I am currently pondering different biasing schemes for an estat headphone amp using DHT output tubes in push pull configuration (see here for a very early concept drawing). Looking at many posts, blogs and websites I found a lot of information about good old cathode bias (be it with a simple resistor or something slightly fancier like a LED). There is also filament bias, which is also a form of cathode bias, but does away with the high impedance cathode resistor. However, it consumes a lot of power and dissibpates a lot of heat.
I couldn't find much about fixed/grid bias (except maybe Valve Wizard). Why does grid bias not get more love?
My current thinking is that fixed bias would be a nice biasing scheme for my application because it provides the following advantages:
I couldn't find much about fixed/grid bias (except maybe Valve Wizard). Why does grid bias not get more love?
My current thinking is that fixed bias would be a nice biasing scheme for my application because it provides the following advantages:
- It's easy to adjust bias for various tubes (so the amp could work with slightly different tube types)
- The cathode can connect directly to GNC/B-, so there is no need for a cathode resistor at all
- Bias current is essentially zero, so it should be easy to minimize ripple and noise on the grid DC bias voltage
- Grid bias requires a dedicated power supply (oh well, should not be a big deal)
- what else?
One disavantage I see is that you need to adjust it for each and every tube because there is a large spread.
Also, without a cathode R there is no stabilizing mechanism and Ia may drift with tube temp and tube aging.
Jan
Also, without a cathode R there is no stabilizing mechanism and Ia may drift with tube temp and tube aging.
Jan
- Bias current is essentially zero
... not always true. If you use only small tubes or only A1 mode, it's true, but in case of large/r/ tubes it's not fully obvious.
Power tubes requires smaller grid leak resistor (or low impedance grid driving -CF, SF-) to ensure grid current deduction.
Cathode bias is safe, but required AC blocking of cathode resistor.
Filament bias is -practically- usable for low/er/ biased DH tubes.
Fixed grid bias may cause a problem, if bias voltage left out (thermal run of tube, due untreated anode current), or bias and anode (B+) voltage moves separately (stabilize both, or both of them must unstabilized, feeds from same PT).
Fixed bias generally requires input capacitor (separate negative grid voltage from previous stage), or input transformer (secondary connected to negative bias voltage).
Possible to use series battery with grid (if bias voltage enough small) ... but noiseless (and "toneless") battery is rare.
I usually use fixed bias for driving (CF, SF) power tubes. Almost never use it for preamp or VAS stages ... but this is my decision.
True.
On the other hand, it does allow using the tubes at whatever bias I want, not where Georg Ohm told the cathode resistor to go. I also consider my amps as pets that need a bit of care from time to time, so I am not worried about a bit of checking and adjusting here and there.
Grid battery bias can be made adjustable by using Li-ion batteries. The voltage of such battery can be finely adjusted between 3.4 and 4.0 V by charging or discharging it. With no current draw, a good Li-ion battery retains steady voltage for years.
NiCd and NiMH batteries also have charge-dependent voltage, but the range is narrower and the voltage is not steady due to relatively high rate of self-discharge.
NiCd and NiMH batteries also have charge-dependent voltage, but the range is narrower and the voltage is not steady due to relatively high rate of self-discharge.