Tube / valve biasing for microphone applications

[Murdock]

I just looked at a few eastern tube mics from Gefell. Most of them use this weird arrangement with one transformer leg connected to the cathode and no bypass cap and often also no grid resistor...

The M92a for example:
http://mikrosammler.de/images/thumb/e/ef/M92a1-Sch.jpg/800px-M92a1-Sch.jpg

UM57 has the same biasing but with grid resistors.

But what I also found interesting was the CMV551:
http://mikrosammler.de/images/c/c9/CMV551-21.jpg
How is the capsule polarisation voltage derived?

I think they used these arrangements maybe because of supply shortage in the GDDR? High value resistors and small electrolyt caps were maybe scarce and expensive so they had to find another way to do it.

Altough Schoeps used similiar arrangements and they came from West Germany...

 

[luckythedog]

Interesting thanks.

....How is the capsule polarisation voltage derived?....

With a floating grid and Va about 75V (ref cathode), grid seems to adopt a potential near -1V ref the cathode. No matter what Vk is. And Ik adopts an operating point about 0.5mA. For an ECC81, assuming same as EC92. By choosing Rk we can set Vg and so provide polarisation voltage.

Rg, if present, might also contribute. Or it might not I suppose if Ig is zero !

T

 

[luckythedog]

Not sure I explained that too well. In these circuits if we float the grid, the grid will find an equlibrium potential about 1V below the cathode. Assuming the valve has found its happy operating point. We can arrange for the cathode to have whatever potential we choose. And the grid will track. So grid potential can provide polarisation voltage all by itself. No Rg needed.

How does capsule C get charged from cold? From grid current that arises until the valve is in equilibrium.

In principle - and it seems to hold in limited tests.

Hope this might be clearer!

T

 

[MagnetoSound]

How does capsule C get charged from cold? From grid current that arises until the valve is in equilibrium.

Actually, no. The clue is in that 60k resistor at the bottom of the cathode ladder. The polarisation voltage is determined by the plate current, since the plate resistor is effectively cut in half and the other half is placed behind the cathode. Together with the 50k plate resistor, this forms a voltage divider which sets the operating point of the entire grid/cathode arrangement some good way above 0v. Assuming ~1mA, that’s around 60v across 60k.

The 1k5 provides the grid bias differential. The 10uf to ground gives the cathode it’s AC path and decouples the cathode signal from the grid.

Coupling the transformer to the cathode provides AC feedback but also allows using an output cap of a lower voltage rating, which may have been an important space saver at the time of design.