On the other hand, in the single-ended amplifier the output tubes must see an idle current equal to the peak output current; for example, if the peak output current swing is 2 amperes, then the idle current must equal 2 amperes, with eight output tubes in parallel, each tube must draw 250 mA at idle. 
    The problem an SE para-feed OTL is that 40 watts of output implies 3.16 amperes of peak current, which in turn implies 400 mA of idle current per output tube in the single-ended amplifier. This current against the B+ voltage of 170 VDC equals 68 watts of dissipation! If we back off the idle current, the output wattage falls quickly, as halving the idle current quarters the output wattage.
    So eight EL509s would only yield about 10 watts of output power. Of course, this just might be the price we have to pay for a truly great sounding amplifier, possibly the best amplifier in the world. Still, the inefficiency glares.
    Bytheway, the coupling capacitor may not even be needed, if a negative power supply is used and the choke's DCR is sufficiently low enough. Remember a perfect inductor displaces no voltage, as it has no resistance. A real life inductors are made of wire that carries a resistance, but if the wire gauge is low enough, the DCR can be as low as 0.05 ohms, which will drop 0.2 volts with an idle current of 4 amperes, an amount of DC offset that is unlikely to cause too much trouble.
    Turning this idea on its head results in configuring the output tubes as cathode followers, which will greatly reduces the output impedance, the choke's required inductance, and the distortion. And in this topology, we can forgo the negative power supply.
    So in sum, I do not think that you and those readers in need of more power are going to be happy this single-ended OTL. The advantage of the push-pull amplifier is that it allows Class-AB operation of the tubes, which greatly increases the life span and power output of the devices. Maybe MOSFETs in place of the EL509 would be a more practical choice in a para-feed.

Subject: Solid-state 6L6
     Have you ever seen a "solid-state 6L6" device that could be plugged into a tube socket?
     I picture a circuit that would use high-voltage HEXFETs or something like that.

Chuck
?

    The April 2001 issue of Electronics World holds an article on such a device. Although 1,000 volt MOSFETs are available, my fear is that the MOSFET will die when it sees an over voltage due to inductive kickbacks form the reactive part of the loudspeaker.

Subject: EL509
    I read your magazine with much attention and it is indeed a follow-up for glass-audio. Data sheets for the El 509 you can find on
    www.machmat.com ( a Dutch tube maniac)
I saw there the triode-connected 509 and I didn't like what I saw. Very strange curves. I bought some Siemens Pl 509, just for experimenting. I think the EL 504 is a similar tube of half the power but more audio quality.
     I read about your idea of getting the output power from the grid. Half a year ago I read something about this on the internet, but I don't know where.
    They said the grid could be used for a low impedance output. The grid dissipation is smaller than the plate but with lower voltage you could get the same output current. Sorry but you are not the first with this idea.
    Keep on writing about tubes and I'll keep on reading it.

Steven
Netherlands

     Yes, I agree the EL509 curves do look strange. The problem with testing a high current tube like the EL509 is that the curve tracer must be up to the task. Maybe some of the wobbling at the end of the grid lines at the higher voltages

< PREVIOUS

www.tubecad.com   Copyright © 2001 GlassWare   All Rights Reserved

NEXT >

Pg.

17