[adamus]

True but I am not one for fashion. The datasheet suggests a very low dynamic impedance and what is there looks more linear than an led. As you say, proof is in the listening.

 

[regal]

Also might want try one of these http://www.fairchildsemi.com/ds/2N/2N7000.pdf  with the gate tied to the drain to the cathode would give about a similiar bias.

 

[adamus]

update: I have it on a perf board ready to test
 

 
It'll find a test home in my tweakers amp which is currently a c3g parafeed. By the way, the c3g ccs loaded parafeed is currently biased by a tl431.... i think i prefer it to leds!!! the bass seems better... which could always be a small change in operating point.
 
test home
 

 

 

[regal]

Adamus is the tl431  adjustble under the cathode,  could you set it for 10V's or higher for tubes that require a lot of bias?

 

[adamus]

its adjustable upto 36v (max rating) and achived using a couple of resistors - the datasheet explains better than i can. Watch the power rating though, it soon falls off with increased temp. Another thing to add is that i have a slightly increased noise floor with tle431, not noticable with senns but is with grados. I am looking for an osciallation but cant spot one, but my scope is a bit rubbish. The c3gs and d3a LOVE to oscillate, i doubt this would be a problem on non 'extreme' valves.
 
another method that would probably yield better reults is using a pnp in emitter follower.. you get the stable vref from the leds and the low dynamic impedance of the pnp - you can also scale power requirements with pnp type. cathode voltage = vref + Vce(which is about 0.7v)
 

 

 

[regal]

Quote:

 

 

Brilliant.   Thankyou very much adamus