Continued from the previous today's post.
At your risk. All safety precautions should be followed when working with any voltage or current.
A Saratoga (from Collins equipment?) 5KΩ to 4Ω output transformer is connected to anode circuit of WE417A. Actually the transformer is a Chicago Transformer Div 677 0430-009, similar to Saratoga Industries TF1A13YY 5KΩ, 35mA to 600Ω or 4Ω.
Sony CD Walkman bridged line out is connected to floating grid through a coupling MKT? blue Philips capacitor of less than 820nF.
The linoleum Pleiades Auratone 5C type little cube monitoring speaker is connected to output transformer.
The Power of Goodbuy - Madonna is used for testing.
First test is at about 80V anode potential, 8mA anode current. Very clear sound. A bit boring? Not loud at all. It seems electron tube is musically under driven.
It is noticed while monitoring anode potential that it increased to about 200V. Disconnecting the CD line out it returned back to 100V. It seems the extremely high impedance of an open grid electron tube is a so sensitive condition that even the slightest leakage of a capacitor?, RF energy? can be read. Later addition: No, it seems this happens due to the grid rectification effect, electron tube may behave as an automatic gain control amplifying device. See below or other posts.
Sound as mentioned is not loud. A 1:10 input transformer is connected between CD player line out and electron tube, through the same coupling capacitor sending signal to control grid. There is severe distortion. A big sound with bass. Of course bias does change also by the grid rectification effect.
A Sowter transformer of 1:3 was connected. Less distortion but electron tube still overdriven.
10KΩ series resistor to power supply of 370V was tried. Anode current increased to 20mA. Anode potential is greater than 200V. Electron tube is somewhat exceeded from max plate dissipation power. Sound a bit louder. Still not as impressive as expected. The Pleiades 2N3055 (a fake 2N3055?) one transistor amplifier, so small and simple, is more preferred.
During these tests the grid potential was measured to go as low as -1V. Open grid biases itself from electron cloud and the grid rectification effect from input signal. The latter is a bit like a compressor side chain. It should happen on Neumann U47 too at high SPLs. Gain reduction by change of bias. Automatic level control.
Open grid seems not to work very well so far for this application. Problem seems to be the fact that the electron tube settles to a relative high anode voltage and relatively low anode current. Ie too less transconductance. Later addition: It sounded nice on following experiments.
Pleiades bias with a high MΩ resistor from anode to grid will be tried and as little anode voltage as possible for getting say 30mA anode current. How would it sound like?
It may seems strange to bias from anode to grid but it must be remembered that grid is already less than -700mV by cathode emitting electrons. So Pleiades bias is pulling up bias from a negative value to a less negative one.
Another observation was that when switching off the amplifier and back to on a few seconds later, sound was louder. Possibly due to the reduction of cathode temperature making grid bias less negative.
A potentiometer will also be added after the input transformer for volume control.
Pull up bias will also be tried by connecting grid to positive cathode through a less high megohm resistor.
Cathode resistor pull down bias will be tried too. Pull down bias is needed as in such cases grid is returned without a coupling capacitor to ground. Cathode is connected to ground through the Rk resistor. Since cathode is at 700mV positive with respect to grid by lack of electrons to electron cloud if Rk were 0Ω the grid bias would be forced to become from -700mV to 0 V possibly damaging the electron tube by excessive anode current. Cathode resistor creates such voltage drop as to bias the electron tube back negative. More or less negative than free grid should depend on value of cathode resistor Rk.
Later addition: after a few days open grid or grid at space potential has been retried with input circuit etc hard wired. Sound very nice. So far indistinguishable (or better?) from pull up bias from cathode to grid with 10MΩ resistor and cathode to ground. See forward in time post:
https://euroelectron.blogspot.com/2019/06/we417a-5842-electron-tube-with-pull-up.html
No comments:
Post a Comment