY7 which I had prototyped a few years before using that very transformer.

The Electrical Design

This amplifier is based on another of the vertical deflection dual triodes built for televisions; the 6CY7. This particular tube has a power section with a peak plate dissipation of 5.5 watts. This should translate to a little over one watt per channel. As I said, I had prototyped this amplifier a few years before while exploring alternate tubes. So I went back and revisited the load line designs I had done at that time. Here is the load line for the power section.

6CY7 Load Line Sec 2

This indicates a peak power into the output transformer of about 1.6 watts. The distortion also looks well controlled. I need about 29v-peak to drive this power stage. From a line level signal, that means that the driver needs a voltage gain of at least 23dBv. Given the nature of the 6CY7 section 1 triode, this should not be a problem. The published amplification factor is 68 so we should be able to get at least 2/3rds of that. The driver stage load line is shown below.

6CY7 Load Line Sec 1

This design shows that this stage will meet both the gain and voltage swing requirements without a problem. The interstage coupling capacitor is calculated as 0.01µf. Here is the final schematic for the amplifier.

6CY7 Schematic - amp as built

This is a very simple circuit that should be very easy to drive. With the volume control at max, it requires about 1.2v peak (or 0.85v-rms) to drive the amplifier to full power. Most small devices like iPods, phones, and portable CD players will easily drive this amplifier. Here is the power supply design for this amp.

6CY7 Schematic - ps as built

The power supply is a typical split rail design with separate filtering for left and right channels. This helps provide good channel separation and a nice wide sound stage.

The Build

The build for this design needed to proceed along a little bit different lines as the chassis was already built. The new design needed to make use of the current wood chassis layout as this was not going to change. A new top plate was in order because of the different number of tubes, the headphone jack went away and was replaced by a power indicator, and a new rear plate was made to accommodate the speaker binding posts. I also decided that I wanted a nice inductor up top to match the power transformer, so I ordered a Hammond 193H for the primary filter choke. I would be reusing the power transformer, IEC connector, fuse and holder, power switch, tube sockets, input jacks and volume control, 2H filter chokes, and the handles on the top plate.

This level or redesign meant that I only needed two new pieces of metal. These are shown in the picture below, just before painting.

6CY7 Metal

The only real challenge to this design was to decide how to mount the output transformers. I decided to use Edcor open frame XSE100-8-5K transformers and I needed to mount them both inside the chassis. This turned out t be not too difficult. After this decision was made, it was time to start wiring. I decided that I would do as much wiring as possible with the top plate removed from the chassis. This way there is plenty of room to work and seeing things is not a problem. Here is an in progress picture of the top plate wiring.

6CY7 wiring in progress

Here the individual components for each channel are mounted to the sockets and the primary portion of the power supply and filter are mostly wired. Wiring in this fashion is significantly easier than doing it with the plate installed on the chassis. Here is an inside shot of the final amp all wired up.

6CY7 Amp Inside

The two filter caps on the upper left are the individual 100µf capacitors for each B+ channel. They are suspended in mid air over (or under) the 2H chokes.


So this amp went through the typical test regiment I give most of my new builds. Here it is all hooked up and in progress. The signal generator is on the left, the dual channel AC voltmeter for checking response is in the middle, and the oscilloscope for checking waveforms is on the right. The unit right to the right of the amp is a dummy speaker load.

6CY7 Amp Testing

At full output the amp produces a clean 1.3 watts output per channel into 8Ω. The Edcor XSE output transformers are rated from 70Hz to 18kHz. This amplifier did a little better than this having a -3dB bandwidth from about 46Hz to about 30kHz. Here is a plot of the frequency response measured into an 8Ω load.

6CY7 Frequency Response


I still can’t get over how good and how loud this amp sounds. It handles everything from classical, to blues, to choral music equally well. And the sound stage is awesome. I will never go back to building common power supply filters for both channels ever again. I’ve had this amp plugged into my system on and off for the last couple of weeks and I am absolutely blown away by the quality of the sound coming from this little unit. I don’t know what it is, but these little triodes are just the best sounding tubes I have ever heard.

