I wish I culd help, but I have just one pair of "Sylvania" 12B4A...
But, I have to ask - what operating point you plan to use, and way of biasing? I mean - with similar OP (100V/15mA) like with 5687 version, bias is about -12...13V with 12B4A.

I have had good luck with 12B4 and noise or microphony. I think that RCA were the most microphonic, and the others about equally risky of being noisy. Most were hard run power supply tubes, so don't take too much stock in that.

If you meant a 750 mV sensitivity, you won't need much gain I think.

On the power TX design, you want enough turns on the primary to yeild a flux density below the knee of the B-H loop. Keep the flux density low and the core loss will be lower. More turns does mean more copper length and thus resistance. There is nothing really magical about TX design. Nothing really complicated either. The output voltage will then be determined by the turns ratio primary to secondary. There is a bit more, and I'll help you find the answers. Not even output TX design is too complicated for a public forum...:)
cheers,
Douglas

Now are we considering audio outputs, or power Iron for the Leackage L question?

cheers,
Douglas

• http://audioroundtable.com/Tubes/messages/929.html

Seems to me leakage L is improtant for at least two reasons. One is frequency response. Leakage L looks like a series L with the primary L/reflected load parallel combo. The other is for transient voltage tolerance. Leakage L is going to offer an independant L to create a high voltage transient if current is interrupted.
cheers,
Douglas

I have been trying to track down some of the DN2540N5 depletion mode transistors, but have been stiking out.
I have found a few larger electronic firms such as Mouser, that will be happy to sell me 50 at a time.
None of the firms where I have looked, stock this model of transistor.
I am not sure I would need 50 of these.
Does anyone have a decent supplier where I can buy a smaller lot of 10 or less of the DN2540N5 transistors?

Douglas did mention a TO-92 variant that would work, DN3545N3.
But, I would rather have the TO-22 that appears to be a little more robust and has more heat sink realestate.

Also, I would love to hear about others experiences in building a power supply for the Guinevere?

I am under the impression that the 12B4 has a slight problem of being susceptible to noise from the power supply. At least this is what I have read on a few other forums, or believe this is what they meant. Does the CCS reduce this susceptability to noise?

Thanks

Norris Wilson