Octal equivalents for 6C45?

[Brian Beck]

The key to LED biasing without sonic "funnies" is current. If you get the current up past 10mA or so for most LEDs, the dynamic impedance becomes lower and more constant. One cool trick for using them with low current tubes like 6SL7 is to run a current source from the supply rail to the junction of the LED and the cathode to make up the difference. That does not inject supply noise, a common worry- the output impedance of the CCS and the dynamic impedance of the LED form a voltage divider which knocks down any supply noise by a whole lot of dB.

That should work. Another possiblity: if any subsequent stage is a cathode follower, instead of connecting the CF's cathode resistor (or CCS) to ground, connect it to the junction of the first stage cathode and LED. This will dump more current into the LED "for free" without having to drag more current out of the B+. If the CF is the very next stage, a small amount of positive feedback is thus created, but before getting all bent about it, consider this: if the CF's cathode resistor is, say, 33K, and the LED's dynamic resistance is, say, 5 ohms, then the positive feedback loop "suffers" a 5/33005 loss due to this voltage divider - that's a 76dB loss, so there won't be any oscillation or other squirrelies. With a CCS, the loop loss would be even more. Just a thought...

 

[SY]

Nice idea, saves some parts. That appeals to my cheapskate nature.