6DJ8/ECC88 mu-follower

[SY]

Bipolar CCS (e.g, the diyAudio boards or something like that) or simple DN2540 cascode.

 

[Shoog]

I just built one in an evening from some salvaged TV transistors and a pin board. Worked first time (almost) and cost absolutely nothing. Works wonderfully.

Shoog

 

[Tyimo]

Thanks SY!

Tyimo

 

[Tyimo]

Hy Sy!

Untill I get some DN2540 should I try the circuit with 20K/2W resistor instead of the solid state CCS? (on 160 B+)

Tyimo

 

[SY]

You could slap together a very high performance bipolar CCS with parts on hand, I'll bet.

 

[SY]

BTW, the reason not to use 20k resistors is the large drop across them- at 10mA, that's more than your B+ supply. CCS is ideal in that situation to minimize the required drop and provide better power supply rejection as a side benefit.

 

[Tyimo]

Thanks Sy!

Could you show me the schematic of your DIYaudio Bipolar CCS?

Tyimo

 

[SY]

Use the "search" function and you'll find far more than you could ever hope for.

 

[dsavitsk]

I did basically this here: http://www.ecp.cc/modular_dac.html I am currently using a 8416 instead of the 6dj8, but the specs are the same. ra is ~2.4K, so into a 50K load, there is plenty of gain and plenty of current. I'd only add the buffer if you find you need it.

 

[Tyimo]

Thanks!

 

[JoshK]

I looked over SY's schematic a bit more, and I am thinking the grid leak resistor and the grid's impedance is quite tough for this dac. It wants a low impedance load and resistor to ground should be higher than 100R for optimal linearity from the dac. This implies driving the cathode, ala grounded grid, for first stage.

JB gives details on such a circuit in his blog. It requires a servo to maintain the 0V that the dac wants. (His circuit is drawn for TDA1541 but the same is true for the BB1704 iirc.)