DCAudioDIY.com

Cathode (self) bias vs Fixed bias is another trade-off. Cathode bias has the advantage of being fail safe, but you need to have a higher voltage supply since with power tubes you drop a good amount of voltage across the cathode resistor (for a 300B, maybe 80V or so). On the other hand, fixed bias you need to have a good clean negative voltage at the grid of the tube. The current is very low, zero at the grid, so filtering is not as much of an issue, the final bias supply capacitor is fully charged and therefore supplies DC without ripple if it is sufficient in size. Many bias supplies consist of a half-wave rectifier into a capacitor. I always like to use a CLC filter, and voltage dividing pot + resistor for adjustment (mandatory). Regulation is not required or recommended. It works just fine.

For small signal circuits, I often use fixed bias using LED or diodes at the cathode instead of a cathode resistor + bypass cap. Batteries also work. But it is less flexible for setting up the operating point of the tube (since you have discrete voltages by these schemes, instead of using cathode resistors that can be selected to achieve the bias voltage. But where it works, it works great, and you have no cathode bypass capacitor.

David

[/quote] David and Charlie turned me on to DC Link caps in power supply use. The improvement over electrolytics there is not insignificant!

I had thought the same about using them (bypassed) in large value speaker crossover applications. But there, as well as in the cathode circuit makes REALLY good sense.

I'll also agree about most common applications where grounding the cathode means that a negative grid supply is needed. The complexity of that, combined with the caps needed there to ensure quiet operation, are probably worse sonically than a really good cathode bypass.

Stuart[/quote]

Each method has its positives. Cathode bias was strongly supported by Sid Smith (Marantz) because of the way the cathode bias "went along for the ride" in terms of dynamics. Sid's opinion was that fixed-bias can suppress dynamics unless you implement it with a grounded cathode rather than a small-value cathode resistor (10-ohms) typically used to prevent excessive plate current in the event of a tube failure. Jadis did something different. They did an old G.E. trick of a hybrid scheme where fixed and cathode bias work together. The fixed bias sets a starting point and the cathode-bias controls the overall operation of the tube. Providing "the best of both worlds."

_________________
Walt

I think the function of 10 ohm resistor between cathode and ground for fixed bias is more for measuring the plate current for bias and/or balance adjustment (PP). I remember Dave Berning suggesting the hybrid bias (fixed + cathode) to me years ago when I was converting my GM70 amp from cathode to fixed bias. The another advantage of a hybrid bias arrangement is increases the effective plate voltage across the power tube for the same B+ compared with a cathode biased amp. This will provide more power output so long as you keep within the plate dissipation. Sometimes a 100% fixed bias conversion results in a higher plate voltage than the tube can handle. So a useful method for getting a little more out of cathode biased amp. Still, bypassing is needed, and a much larger bypass cap will be needed across the smaller value cathode resistor. Also, failure of the fixed bias supply will result in the tube at very high plate current, but no where as bad as fixed bias failure. One can protect against a fixed bias failure by a large value cathode resistor in parallel with a fuse with a current rating perhaps 2 or 3 times the quiescent current. Of course, this is basically a plate fuse in the cathode, when it fails, the tube reverts to cathode bias at a very low current to protect the tube and output transformer. I use this scheme in my 211 amps.

I must apologize. Until reading these last few posts, I did not realize we were talking about OUTPUT tube bypass caps/cathode bias operation. Like an idiot, I assumed we were talking about a voltage gain stage. In the output tube case, any of the methods make sense, all with compromises. The complexity of fixed bias isn't such a big deal here as opposed to applying it to a signal tube grid (and needing to be really quiet).

Sorry. Carry on!

Stuart

Stuart,

It wasn't clear in the initial post and initial discussion. However, fixed bias is very easy to do with voltage amplification stages. Diode (inc. LED) bias, and battery bias are two methods that are very low noise, even suitable for phono preamps. Check out Morgan Jones and others. Using a string of cheap red LEDs (1.7V each), SiC diodes (0.9V each), or rechargeable batteries in the cathode in place of the cathode resistor work great. You can even place a battery inline to the grid. Of course, these are pretty coarse steps for determining operating point, but you can use different color LEDs (each with their own voltage drop). The "Red Light District" PP EL84 amp project uses a parallel string of red leds (to get around the 20mA limit) to bias the output tubes. So I would argue that fixed bias schemes are very practical for voltage amplification stages and get completely around the use of cathode bypass capacitors (which due to the large value needed, are usually electrolytic capacitors).

David

@Pelliott321 -- A "example" schematic [with said cathode cap circled] would really helpful to understandinhg thie jist of this post. Just saying.

_________________
"We are all inventors, each sailing out on a voyage of discovery, guided each by a private chart, of which there is no duplicate. The world is all gates, all opportunities."
Ralph Waldo Emerson