Tube vs. Solid State Preamp
in [Design]
|
Prev: How do you call
Next: Internet via Cellular
From: Tim Williams on 1 Apr 2010 14:01 "Joerg" <invalid(a)invalid.invalid> wrote in message news:81k39hF41iU1(a)mid.individual.net... > The TL598 has dead-time control but it's usually just too weak to drive > really fat MOSFETs. Got to follow it with some real gate drivers, or use a > better chip from LTC. Ah, but these *are* the gate drivers. http://myweb.msoe.edu/williamstm/Images/Highamp_PWM.png Deadtime on these means ~5x the components, for which I'd rather just use a quad of TC4420's. But these are half price, and at least as powerful (if not more so). > But a tantalum? I can already picture it, a sharp popping sound, some > hissing, an orange puff of smoke wafting through the room ... Ya, that was before I realized these things are a little bigger than your average CMOS logic. ;-o To be fair, despite the rated ESR, I think the transistors were burning most of the heat, not the tantalum. Maybe its ESR is way less than the datasheet claims. Power supply rail basically just looked like deltas, no ringing or recovery inbetween, which corresponds to my experience (i.e., a tantalum can be quite excellently modeled as ideal C + ESR + a few milimeters of wire, no spooky higher order elements). Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
From: Tim Williams on 1 Apr 2010 14:11 "Phil Hobbs" <pcdhSpamMeSenseless(a)electrooptical.net> wrote in message news:4BB4AA5D.4060007(a)electrooptical.net... >> ....I had wonderd if a device using positive ions might work... > > The mass ratio between even a proton and an electron is about 1800:1, > which makes holes in silicon look very speedy! I'm going to have to make one, if I ever get access to the equipment (and glassblowing expertise!) to do it. There's something fundamentally wrong about making a tube with gas, though. It can work, but it can't work very well. It's very difficult to pump the electrode section low enough so that it's only ions, no electrons and definitely no secondary ionization taking place. And as soon as those ions hit the plate (= cathode), they become hydrogen again. The plate material has to resist atomic hydrogen, which is rather acidic (literally, 100% Bronstead-Lowry acid). Plus, ions carry enough momentum to sputter the cathode, which is even worse. Even if your P-tube has a tranconductance ~2000 times lower than a 12AX7 (i.e., about 1umho), it's worth considering because hey, a current mirror is a current mirror. Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
From: Tim Williams on 1 Apr 2010 14:12 "Jim Thompson" <To-Email-Use-The-Envelope-Icon(a)On-My-Web-Site.com> wrote in message news:6kj9r5l3fbp0bodgbg9n04r1t5cqa4f8sd(a)4ax.com... >>> A good jfet can have noise below 1 nv/rthz and Gm of 40,000 (in tube >>> terms). And very low gate current compared to a tube's grid current. >> >>FETs are the nearest things to superior valves. Interestingly they have >>gently curved characteristics like valves too - and they can be used >>without feedback (if you don't mind the spread of characteristics). > > All that attitude reflects is no knowledge of how to properly design > around BJT's. Admittedly, I've been doing it for 55 years, so _maybe_ > I've acquired some skill at it :-) Well, the man said FET, which is distinctly not a BJT. Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
