Subject: July 2000 issue
    I read with interest the "heater issue." I have heard of the heater-as-cathode-resistor trick (I think it was used in an old guitar amp?) but it was great to be reminded about it. The interchangeable ?SN7 item was very clever, I liked that one quite a bit too.
   I saw the circuit for voltage-regulating the 2A3 filament : do you have a good suggestion for current-regulating it? The 2.5A draw seems to preclude the usual approaches as I understand it.
    Going back to an earlier issue (May, the second circuit in the "Safe Loftin-White" article), I thought you might like to know I have built a very similar direct-coupled amp with excellent results. It uses the voltage drop across the driver's choke load (plus a resistor) to bias the output tube, whose cathode feeds the driver/resistor/choke combo. I had not noticed your similar circuit until it was pointed out to me, and in any case I am not sure you have built the design. I wanted to encourage you to try a listen -it is the best amp I have built (that is not such a long list though) but more importantly there have been some much more experienced listeners who have found it very good listening too. My friend Bob (who hasn't heard it yet) has done some modeling and says the modeling results are surprisingly good.
     I call it the "Free Lunch" - by re-using the 2A3 current for the driver, and by using a choke load, I am joking that you are getting some great benefits for "free" this way. The circuit is attached, although I have revised it as follows : the driver uses 2 x AA batteries for bias (which sounds MUCH better than R bypassed by C), the cathode R for the 2A3 is 4500 ohms, the resistor in series with the choke is 1K, and there is now a second choke between the two positive terminals of the dual 47uF cap (the B+ is taken from the SECOND 47uF section, of course.)
     I have experimented with "ultrapath" connection of the last power supply capacitor (connecting it between B+ and cathode of the 2A3 instead of to ground) but I only had a short

time to listen this way and I did not develop a preference either way.
    Thanks for all the interesting ideas,
-j

     As the point of the Safe Loftin-White article was safety, let's see what safety modifications can be made to your circuit. The resistor used in series with the choke is needed to define a voltage drop to bias the output correctly.  Unfortunately, if the input tube is jiggled or removed or even fails while the amplifier is in use, the amplifier can be damaged. Replacing the resistor with a zener will not help, as the zener can only define a voltage under load; no current, no voltage. On the other hand, replacing the resistor with a transistor or MOSFET or even a tube will help. These devices will lock the choke's nominal DC voltage to the desired bias voltage for the output tube, even if the driver tube is removed from the circuit. Since you are bold enough to use battery bias for the input stage, you might want to replace the resistor with a series of rechargeable 9 volt (7.2 volts, in reality) batteries. This would also define a fixed bias voltage that would be renewed when the amplifier is in use!