connecting the shield in the EF80/EF800

Hey.  I'm building a mic with an EF800 tube.  I had someone give me the following advice, but I haven't been able to get them to clarify (they're probably busy, or just missed my response).  I figured maybe the people here could help.

Here was the observation:

For the EF800 they sound pretty cool, but currently the way you have it hooked up is without the shield. It works but you might end up with the strongest AM source in your audio signal if your PE (ground) of the outlet has more than a few Ohms in line.
There are two ways to hook up the EF800 Postive pin 4 and negative (0V) pin 5 bridged to the shield pin 6, or Positve pin 4 negative pin 5 and the shield pin6 hooked to the plate. Both ways sound slightly different so give it a try.

Click to expand...

I currently have pin 6 unconnected.  I'm not sure I exactly understand what is advised here.

Does this mean pin 6 (the shield) should be connected to either pin 5 (the filament) or to the plate?  Or should the shield be connected to ground?

Thanks. 

Here's the EF800 pinout

S/he is saying that you can either tie the shield to the cold heater pin (0v) or to the plate - and that either will work just as well but both ways will sound different.

You are building a very sensitive equipment and manufacturer of the tube provides you with a very nice internal electrostatic shield. What do you do?
You ground it of course. "Ground" can of course be at elevated DC potential, so if wire is very expensive (haha) you can use also cathode (if it there is no signal) or heater (you surely use DC heater) But obviously there is no reason to use anything else than the real pure 0V signal ground potential.

Yes, joining shield to anode increases stray capacitance so frequency response problems and/or oscillations are bound to happen.

So, is it better to connect pin 6 (shield) to pin 5 (0v)?  ???

Why does shield to anode increase stray capacitance?

mista min said:

So, is it better to connect pin 6 (shield) to pin 5 (0v)?  ???

Why does shield to anode increase stray capacitance?

Click to expand...

Because anode voltages can vary with signal, and you have five bits of metal conductor (cathode, anode, shield, screen grid, and grid) separated by an insulator (vacuum) = a bunch of capacitors.

They may be very small value capacitors (few pF), but they are "stray" capacitors nonetheless. Connecting anode to shield can cause additional feedback coupling of any signal on the anode back into the very high impedance input of the grid, 180 degrees out of phase and with high open loop gain, and hence produce HF oscillations, or frequency response problems.  "Miller Effect" ring a bell?

If you are using DC heaters, there's little or no AC voltage variation at that point in the circuit, so no stray capacitance coupling back to the input that is dependent on signal. Otherwise you can also use a separate connection to true signal ground for the shield, which would be my choice. Many tubes already have the shield tied to the cathode internally. EF80X gives you the freedom: anode, cathode, true signal ground, heater. You choose.

I connected it to ground, and it works great.  Thanks everyone.