Earthing the braid of co-ax on the plate.

[Mark]

I know Ken Fischer and subsequently Gerard Weber were fans of earthing the braid of co-ax on the plate/anode of a 12AX7. Though in Ken Fischer’s case he used this wiring of co-ax on the second triode which would result in only damage to the volume pot.

There are others who claim the voltage on the plate/anode can get on to your guitar and cause electrocution. I could be mistaken but I think that the voltage is more likely to damage your pickup rather than electrocute the player.

At this point in time I think I would prefer a small value cap from plate/anode to the cathode.

What is the general consensus out there?

IMG_0605.jpeg

[TUBEDUDE]

This practice is giving an A.C. path to ground through the power supply node.
It also acts as a capacitor, providing mild feedback, as the phase of the plate signal on the shield is 180° to the grid signal.
I trust teflon coax, but i think on the first stage, the risk would be too great.
I've also never needed to this technique combat noise, YMMV.

[bepone]

this is to tame high frequencies in high gain with 100-200pF value of capacitance in coax cable.. not technical method, nor smart idea, better to use some ceramic 100-200pF 1kV if needed

[martin manning]

It also acts as a capacitor, providing mild feedback, as the phase of the plate signal on the shield is 180° to the grid signal.

The small capacitance is in parallel with the internal Cg-a, so it gets multiplied by the gain (Miller effect). Even if the cable capacitance is only a few pF, it will cause a significant HF roll-off. The same effect can be had by placing a snubber cap from plate to cathode, say 200-400p.

[Mark]

Thanks for your replies. I thought much the same. I estimate the capacitance from a two inch piece of co-ax to be in the order of 10pF. Even with Miller’s effect I doubt it would be all that effective.

[martin manning]

I estimate the capacitance from a two inch piece of co-ax to be in the order of 10pF. Even with Miller’s effect I doubt it would be all that effective.

Let's say 2" of a low capacitance 20 pF/ft cable is used. That's 3p3. Here's the result for a typical Fender 12AX7 stage with 34k Rg, 100k Ra, 1k5 Rk, 22u Ck with and without 3p3 Cga added. The -3dB point drops from ~22 kHz to just over 7.5 kHz (not 5kHz).

Edit: Corrected an error in the model which raises the -3 dB point for the case with added Cga. Also note in both cases the input signal has 39k source impedance.

[Mark]

I estimate the capacitance from a two inch piece of co-ax to be in the order of 10pF. Even with Miller’s effect I doubt it would be all that effective.

Let's say 2" of a low capacitance 20 pF/ft cable is used. That's 3p3. Here's the result for a typical Fender 12AX7 stage with 34k Rg, 100k Ra, 1k5 Rk, 22u Ck with and without 3p3 Cga added. The -3dB point drops from ~22 kHz to just over 5 kHz.

Thanks for illustrating the difference between the two outputs. I’m surprised that 3p3F makes that much difference.

I know the top end of my hearing has rolled off over the years (doing a quick test of my hearing I hear 8Khz okay but 10Khz is quite difficult), and I guess it’s expected that I didn’t hear any difference when removing the co-ax cable.

[martin manning]

I know the top end of my hearing has rolled off over the years (doing a quick test of my hearing I hear 8Khz okay but 10Khz is quite difficult), and I guess it’s expected that I didn’t hear any difference when removing the co-ax cable.

See my correction above, which seems to make the simulation result line up with your observation a bit better (-3 dB at 7.5k). The effect will of course increase with longer cable runs, such as from the front of the chassis to the tube pins, where the length could be 7-8 in. Six inches, 10p added Cga, brings it down to 3.2 kHz.