This was a custom one-off, designed to the customer’s requirements.
This is a project that represented a change of direction from usual.
The design requirement for this one was simple. A very simple line-level preamp, with the capability to drive low impedance loads. No remote control, no phono input, no tone controls needed.
So far so good.
There was also a requirement to have a switchable +6dB gain for linelevel sources that may be a little low.
And the most unusual requirement. This needed to be a 3 channel amplifier. The customer was a SACD user (remember those?) and wanted to run the centre channel from the SACD through the preamp as well. So one of the inputs needed to be 3 channel, the rest 2-channel.
This last requirement forced a fundamental circuit topology design change. I had planned on a simple gain stage with negative feedback DC-coupled to a cathode follower for output. Instead this is what I built (click image to expand)
First the choice of tubes. If it was going to be a 2 channel preamp I’d have used a dual-triode 6N1P as the initial gain stage, and a dual-triode 6N6P for the cathode follower output. However the 3-channel aspect necessitated using the 6N6P as the gain stage and the Cathode Follower.
However the other problem was the volume control. There is no such thing as a 3-gang potentiometer that I could find, and I was not prepared to use an electronic volume control since this would be putting silicon in the signal path, plus the ones I found had unexpectedly high distortion characteristics.
In the end I found a 4 gang ALPS potentiometer, but it was only 50K – no 100K part could be sourced.
I don’t like 50K impedance on the input, in my listening experience having a high input impedance has a positive benefit on the sound. So I decided to put the volume control between stages.
R2 and R3 form the negative feedback, when parallelled they provide an amount of feedback that precisely matches the gain of the first stage, and when S1 is open, the gain is 6dB. This high NFB has the added benefit of greatly minimising distortion as well.
Due to this customer’s preference for the tube aesthetic, I decided to depart from my usual solid state rectification and go with a tube instead. Hence the several levels of step-down in the power supply, since I did not utilise a choke, which is the conventional approach with tube rectification.
The amplifier was assembled using 100% point-to-point wiring in an AliExpress aluminium case, suitably laser-engraved and CNC-milled. The customer had several strongly expressed preferences for the aesthetic and when the precise style of knob desired could not be sourced (except in quantities of thousands!) one was designed and 3D-printed.
About the choice of tubes
The tubes in this amp are Soviet-era military-spec New Old Stock (NOS) from 1976. The story goes that during the Cold War, the Russians persisted with vacuum tube technology because they are far more resistant to the EMP discharge that occurs with a nuclear detonation, Whether or not this is responsible for the flood of Soviet-era NOS tubes onto the market now is a cause for speculation. The quality, consistency, reliability, high availability and low cost of the Soviet tubes however is not. You can buy these 6N6P tubes for as little as $10 each on some sites. The closest Western equivalent – the ECC99 – is much dearer. You just need to be able to read Russian datasheets (with practice, this isn’t too difficult)
Around the back
Just what you’d expect, nothing more or less. I like the laser engraving, it came out as well as I’d hoped.
Once the construction was complete, it was time to test the circuit. This is where I got some surprises, because the results on the test bench were far better than I was expecting. The highlights were:
Distortion – measured at 0.1% at 2v RMS input
Freq Response – given that it’s AC-coupled on the output, this will depend on the connected load, so I torture-tested it with a 5K load and got a ±1dB response from 15Hz to around 60KHz. There is no output low-pass filtering given the stated intention to use SACD as a signal source
Hum and Noise – which was below my measurable limit of 1mV
Headroom – exceeded my ability to measure – my signal generator will give up to 20v p-p and at this setting I was faithfully getting 20v p-p out with no distortion. Then I hit the +6dB switch and the output jumped happily to 40v p-p with no increase in distortion. This represents a headroom of at least 32dB
I then moved to listening tests, in the listening room I connected it to the inputs of a 80W tube integrated amplifier which I used to simulate a power amplifier by turning its volume all the way up (this removing the volume control from the circuit). The speakers were silent at this point (As they should have been). For a signal source I used my DAC, a Topping D30. Streaming 16/44 FLAC files to it (CD quality) and hi-res from TIDAL resulted in a transcendant listening experience. The imaging was so three dimensional it was holographic. The speakers – floorstanding early 90s-era KEFs – were projecting a soundstage that had not only height and width but also depth.
I was firmly reminded what tube audio is all about while listening to this preamp.
It remains to be seen what sort of account it gives of itself when connected to a solid-state power amp as it will be with its new owner.