U-47 Bias with another tube.

[oscarhuang]

Ohm’s law. U= RxI. So if you want 1 volt on cathode. Filament = 0.2 amp. The current in the series circuit does not change. So the formula is
1= R x 0.2 amp. R= 5 ohm.
1v+ 6.3v = 7.3v.

 

[vmanj]

And if you do it differently, it will be the same thing?
If the cathode current is small, then I increased the values of resistors R1 and R2 so as not to "overload" 6.3V.

 

[Monte McGuire]

Yes, but we are talking about the fact that the filament was no more than 6.3V.

Yes... But do you want the standard lifetime of an EF86, or the 50-100 year life you get by underheating? What's the point of making the heater 6.3V? No tube mike I know of runs the heater at nominal voltage.

 

[Tubetec]

It should still work like that ,but even with resistances R1 and R2 both scaled back to 1/10 of the value you shouldnt cause any overload of the LT supply , as we said before the heater at switch on is 5-7.5 ohms , your voltage divider is 1260 ohms in parralel , lets calculate the extra current in the heater supply due to the voltage divider , 5mA extra , now try 96 and 30 ohms , now our potential divider draws an extra 50 mA from the heater circuit , thats not going to be an issue , a 6.3 volt indicator bulb would draw more current that that . The higher value of R2 will in itself cause a small voltage drop due to the cathode current of the tube which will offset the calculation to some degree , a lower value R2 sits the cathode closer to ground as far as AC signal is concerned , usually thats a good thing .

As was previously mentioned if you supply the cathode/heater in that way you may need an extra quiet 6.3v rail , the grid bias arrangement like the U67 allows extra passive smoothing of the grid reference voltage , likewise in U47 the 1780 ohm series resistor not only drops a lot of voltage but helps passive smooth the heater/cathode supply .

The NU67 psu spec says 6.5volts LT supply , To allow for some voltage drop in the cable the mic schematic shows between 5.7-6.3 volts , If I was to apply the voltage to the cathode Id favour the U47 type setup , if you want to apply negative volts to the grid just go with the standard U67 setup .

 

[Tubetec]

Its worth having a look at Ruud's U67-ish mic topic , theres some discussion about how the bias of the U47 and U67 differs .

 

[vmanj]

If you look at a fixed bias, as in u47, how much can you increase the value of a 29 ohm resistor, for example, to 3K?

 

[dmp]

If you look at a fixed bias, as in u47, how much can you increase the value of a 29 ohm resistor, for example, to 3K?

The whole point of the design is to minimize that resistor. Making it larger moves the mic back to cathode bias. The larger the cathode resistor, the more negative feedback happens due to the cathode voltage changing with signal. Higher cathode resistors need a stabilizing cap to counter this effect.

Check out the U47 Alt by Oliver for using H+ instead of B+

One thing that is often lost when looking at the different biasing of the U67 & U47 is that the power supply is inherently a part of the mic.
In the U67, the cathode isn't connected to an 'all stable' ground, the cathode is connected to the (-) of the filtering capacitors.
And in the U47, the cathode is connected to the filtering caps through the 29 ohm.
In a cathode bias amp, the cathode is connected directly to the (+) side of a cap right at the mic.
Once you consider the power supply as part of the mic, the differences seem less significant. The cathode is still 'seeing' a capacitor.
But, in my experience, these 'fixed bias' mics have a significant difference in character that is really cool.

 

[doqmemory]

Check out the U47 Alt by Oliver for using H+ instead of B+

This works pretty good with the Matador/Chunger power supply, too.

 

[dmp]

This works pretty good with the Matador/Chunger power supply, too.

Sufficient H+ filtering? Good to know

 

[doqmemory]

Sufficient H+ filtering? Good to know

It is just slightly noisier than the p/s from D-47 by @poctop, but still acceptable.

 

[vmanj]

And in the U47, the cathode is connected to the filtering caps through the 29 ohm.

If we consider the filter caps in the power supply (as in the image), then the cathode is connected to the filter caps not through R1 29 Ohm, but through R2 + R3.

 

[vmanj]

Once you consider the power supply as part of the mic, the differences seem less significant. The cathode is still 'seeing' a capacitor.

Bypass cap of the cathode resistor 29Ω, connected through 2-3K (conditionally).
Can such a "shunt" have a noticeable effect on the sound?

 

[dmp]

If we consider the filter caps in the power supply (as in the image), then the cathode is connected to the filter caps not through R1 29 Ohm, but through R2 + R3.

It's also connected through 29 ohm. The (-) side of a cap is also a capacitance. The path through 29 is more signifcant than the path through 1780. But you are on the right track!
A good analogy to understand whats happening is to think of a capacitor as a bucket of water, resistors as valves, and current as the water flow (voltage is water pressure). The bigger the capacitance the bigger the bucket, and the more 'damped' waves will be from the water flow. The tube current is going from the cathode to ground (into the bucket). Whereever a node is attached to a capacitor, (-) or (+), there is a bucket dampening out waves.

The DC bias can all be worked out easily with ohms law. Then the AC bias can be considered. If the tube current varies by some small perturbation, how much does the grid change? (negative feedback) dV = dI*29

 

[dmp]

Compare the U47 to the C12
Then: what are you trying to accomplish?

 

[luckythedog]

My 2p worth. When grid impedance is so high, it is effectively floating. Then the valve adopts a 'self bias' grid-cathode potential as set by the valve. Mucking with cathode resistor doesn't change operating point in this arrangement.

If one wishes absolute zero grid current, one needs to return the grid leak resistor to grid potential. It follows to return grid leak resistor to a tap of a dc heater (or cathode divider) if that can provide correct overall potential.

T