hi all
was looking at the Lighthouse Electric link that was posted for tube newbies sticky.
on the page explaining cathode and grid leak biasing here.
grid leak biasing requires Cg and Rg connected to the grid of the tube
cathode biasing requires Cg and Rk connected to the cathode, in addition to the Cg and Rg at the grid of the tube. i would assume that this is correct. why is it that there are circuits that do not have Cg and Rg connected at the grid for cathode biasing?
if cathode biasing requires 2 more components at the cathode that might affect sound, why not just go for grid leak biasing instead?
thanks
was looking at the Lighthouse Electric link that was posted for tube newbies sticky.
on the page explaining cathode and grid leak biasing here.
grid leak biasing requires Cg and Rg connected to the grid of the tube
An externally hosted image should be here but it was not working when we last tested it.
cathode biasing requires Cg and Rk connected to the cathode, in addition to the Cg and Rg at the grid of the tube. i would assume that this is correct. why is it that there are circuits that do not have Cg and Rg connected at the grid for cathode biasing?
An externally hosted image should be here but it was not working when we last tested it.
if cathode biasing requires 2 more components at the cathode that might affect sound, why not just go for grid leak biasing instead?
thanks
Hi,
It's a good thing to bypass the cathode resistor by a small capacitor. This makes the amp more musical. For power tubes, I have seen capataince up to 100uf. For rectifying tubes, I use tiny ones, like a few nf to one uf.
It's ok without bypassing, but it could be noisy sometimes. I don't have any experiment to support my last statement, noise part. Music sounds quick and stiff without bypassing.
C.C.
It's a good thing to bypass the cathode resistor by a small capacitor. This makes the amp more musical. For power tubes, I have seen capataince up to 100uf. For rectifying tubes, I use tiny ones, like a few nf to one uf.
It's ok without bypassing, but it could be noisy sometimes. I don't have any experiment to support my last statement, noise part. Music sounds quick and stiff without bypassing.
C.C.
actually the capacitor at the cathode is to help reduce the Rk's effect on the plate resistance of the tube or something. i've read this somewhere. thus a capacitor is normally required in cathode biasing.
but my question is regarding the Cg at Rg at the grid of the tube.
are these 2 components absolutely required?
it seems that for cathode biasing, as part of it's layout, a grid leak biasing layout is used, but with more components. what are the benefits, if any?
but my question is regarding the Cg at Rg at the grid of the tube.
are these 2 components absolutely required?
it seems that for cathode biasing, as part of it's layout, a grid leak biasing layout is used, but with more components. what are the benefits, if any?
Hi,
Because the value of the gridleak resistor needs to be very high for the bias voltage to develop.
Resistance value in the order of 10Meg and more are not uncommon.
This method of biasing was rather popular with Western-European manufacturers wherever low negative bias was required. I.e. phonopreamps and such.
Rg is necessary unless the previous stage is direct coupled to this one.
Cg is often not required as grid current is usually very small and the output DC blocking cap of the previous stage will take care of blocking DC anyway.
Cheers,
Because the value of the gridleak resistor needs to be very high for the bias voltage to develop.
Resistance value in the order of 10Meg and more are not uncommon.
This method of biasing was rather popular with Western-European manufacturers wherever low negative bias was required. I.e. phonopreamps and such.
Rg is necessary unless the previous stage is direct coupled to this one.
Cg is often not required as grid current is usually very small and the output DC blocking cap of the previous stage will take care of blocking DC anyway.
Cheers,
Cg is necessary if there is a voltage difference: the goal of the cathode-biased stage is to keep grid at zero volts while the tube's bias current produces a voltage across Rk, biasing it off just the right amount to be in the linear range (depending on Rk of course, which needs to be suited to the tube and intended operating point). Likewise Rg is needed to ensure the grid stays near ground; there is very little grid current, but then, with no connection (it's basically a floating piece of wire inside the tube), it takes extremely little current to change the grid voltage a troublesome amount!
If the signal source is at 0VDC and has a lowish DC impedance (that is, it will tend to keep the grid at 0V if something tries to change it), you can omit Rg and Cg because it will have the same effect. You cannot omit just Rg (leaving Cg) because that will keep it isolated and will let the grid float wherever it pleases. Not good.
If the source is as mentioned, you can omit Cg, but Rg now only loads the source and serves no function in the circuit's operation.
As for grid leak biasing, it can be done but isn't a good idea because:
1. it loads the signal source (the grid, when positively biased, conducts; it forms a diode with the cathode. This rectifies the signal source, producing a negative grid voltage automatically). This can be partially overcome by using very large (often 10meg) grid leak resistors and small capacitors (less than .01uF).
2. Operating point varies widely and depends entirely on the signal strength. For none, or very small signals, the tube is fully on and plate voltage is low. With a small signal, it will remain in a good linear range. As signal is increased, bias increases and it will leave class A. Works good for radios though, as a good 40Vp-p from an oscillator, into a 12AU7 will provide -20V bias which is a good class C operating point. Of course, with tubes that conduct any better (especially sweep tubes, where this method was often used), you risk melting them if you lose drive signal.
