G7 voltages, operatin point, setting cathode resistor for bias without charts ?

[ln76d]

I want to set proper bias for E80F tube in my G7.
Unfortunately there is no charts for E80F in triode mode like for EF86...

How?

E80F is not related with EF80, it's similar to EF86 - datasheet:
http://www.shinjo.info/frank/sheets/009/e/E80F.pdf
http://www.shinjo.info/frank/sheets/030/e/E80F.pdf

In attachment is updated schematic with voltages measured for E80F(blue) and EF86(green).
Measurements refered to ground except plate-cathode.
Since E80F draw 100mA more heater current, measurements for both tubes were taken with adjusting heater voltage to the same value  (6.28V). Measured for cardioid pattern.

From subjective tests - increasing cathode resistor value gave worst results in response...

Since am tweaking polarisation voltages by voltage dividers resistors values, there's an option to go up and down with plate voltage. 
Should i think about different value of plate resistor?
My next idea is to remove 10k resistor in pattern voltage path and replace first 100k resistor in PSU divider for potentiometer and resistor - to set accurate voltge for "back" diaphragm

 

[ln76d]

Any option for using osciloscope and generator?  ;D

 

[PRR]

> set proper bias for E80F tube in my G7.

What is "proper"?

Triodes can usually be run over a w-i-d-e range of values without much change.

Heater voltage should not matter (if correct). In the G7, capsule bias has no effect.

Polarization directly sets output level but does not have much other effect.

> Should i think about different value of plate resistor?

No; or not first-off. Plate resistor is selected for the load.

Gyraf has EF86 with plate at 67V. 91

You have E80F with plate at 47V. 107

The '80 sucks more current than the '86.

The obvious way to reduce current is to increase the bypassed 1.6K. Try 2.2K.

> increasing cathode resistor value gave worst results in response...

Increasing the 220r will reduce output, which always sounds "less good".

If increasing the 1.6K (how much?) is "worse", then I do not know what you are seeking.

Are your "subjective tests" controlled against simple change of sensitivity? Sensitivity is not important: you have a preamp knob for that.

 

[ln76d]

> set proper bias for E80F tube in my G7.

What is "proper"?

Triodes can usually be run over a w-i-d-e range of values without much change.

Heater voltage should not matter (if correct). In the G7, capsule bias has no effect.

Polarization directly sets output level but does not have much other effect.

I mean proper by current consumption  - bulb temperature, output level etc.

> Should i think about different value of plate resistor?

No; or not first-off. Plate resistor is selected for the load.

Gyraf has EF86 with plate at 67V. 91

You have E80F with plate at 47V. 107

The '80 sucks more current than the '86.

The obvious way to reduce current is to increase the bypassed 1.6K. Try 2.2K.

Comparing both tubes datasheets, there's a common example of "A.F. amplifier".
Of course it's different circuit for pentode operation mode.
First what i found that the plate load for EF86 is 100k and for EF80 220k.
Cathode resistor (100v supply) EF86 1.5K, EF80 3.3K.

Doesn't increasing of plate resistor and higher voltage (right now i set it up to 170V) give higher gain, lower distortion, more headroom?
At the same time increase cathode resistor and get less current to get higher output impedance still more gain and less heat?

> increasing cathode resistor value gave worst results in response...

Increasing the 220r will reduce output, which always sounds "less good".

If increasing the 1.6K (how much?) is "worse", then I do not know what you are seeking.

Are your "subjective tests" controlled against simple change of sensitivity? Sensitivity is not important: you have a preamp knob for that.

Yes, before starting topic i increased 220r It sounded "less good".

Yesterday i did some tests and tried to see something on the oscilloscope
Nothing happend when i put 1kHz sine wave to the input and change the value of cathode resistor in place of 1.6k.
I used 2.2k pot.  I did tests for other frequencies.
Also signal was sent to the mic preamp and power amp with cheap vu meter.
Adjusting resistance there was visible vu meter needle movement .
From 1.25k value of pot there was no change in movement when increasing.
Noise?

I did also tests with the scope and subjective listening for output capacitor. I put on the switch four capacitors 0.5uF, 1uF, 2.2uF, 3.3uF.  Increasing value of capacitance didn't gave any significant change in usable frequency spectrum but in subjective test better result was with lowest capacitance... Less capacitive load on the transformer?  Right now  0.5uFstayed at the output.

Another test was with HF damping capacitor from plate to ground (like tele elam) i did scope tests because subjective tests gave me only audible results with 680pF and higher.
Four capacitors 100pF, 300pF, 450pF, 1nF. I chose 300pF,  which gave only 1dB reduction in 10k-20kHz range.

 

[ln76d]

Could someone explain what for is 220r cathode resistor?
Is it only for feedback additional option applying or it has some significant role?
Why not use  1,82K instead (assuming that the original values are used and no feedback) ?

 

[Matt Nolan]

As far as the DC operating point is concerned, it sees 1.82k. As far as the small AC signal is concerned, it sees only the 220R as the 1.6k is bypassed by capacitors. So, the 220R provides a little negative feedback for AC, reducing gain a little, increasing headroom a little and reducing distortion a little.

You are right that it also allows the post-transformer feedback option. Tapping that in across a bypassed resistor wouldn't do much.

 

[PRR]

And with that OT, the "NFB" doesn't really do anything. Gyraf's diagram says it should be omitted.

 

[ln76d]

As far as the DC operating point is concerned, it sees 1.82k. As far as the small AC signal is concerned, it sees only the 220R as the 1.6k is bypassed by capacitors. So, the 220R provides a little negative feedback for AC, reducing gain a little, increasing headroom a little and reducing distortion a little.

You are right that it also allows the post-transformer feedback option. Tapping that in across a bypassed resistor wouldn't do much.

I really like the switch method for subjective tests - my tube probably not...
I did test for switching 220r resistor, keeping it in the circuit in both ways so for DC is the same (measured), unfortunately this method don't give a quick comparision. Tube need a while to stabilise voltage after switching.
Problem with these tests is that sometimes i'm not sure what's better. There's a minor difference but since it's late i will try again tomorrow.

 

[ln76d]

I gave up with  plate resistor value changing

Definitely i prefer mike without 220r cathode resistor.

Right now microphone is running close to 170V.
With 1.82k at cathode:
100k plate resistor - before  155.46V, after 51.43V = 104.03V drop at resistor - 1,04mA
Cathode-plate 49.49V, cathode 1.895V

With 2.7k at cathode:
100k plate resistor - before 156.56V, after 60.10V = 96,46V drop at resistor - 0,965mA
Cathode-plate 58V, cathode 2.60V

What do you Guys think about runing mike with 2.7k nd 2.6V at cathode?
Something stands in the way to use it?