6CY7 Amp Left Profile

This is quickly becoming my favorite amp. I currently have a set of NOS Sylvania tubes in it, but I have also used a set of old used GE tubes which sound just as good. This is really a pretty inexpensive build. I highly recommend that anyone looking for a little iPod or computer amplifier give this one a try. You will NOT be disappointed!

25 thoughts on “The 6CY7 Amp

  1. Hi Matt. I am about to build this amp but have to use the Hammond 369ex transformer. I am new to amp building and was wondering about the center tap on the 6.3v secondary. Should I ignore this or connect it to ground or one of the cathodes and remove the 100ohm resistors.

    Thanks for all your work


    • Just omit the two 100Ω resistors on the filament lines and tie the center tap from the 6.3v winding directly to signal ground.

      The two 100Ω resistors form a virtual ground for the filament circuit if the transformer does not have a center tapped filament winding. I actually prefer using center tapped filament windings if they are available.

      • Thanks. Very interesting blog. I always wondered why feedback would really be that necessary with these SE amps with just two stages! But I can see how, as far as the perceived way that sound is heard, is very subjective and each person is different. Enjoy your write ups on each project…and how thorough you are with each one…also the very neat wiring and look of the finished project.

  2. Hi Matt, I’m a little confused your load line for section 2: The schematics indicate a B+ volts, but the load line shows a B+ of about 325 volts @ zero plate current and at zero plate voltage the plate current is 68mA which gives a B+ of 340 volts for a 5K load. Can you explain these discrepancies to me?

    • You are confusing the power stage load line with a restively loaded stage. The load line you are looking at in the power stage diagram is the AC load line. The DC load line is on the order of 200Ω. The static bias point is determined by the DC load, not the AC load.

      A value of 200Ω is essentially straight up and down at this scale. For this much difference between AC and DC load values (i.e. 25:1) you can safely ignore the DC resistance and put the bias point at the B+ value. You’ll notice that in the table embedded in the graph, the power stage bias point is at 188v/29mA.

      The required B+ is derived as 188V (i.e. the quiescent plate voltage) plus the cathode bias voltage (i.e. 29V) plus the dc drop in the output transformer primary. This is shown in the bottom line in the embedded table in the figure. The actual total is 188v+29v+5.8v=222.8v. Given the accuracy with which lines can be drawn and the fact that the plate characteristics are average values over a family of samples, I simply rounded to a B+ of 220V for the power stage. The error in doing this is only (2.8v/222.8v)*100% ≈ 1%.

      This process yields a nice round number for the B+ that doesn’t cause people to fret too much over the accuracy of the B+ voltage in their builds. If I had indicted a B+ voltage of 222.8v, some people would have fretted over a B+ voltage just a few volts higher or lower. The nice round value has the psychological benefit of putting people’s minds at ease.

      I hope this helps.

      • Matt,
        Thanks for the quick reply and yes it helps greatly; I concluded the same thing myself after thinking about it. All the references I have only mention resistively loaded valves so thats why I got confused. Another point of confusion is that in the embedded table you indicate B+ = 217V + 0.029A * rp but you left out rp in your calculation above which would add another 30.5 volts? Also you mention that it would take about 29 volts peak to drive the output to full power, how do you determine that? I’m thinking of building a similar amp with the 6DR7 as a design/learning experience so I’ve studdied your design extensively but I’m fairly new valve amps but I have built both your Lacewood V2 and the your 4S preamp an I like them a lot especially together.

        • First, 0.029A x 200Ω ≈ 5.8v. I’m not sure where you got 30.5v. You’ll have to explain your calculation. You’re correct that I did leave the output transformer primary DC resistance off the load line design. That’s because the design came first and I had not as yet chosen the exact output transformer. Once I did, I designed the power supply for the correct voltage.

