CCS bias question

[alexmoose]

hello! I am going to change to CCS bias in the next day or two in my EL34 SE amplfier because I found some LM317s at Radio Shack (finally something useful at Radio Shack). and I wanted to know what to set the current to on the LM317.

I buy all of my tubes from Tubedepot (they have two day shipping for 4 bucks!) and they all have a "Bias point" printed on them. this amp runs class A, so should I set the 317 to run at the "bias point"?

Thank you so much

-Alex

[ray_moth]

. . . should I set the 317 to run at the "bias point"?

That would seem the logical thing to do. However, don't forget to bypass the CCS with a large cap (same as you would use for conventional resistive cathode bias) or you won't get any output signal! Also, the LM317 is not known for its good frequency response and may not be a good choice here, IMHO.

[ErikdeBest]

Besides Ray Moth comments I would add that the bias value of an EL34 approximates dangerously the max value of the LM317, which takes 37V between input and output! You can try some zeners in serie

http://www.glass-ware.com/tubecircui...o_Biasing.html
http://www.glass-ware.com/tubecircui...Biasing_4.html

[Brian Beck]

Originally posted by ray_moth

That would seem the logical thing to do. However, don't forget to bypass the CCS with a large cap (same as you would use for conventional resistive cathode bias) or you won't get any output signal! Also, the LM317 is not known for its good frequency response and may not be a good choice here, IMHO.

The bypass cap will, by definition, bypass all audio currents to ground, and keep the voltage of the cathode and the 317 constant. Therefore, the poor performance of the 317 at higher frequencies will not a problem. All the 317 has to do is to hold the DC cathode current constant, which it does well.

The voltage rating concern is another matter though...

[Brian Beck]

Further thought on the voltage rating of the 317: Why not just add a series resistor to the 317 CCS (with its setting resistor)? All three components would be enclosed "under" the bypass cap of course. I don't see why you'd need cascode FETs or zeners. The idea of a current source is to make a high resistance anyway, so adding another resistor in series won't hurt a bit, and it will reduce the power dissipated in the 317 too. The value is easy to pick. How many volts do you want to drop? Divide this by the set 317 CCS current.