The Nuvistor 7586 did not operate at 9V with control grid at space potential

Using variation of the Pleiades V6 schematic...


(Always use a fuse with any voltage power source for safety).





When 9V were applied to the plate and the grid open, floating, or at space potemtial there was practically no anode current. The sound output was very low and buried in hiss.


This is because the grid is at cutoff potential for 9V at the anode.


It seems the missing electrons by cathode thermo-emiision make the cathode much more positive with respect to grid or the grid much more negative with respect to cathode.


By connecting a 20MΩ pull up or Pleiades bias resistor from plate to grid the anode current jumped to 140μA and the sound was a big as thunder. Very low noise too. The ticking clock at the far away room could be heared through both natural ears and headphones.


Signal path, setup:


Male singing voice at 12in - Sennheiser MD441 U3 - Pleiadss V6 with Altec 4722 input transformer - Sony TC-D5 Pro - Sennheiser HD580


When disconecting the pull up resistor there was almost no sound in 1 or 2 seconds. Apparently the grid was becoming egative by a time constant


When connecting the YFE voltmeter between ground (ie cathode) and grid in order to measure the self assumed grid bias, the amplifier operates.



This shows that the voltmeter although of specified 1GΩ input impedance pulls up the grid by the more positive cathode. The cathode is positive as it is emitting electrons.


The voltmeter showed more than -300mV. And in fact the grid is even more negative as the voltmeter itself changes or pulls up the bias.


An electrometer or a PH meter-voltmeter is needed which have a TΩ input impedance.



On the other hand the 1H4 electron tube operated at free grid or space potential when plate supply is 9V as the bias at grid increases to -600mV from -850mV (50mV measured between negative filament and grid, the average from 0 to 1.3V must be taken as cathode potential) when anode potential is applied.


On the other hand the 1H4 does not operate with 1.5V plate potential even with pull up bias but the 7586 does operate.


Some measurements and comments on the 7586 on Pleiades V6.


At 3.9V plate potential and 6MΩ pull up bias the anode current is 36μA. Bias is -112mV.


With 10Ω in series with heater (further reducing heater voltage from 3.9V) the anode current becomes 20μA and the sound is interesting. Grid bias becomes +100mV to 170mV. The microphone can be used at 6in as the bass is reduced.


At 9V plate potential anode current is 20μA using a 20MΩ pull up resistor from grid to (positive by missing emited electrons) cathode. Vg is -300mV. The sound is too thin especially if using a Pleiades (130Ω,40mH) filter. When connecting the 20MΩ to pull up anode, the anode current becomes 140μA and it is a full nice sound, very nice with the Pleiades filter to cut bass heaviness. A very nice sound with great definition, very well defined treble and world class studio quality sounding presence.


As mentioned above when grid is open it self biases to cutoff, by the boiling electrons near it? Reducing heater voltage by a 10Ω resistor made no difference (still no sound).


Connecting 20ΜΩ to plate produced a big sound like thunder. So the Pleiadss filter is needed.


Connecting instead 20MΩ to ground the sound is confused and proximity is not compensated in a euphonic way.


With 3.9V at plate and 20MΩ pull up, bias becomes -130mV, the sound is nice but perhaps not as low noise as when Ia is 140μA. And the proximity effect is not as well compensated. But 20MΩ seems to sound better that 6MΩ pull up with particular setup and mic.


A conclusion:


The 7586 electron tube self biases at less than -300mV. So with open grid or grid at space potential there is no sound when plate voltage is 9V. Connecting the voltmeter to measure bias was biasing the amplifier (pulling up the bias) and the amplifier was semi operating.


The 7586 at 140mA anode current with the aid of pull up bias sounded (at only 9V plate potential) as big as thunder. Low noise too.


A Pleiades (130Ω,40mH) had to be connected to the brilliant MD441 to reduce bass due to Fletcher-Munson, voice effort, proximity effect. The sound was clear, big, detailed.