Custom Tube Consoles

 

Custom Tube Consoles came about because of a question asked on the rec.audio.tubes newsgroup.  Someone wanted to know if there was anyone out there who could design and build them a custom 6 into 2 all tube mixer.  A very lively discussion followed on both the specification and design of such a mixer. Many design ideas and topologies were discussed in general terms but in the spirit of friendly competition people mostly kept their design cards close to their chests.  Over time the discussion died down but it had got me thinking.

Tube mixing consoles have been around in broadcast stations and recording studios almost since tubes were invented. The early ones were relatively simple like this RCA broadcast console.

Its controls are really quite simple; little more than a rotary fader and a couple of switches per channel. As requirements became more sophisticated, more complex circuits with linear faders and equalisation were introduced. For some time, individual pre-amplifiers had been built into modular enclosures such as the Edyne example below and of course the well known V72.

Note the ventilation holes at the top and bottom. Soon it became economic to build each channel in its own module. Here is an example from a EAB mixer (note the transistor in the bottom right hand corner).

As well as a linear fader and switched microphone gain (bottom left rotary switch) this channel provides HF and LF boost/cut and a switchable high pass filter.. The tubes are E283CC types which are high reliability versions of the ECC83/12AX7. Notice how the bottom is completely open.

Next is a Telefunken channel module.

Note the input transformer at the bottom right and the phase change switch above it. The first stage amplifier appears to be to the left of the input transformer and there seems to be no way to vary its gain. To the left of the input tube is the output transformer and to the left of that the output tube. The third tube is for the equalisation circuit. The linear fader is a strange shape because it uses a rotary potentiometer operated via a flexible steel band and some gears. In addition to HF and LF boost/cut, this module has mid cut/boost at selectable frequencies. You can see ventilation holes below the input tube and to the left of the EQ and output tubes.

These modules gave a clear idea of the the construction techniques used and in particular how cramped the wiring is. Note also that the 12A?7 series of tubes seems to be in common use.

Now, some months earlier on rec.audo.tubes there had been another lively discussion about negative feedback (NFB) in tube circuits as a result of which I wrote a short  introductory paper on the subject to help designers use it effectively. Privately I had concluded that NFB around more than a single stage was fraught with compromises in tube circuits not least because it is very difficult to close the loop at dc. Unfortunately, in researching microphone pre-amplifier designs I soon discovered they all used twin triode topologies (12AX7 and friends) with NFB around two or more stages. There was no way I was going to be able to base a modern design on these old topologies.

I then came across some work by Morgan Jones into the intrinsic distortion in tubes. He used a topology that effectively removed all circuit based distortion mechanisms allowing the intrinsic distortion in the tube to be measured. To cut a long story short, he found the 12A?7 series of tubes had relatively high levels of intrinsic distortion; no wonder they had to be used in circuits with global negative feedback. However, one tube shone out as having a surprisingly low intrinsic distortion even at high operating signal levels and its distortion was primarily second harmonic with the higher harmonics falling away nicely.  This tube was the 6SN7. I wondered if such a tube could be used in a microphone pre-amplifier without the need for global NFB.

Unfortunately (again), the 6SN7 is a largish tube housed in an octal base; not really appropriate for a compact mixer.  After some further research I discovered that a 9 pin glass base version had been developed; it is the 6CG7/6FQ7.  This tube is compact and is still in current production. If it was as good as the 6SN7 there might be a route to an all tube mixer without global NFB.  Morgan Jones circuit is not really suitable for a practical circuit so a common variation of it, the mu follower, was tested instead. It turns out that a 6SN7 in a mu follower topology has almost as low a distortion as in Morgan Jones test circuit. And best of all, after testing about 50 of them, I found the 6CG7/6FQ7 is just as good. So here was the basis of an all tube mixer without global NFB.

There’s a lot more to mixing console design that having a decent gain block, so after a lot more work a basic console design was created which is now being manufactured. In the mean time I thought it was important to verify that sonic integrity of the basic channel, so a single stand alone channel with its own power supply was built and sent the client for him to test.

 

Fortunately, he loves it!