Fixed bias question.

[bluefoxicy]

Fixed biasing means the tube is biased by means of a DC voltage, which is usually a negative voltage applied to the grid of the tube with respect to the cathode. As the negative grid voltage is adjusted, the bias current will increase or decrease, depending upon the direction the bias voltage is going. In general, as the bias voltage becomes more negative, the bias current becomes smaller, and the tube is biased "colder". As the bias voltage is adjusted less negative, towards zero volts DC, the bias current becomes larger, and the tube is biased "hotter". This is because a tube is a "normally on" device; that is, it allows current to flow from the cathode to the plate when the grid is at zero volts with respect to the cathode. The tube can be turned off, and the current flow stopped, by making the grid voltage negative with respect to the cathode. The tube can also be biased by referencing the grid to ground, or zero volts DC, and applying a positive DC voltage to the cathode. This is the same as keeping the cathode at ground and applying a negative DC voltage to the grid, because it is the grid voltage with respect to the cathode that determines the amount of bias current in the tube.

Since vacuum tubes are "normally on" devices, a trick can be used to bias them without having to supply a negative DC voltage source to the grid. If a resistor is placed between the cathode and ground, and the grid of the tube is referenced to ground (usually by connecting a large value resistor, such as a 1Meg, from grid to ground), the tube will try to conduct a large current from cathode to plate, since the grid and cathode are initially at ground potential. However, this cathode current flow will cause a voltage drop across the cathode resistor, making the cathode voltage positive with respect to the grid. Since the cathode voltage is now positive with respect to the grid, the current flow will decrease, and the tube will head back towards cutoff. A point of equilibrium will quickly be reached where the increase in current is offset exactly by the increase in cathode voltage, and the bias current will stabilize at some particular value. It will remain at this value unless the resistor value is changed, or a different tube with different characteristics is plugged in. This allows the desired bias point to be set by varying the value of the cathode resistor.

Uh, I'm having a hard time understanding the "Apply negative voltage to the grid" thing here, and understanding how this is different from cathode bias. See this diagram (cathode biased EL84):



What I notice here is that the positive rail goes to the anode, and the negative to the cathode. This has to happen, because electrons flow from the cathode to the anode; a pentode still acts like a diode, reversing polarity simply will not work.

Now, my negative power source is common to ground. Where the hell do I get a negative voltage with respect to the cathode, without dropping a resistor under the cathode? It looks like if I want to do this, I need to apply a positive voltage to the cathode... i.e., drop a small positive rail right on the cathode, using a zener to limit voltage and a resistor to limit current so I'm only shorting a couple microamps to ground/common (wtf?). Then I can fixed-bias the amp by referencing the grid to ground with a trim pot at (I guess) R5.

... what the hell? This looks like complicated brain damage. Cathode bias seems much simpler.

[BackwardsBoB]

Cathode bias is simpler.

Most fixed-bias amps either have a separate winding to generate the negative bias voltage, or they use a center-tapped plate winding with only two diodes (perhaps a tube rectifier) instead of a bridge, and the center tap grounded. They tap a bit of the large negative excursions of one of the ends of the plate winding to generate a negative voltage.

[bluefoxicy]

This seems overcomplicated and pointless, except that it leaves the cathode more negative than cathode biasing and so would probably generate more current flow. Cathode biasing seems fair and simple.

[Steve Conner]

You are right. Cathode bias does more or less the same as fixed bias, but it's simpler and uses fewer components. However, it isn't used in higher powered amps because the cathode resistor has to dissipate a lot of heat. (if you cathode biased a SVT, think about 60 watts!)

At some size of amp, the cost of the large resistor gets more than the cost of the extra parts for fixed bias. There's also the factor that you can get more power from a given tube lineup with fixed bias.

[Paul P]

... what the hell? This looks like complicated brain damage. Cathode bias seems much simpler.

I doubt that anything pointless would have survived in something that's been
mass-produced for fifty years .

Paul P

[BackwardsBoB]

There are a few side-effects from cathode biasing.

If you leave out the capacitor around the cathode resistor, you'll reduce gain a bit and increase linearity due to the negative feedback induced by the resistor. If you put a big capacitor in parallel, creating a virtual ground at the cathode, this doesn't happen, and simulation with sine-waves looks good, with little difference from fixed bias. Transients in actual musical content do get distorted a bit, as the cap charges or discharges, temporarily changing output current.

These effects don't necessarily sound bad, and cathode biasing has been used successfully in many great sounding guitar amps, with the advantage that you don't even have to think about re-biasing when you change tubes unless you change tube types.

[Steve Conner]

+1 on all that. I think a lot of people swear by those "transients... getting distorted", it's almost like having a vari-mu compressor built in to your amp.

Nevertheless, there is a point in amp size and power beyond which everyone goes to fixed bias. The biggest cathode-biased amp I can think of just now would be the one in the Leslie speaker with its two 6550s.

Where the hell do I get a negative voltage with respect to the cathode

Fixed bias amps have a little dedicated power supply to do just that. There's an extra winding or tap on the transformer that gets rectified and filtered to produce a negative rail 40 or 50v below ground. The power tube cathodes are then connected directly to ground.

[Rob Mercure]

BF,

To elaborate on Steve's response note that the schemo you provided included a pot (wired as a rheostat) to adjust output tube bias - have you priced high dissipation (wattage) pots? Imagine trying to set the individual output tube bias on something like the mentioned SVT using large power resistors with sliding taps - you'd need a separate chassis just for the cathode resistors. Not to mention that potentiometers deteriorate in DC use and get noisy. And as far as parts count you've got a pot and a bypass cap - with some fixed setups you've got a diode, a voltage dropping resistor/cap, and a filter cap - not a whole lot more. In general I guess I like the control of circuit parameters that fixed bias gives me more than cathode bias - but then again you probably like ketchup on your hotdogs <grin>.

Rob