Most of you probably expected the 300B to be this months tube. But after all the attention it got on my blog recently, I decided to set a counterpoint by selecting a totally different tube for October. A Compactron Beam Power Pentode, the 6GE5.
As fond as I am of directly heated triodes, I also like to work with 'sleepers'. These are tubes which are largely ignored by tube amplifier builders but which are very usable in audio applications. One of them is the 6CB5A, also a beam power pentode. Many years ago I proposed the use of triode strapped 6CB5As as low cost 300B alternative. Quite a few 6CB5A amps have been built around the globe since then.
In the 1960ies General Electric introduced the compactron tubes with 12 pin bases. This class of tubes was intended to rival the transistor which started to replace vacuum tubes. With the 12 pins more systems could be integrated into a single bulb. There have been triple triodes, quad diodes, twin pentodes and many combinations. The 6GE5 is a single beam power pentode, thus it does not need all the 12 pins. Yet many such types with single systems have also been put on 12 pin compactron bases to enable TV sets which only need a single type of socket. I already covered some 12 pin compactron types, the 6U10, 6HS5, 6HV5A and 6CG3. There are many more hidden gems in this series of tubes and I intend to present some of them during the next year.
So why the 6GE5? With the 6CB5A there is already a good alternative for the 300B so maybe something smaller would be nice. Also many people don't like tubes with top caps, so I picked one without. The 6GE5 was intended as horizontal deflection amplifier tube in TV sets, a so called 'sweep tube'. It has a lower plate dissipation than the 6CB5A, only 17.5W which also comes with a lower heater current of 1.2A, less than half that of the 6CB5A. The 6GE5 data sheet lists the triode amplification factor as 4.4. Plate dissipation and amplification factor are remarkably close to the 2A3, so maybe the 6GE5 could turn out to be a nice low cost alternative to the 2A3 and get us some 3.5W in single ended configuration? To find this out we need to get the values for plate resistance, transconductance in triode mode and of course we need to see the plate curves. The data sheet does not show the triode curves, so I measured them with a curve tracer:
This is about as good as you can expect from an indirectly heated tube. Quite good linearity in the mid range with the spacings of the pate curves getting a bit larger between the 0V and 10V grid step. How about the transconductance? This can be derived from the plate curves, but I took the easy way and just measured it on the tube tester. With 300V on the plate and 55mA current I got 5300 micromohs (5.3mS) almost identical to the 2A3s 5250. The plate resistance is easily calculated as amplification factor divided by transconductance which gives 830 Ohms. Again close to the 2A3s 800.
Before we go more into detail about the use of the 6GE3 in a single ended amplifier, lets have a closer look at the tube. So far I only have 2 of them so I cannot show many photos of different brands. The ones I have are Made by General Electric:
The 6GE5 has a nice shape with it's rather short but and bottle.
A view from the top showing the grid cooling fin:
Removing the heater:
The heater wire:
A close up:
The beam plate with the cathode still inside, the grids are already removed:
This photo shows how the beam plate is internally connected to the cathode:
A close up of the cathode:
The beam forming plate:
All the electrodes side by side, heater, cathode, control grid, screen grid, beam plate and plate:
The control grid is gold plated and the screen grid has some kind of carbon treatment to increase it's dissipation capability. A close up of the grids:
Another view of the control grid:
So how to use this tube in a single ended audio amp? From the values discussed above it is quite obvious that the tube could be used with typical 2A3 operating points. But maybe this can be optimised a bit. I used the set of plate curves from the curve tracer and extrapolated some additional grid lines since the tracer only works up to 400V :