So um...yeah. I could make more sense of that for 'ya but I'm still hazy... I need some breakfast. (Yes, at 3AM...
)
Tim
If the signal source is at 0VDC and has a lowish DC impedance (that is, it will tend to keep the grid at 0V if something tries to change it), you can omit Rg and Cg because it will have the same effect. You cannot omit just Rg (leaving Cg) because that will keep it isolated and will let the grid float wherever it pleases. Not good.
If the source is as mentioned, you can omit Cg, but Rg now only loads the source and serves no function in the circuit's operation.
As for grid leak biasing, it can be done but isn't a good idea because:
1. it loads the signal source (the grid, when positively biased, conducts; it forms a diode with the cathode. This rectifies the signal source, producing a negative grid voltage automatically). This can be partially overcome by using very large (often 10meg) grid leak resistors and small capacitors (less than .01uF).
2. Operating point varies widely and depends entirely on the signal strength. For none, or very small signals, the tube is fully on and plate voltage is low. With a small signal, it will remain in a good linear range. As signal is increased, bias increases and it will leave class A. Works good for radios though, as a good 40Vp-p from an oscillator, into a 12AU7 will provide -20V bias which is a good class C operating point. Of course, with tubes that conduct any better (especially sweep tubes, where this method was often used), you risk melting them if you lose drive signal.
So um...yeah. I could make more sense of that for 'ya but I'm still hazy... I need some breakfast. (Yes, at 3AM...
)Tim
thanks frank and tim for your replies.
i still have a couple of questions.
suppose we have a schematic as such:
taken from Lighthouse Electric site.
if i have a preamp before this, can i replace C1 and P1 with a 1Meg resistor as LH recommended or should it need other values? this would fall into the condition where tim mentioned (signal source is at 0VDC and has a lowish DC impedance).
the 6BM8 would then be cathode bias with just R1.
and for the ECL82, C2 and R3 would maintain the grid voltage to ECL82 with C3 and R4 would performing the cathode bias? (if i understand frank correctly, the presence of R3 in this circuit would mean that the previous stage is not direct coupled to this)
rgds
vincent
i still have a couple of questions.
suppose we have a schematic as such:
An externally hosted image should be here but it was not working when we last tested it.
taken from Lighthouse Electric site.
if i have a preamp before this, can i replace C1 and P1 with a 1Meg resistor as LH recommended or should it need other values? this would fall into the condition where tim mentioned (signal source is at 0VDC and has a lowish DC impedance).
the 6BM8 would then be cathode bias with just R1.
and for the ECL82, C2 and R3 would maintain the grid voltage to ECL82 with C3 and R4 would performing the cathode bias? (if i understand frank correctly, the presence of R3 in this circuit would mean that the previous stage is not direct coupled to this)
rgds
vincent
As for removing C1, that depends on the preamp. If it's how I build them, which is with a coupling cap on the output, as shown here:
There's no DC path to secure the grid to a reference.
<EDIT> Oh, there's a 100k on the output. 20 lashes with a wet noodle for me, for not remembering my very own schematics...
If you leave P1 (use a 100k pot because you need the lower impedance to drive the triode), it will work because it will secure the DC point. If you removed P1 (and used something elsewhere for volume control), the grid would float wherever it pleases, not good.
If the preamp has an output transformer, or is solid state, chances are output has a stable DC output. Only in this case can you safely remove the grid leak components, hence why I always include DC isolation (a coupling (series) capacitor) in my designs.
Tim
An externally hosted image should be here but it was not working when we last tested it.
There's no DC path to secure the grid to a reference.
<EDIT> Oh, there's a 100k on the output. 20 lashes with a wet noodle for me, for not remembering my very own schematics...
If you leave P1 (use a 100k pot because you need the lower impedance to drive the triode), it will work because it will secure the DC point. If you removed P1 (and used something elsewhere for volume control), the grid would float wherever it pleases, not good.
If the preamp has an output transformer, or is solid state, chances are output has a stable DC output. Only in this case can you safely remove the grid leak components, hence why I always include DC isolation (a coupling (series) capacitor) in my designs.
Tim
Sch3mat1c said:
i see.
thus, if i were to have a integrated stereo tube amp, then it should ideally have a 100k volume pot/attenuator.
and if is is a tube power amp, being driven by a preamp, i can get away with not having grid leak components (Cg and Rg in the first LE pic). coupling cap is required only if the previous stage does not block DC.
ok. now i understand the relationship between the previous stage and the first tube stage of the amp in a cathode self-bias configuration.
thanks.
The cap is used to stabilize the voltage drop on Rk. If there is no cap on cathode, bias voltage alters and that is not good, because amplification and max input signal voltage in A class reduces.
How about biasing with batteries:
1. http://audio.ring.lt/se2/6p9.png
2. battery instead of cathode cap and resistor with + pole to cathode;
3. battery in series with grid cap with - pole to grid?
I't would be nice to hear your opinion about this way of biasing.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.