Still a negative voltage regulator has it merits. In fact, I have meant to write an article just this topic since the first issue. So rather than short change the topic with a short reply, I will definitely write that article for the next issue.
    But here is a teaser to hold you until then.

    I found one article on the net (I guess it was published in GlassWare Tube Circuit of the Month archive.), which I think is a perfect design for this purpose. The circuit uses a floating regulator, which then drives a high voltage MOS-n-FET. The problem is that the transistor would probably dissipate quite some power (approx. 150-250 mA for the TL/10, and almost 500 mA for other amp at 50V drop), so it
would need a large heat sink. As the drain of the transistor is almost always connected to its package (i.e. TO-3, TO-220 etc.) that means that
the whole heat sink would be at around 300V above ground (which is the chassis of the amplifier). Therefore I'd have to isolate the heat sink from the amplifier chassis, which is a great disadvantage.
    A negative voltage regulator could solve the problem. What I have in mind is to use some lower power MOSFET (which would then also have lower gate-to-source capacitance) as a driver for a NPN high voltage bipolar transistor. (I guess I could also use a n-channel MOSFET alone.) Almost all of power NPN-s (MOS-n-FETs) have their collector (drain) connected to the package, and the collector (drain) would of course be at the ground potential. Therefore I wouldn't have to isolate the heat sink from the
chassis, the amp could then be much smaller. I could then also use the package of the NPN (MOS-n-FET) as the grounding point.
     As I lack the knowledge to design such a regulator on my own, I am asking you for recommendations. I also think that such a regulator could be cheaper, smaller and offer better regulation than those expensive high current - high inductance chokes.
Damir,
Slovenia

    First of all, I must point out that any pass device can be electrically isolated from the heat sink with mica washers and plastic screws. Furthermore, many MOSFETs come entirely encased in plastic. In other words, isolating the pass device from the chassis is not big of a deal.

Subject: Thanks
    Thank you for your efforts and thanks for a  great moment in my life. I started reading the Tube CAD Journal a little over a year ago. It was all over my head, but I read all the articles all the same and I have printed every issue so I can read them offline. At first the articles made as little sense to me as did all the articles in Glass Audio. (I could make sense of many of the articles in Sound Practices, but unfortunately I believe that was due to many of the authors actually knowing as little about tube electronics as I did.) But after rereading your articles, something different happened: I finally got it! I kept reading the words "conduction" and "current" over and over again. Then I realized that it is current that makes every thing work in electronics. Now voltage takes a back seat in my mind. I know you have stressed this point several times, but it took a lot to push voltage away from the center stage.
    For example, I could not understand how a cascode circuit worked because as I saw it the bottom tube's B+ voltage was locked by the top tube, so there could be no amplification. The fact that there is amplification and a lot of it bothered me to no end. Then I got it. The top tube's plate resistor is the same current path as the top and bottom tubes. Bingo. Amplification.

< PREVIOUS

NEXT >