From: Joerg on 1 Apr 2010 14:35 Jim Thompson wrote: > On Thu, 01 Apr 2010 10:30:26 -0700, Joerg <invalid(a)invalid.invalid> > wrote: > >> Jim Thompson wrote: >>> On Wed, 31 Mar 2010 18:04:35 -0700, Joerg <invalid(a)invalid.invalid> >>> wrote: >>> >>>> John Larkin wrote: >>>>> On Wed, 31 Mar 2010 13:36:54 -0800, Robert Baer >>>>> <robertbaer(a)localnet.com> wrote: >>>>> >>>>>> John Larkin wrote: >>>>>>> On Tue, 30 Mar 2010 08:56:08 -0700 (PDT), Chris >>>>>>> <christopher.maness(a)gmail.com> wrote: >>>>>>> >>>>>>>> With today's modern technology, is it possible to make a solid state >>>>>>>> preamp that is as quiet as a good tube pre? >>>>>>>> >>>>>>>> I am thinking about building a preamp. >>>>>>>> >>>>>>>> Thanks, >>>>>>>> Chris >>>>>>> Tubes are noisy. >>>>>>> >>>>>>> John >>>>>>> >>>>>> Try running a 12AU7 with a plate supply of 45V and grid-to-cathode >>>>>> voltage of about -1.5V (tweaked for zero grid current). >>>>> What are the voltage and current noise densities? >>>>> >>>> Don't know about the 12AU7 but the 7788 or in civilian E810F could get >>>> to around 1nv/rtHz when rigged as a triode (it wants to be a pentode in >>>> normal life). >>>> >>>> I wish Sovtek, Svetlana or one of those companies would make it. In an >>>> environment where a big zap to an input is sort of normal these are >>>> really useful. Even a real zinger with a blue flash inside the glass and >>>> an audible pop typically leaves a tube unfazed. >>> We've sort of shot ourselves in the foot with our desire to get >>> everything on a single chip... >>> >>> I'm often puzzled why a hybrid... tubes plus transistors... wouldn't >>> be the nicest solution. ... >> >> Well, that can most certainly be arranged: >> >> http://ixbtlabs.com/articles2/aopentube/ > > That's just plain dumb audiphoollery. > But it'll sure turn some heads during a party :-) >> >> I've actually made a very HV transistor by >>> using a tube with a grounded grid and an NPN in the cathode path. > > I needed a 1000V device with characteristic curves much like an ideal > CMOS transistor... no ability to get feedback in there. Piece-a-cake > with an NPN in the cathode circuit of a pentode (with local feedback). > 1kV is a white knuckle ride with most tubes unless you use a ballast triode or something. Although most tubes will stomach the voltage even if the abs max says 700V or 800V. >>> Size? Ever see a Nuvistor ?:-) >>> >> Yup, used them. In fact, one of my RF tools here (a dipmeter) has one in >> there except they called them acorn tube in them's days. >> >> >>> And I can't remember what they were called... a stack of ceramic >>> spacers and metal grids... really small. One jokester had the >>> "heater" from a blowtorch ;-) >>> >> That I don't remember. I'm not _that_ old :-)) > > Yeah, You are! You just don't admit it :-) > But it was you who called me young buck and yourself old buck :-) -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Joerg on 1 Apr 2010 15:26
Tim Williams wrote: > "Joerg" <invalid(a)invalid.invalid> wrote in message > news:81k39hF41iU1(a)mid.individual.net... >> The TL598 has dead-time control but it's usually just too weak to drive >> really fat MOSFETs. Got to follow it with some real gate drivers, or use a >> better chip from LTC. > > Ah, but these *are* the gate drivers. > http://myweb.msoe.edu/williamstm/Images/Highamp_PWM.png > Deadtime on these means ~5x the components, for which I'd rather just use a > quad of TC4420's. But these are half price, and at least as powerful (if > not more so). > That's MOSFET pairs, not gate drivers, that won't improve transitional things all that much. TC4420, now that's what I'm talking about ;-) >> But a tantalum? I can already picture it, a sharp popping sound, some >> hissing, an orange puff of smoke wafting through the room ... > > Ya, that was before I realized these things are a little bigger than your > average CMOS logic. ;-o > > To be fair, despite the rated ESR, I think the transistors were burning most > of the heat, not the tantalum. Maybe its ESR is way less than the datasheet > claims. Power supply rail basically just looked like deltas, no ringing or > recovery inbetween, which corresponds to my experience (i.e., a tantalum can > be quite excellently modeled as ideal C + ESR + a few milimeters of wire, no > spooky higher order elements). > My experience with tantalums is that they are excellent in terms of electrical behavior. Until one fine day ... *PHOOMP* -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM. |