          The maximum power stage AC drive signal is based on the bias voltage. This design is DC biased to 29V. If you apply an AC voltage to the grid with a peak value of 29v ( ≈ 20.5v RMS) then the grid will swing up to zero volts and down to -58v on each cycle. If the AC signal is any bigger, the grid will go positive and start to draw current, inducing lots of harmonic distortion. So max power is at full grid swing with the 29v peak sine wave.

          I’m glad you like the other projects. I try to present designs that are easy to build and yield nice results. I think both the Lacewood and the 4S meet those guid lines.

          • Matt, I thought rp referred to the plate resistance for the valve which is 1050 ohms listed in the embedded table. Thanks for the explaination on the required drive, simple and makes perfect sense. I appreciate all your help help and wonderful contributions to DIY valve audio.

  3. Hello,
    very interesting projects and website.
    I need the simplest hi-fi amp I can build: One stereo input, one stereo output (speakers: 8 ohm, 75W), volume control. For listening to music in my room, from my computer…
    Does this project will do the job? Is it a good start for someone who never have built an amp? I have built many electronic projects and repair an handful of guitar amp (solid state and tubes based) …

    • This amp’s output power is about 1.3W per channel. This is a great sounding amp but it needs to be matched to appropriate speakers. Do you know what is your speaker sensitivity?

      • Personal experience – one watt per channel into standard 86 dB speakers will easily fill a 20ft X 20ft living room with pleasantly loud sound. Average power level runs about 200 mW per channel for this listening level.
        It takes a lot less amplifier power than most people think to get good and loud.

  4. I built this amp with a solid state, RC filtered power supply. I have an output switch for speakers, or headphones through a resistor network for load matching and attenuation. This is my second tube amp build and it was on a tight budget. Sounds very good, lots of clarity and texture, as long as you respect the low power output. I am still chasing down a very quiet, but audible 60 Hz hum, but quite pleased with the project. Thanks for the design!

  5. I’m trying to understand what the 6.81K ohm and 33u Farad filter part on the power supply does. Doesn’t that drop the B+ voltage significantly below the 210V B+ denoted in the graph of section one of the 6cy7?


    • Actually this is an additional filter stage between the signal and power stages of each channel. This provides additional power supply ripple reduction for the high gain stage and helps prevent regenerative action between the stages; a characteristic necessary for the accurate reproduction of certain types of harmonic rich content. The filter stage provides approximately 44dB of isolation at the primary ripple frequency and considerably more at higher frequencies.

      As for the voltage drop, the driver stage only draws about 0.9mA so the total voltage drop due to the filter stage is only about 6.1v. This stage operates perfectly acceptable with this small reduction in B+.

      Does this answer your question?

    • I have a parts list but I haven’t priced it out. The big items are the chassis, transformers and chokes, and tubes. But the piece parts can add up if you’re not careful. Let me see if I can generate an estimate.

  6. Hello

    Thank you for all the work – the amp looks impressive – greatly appreciate your effort.

    I have a quick question, i am using B&W P7 Headphones and their impedance is 22ohms – a bit higher that the 8ohm this circuit is designed for. Would this still work for my headphones, and if not – what should be changed.

    Best regards.

    • You might be able to get by using the Edcor XSE10-16-5K output transformers. It’s still not a great match but some have reported that there is some wiggle room in this amp with respect to output loading. Kenny says he is running this amp with 4Ω speakers on the 8Ω outputs and he likes the sound. So 22Ω on 16Ω output transformers may be ok.

  7. I just finished building this amp and I am very happy with it. It outputs into a pair of ESS PS820 bookshelf speakers with good volume. Speaker impedance is 4 ohms but my ear can’t detect any distortion.

    I am a “digital guy” but this amp has won me over to the dark side (analog) and I loving it. Thanks for the great design.

  8. Hello, iam Santi from argentina. i m really interested in this project, i was looking to make an amp with tubes and this one seems to be a simple but awesom option.

    in argentina the line V is 220v 50 hz, how this affect on the power supply.?

    Do you have a BOM to make it a little more easy the parts search ?

    Cheers Santi

Leave a Reply

Your email address will not be published. Required fields are marked *