Tube vs. Solid State Preamp
in [Design]
|
Prev: How do you call
Next: Internet via Cellular
From: John Larkin on 1 Apr 2010 11:29 On Thu, 1 Apr 2010 09:33:06 +0100, adrian(a)poppyrecords.invalid.invalid (Adrian Tuddenham) wrote: >George Jefferson <George(a)Jefferson.com> wrote: > >> "Adrian Tuddenham" <adrian(a)poppyrecords.invalid.invalid> wrote in message >> news:1jg8ngr.1g7xp3le9s3eyN%adrian(a)poppyrecords.invalid.invalid... >> > �Leo� <leo2100(a)gmail.com> wrote: >> > >> > [...] >> >> This happens so because the transistors are more ideal amplifying >> >> devices than tubes are...so the signal that comes in is the one that >> >> comes out, with as little distortion as possible. Since the tubes are >> >> less ideal, they add different components to the original signal... >> >> this have turned out to be more pleasant to the human ear. >> > >> > I would disagree with this statement. The curvature of the transistor >> > characteristic is much sharper than many valves and the harmonics >> > generated are much nastier-sounding. Valve stages with no feedback are >> > quite common and the distortion they generate with moderate signal >> > voltages is quite tolerable, but transistor stages with no feedback are >> > virtually unusable for audio. >> > >> > Transistor amplifiers can contain more devices in less space for the >> > same price and that allows them to employ much more feedback than valve >> > ones, not just overall but within individual stages; this is what gives >> > the improvement in overall linearity which you have erroneously >> > attributed to the devices themselves. The down side is that when they >> > eventually overload, they do so much more sharply and generate higher >> > harmonics, which sound vile. >> >[...] > >> A tube circuit can never beat a properly designed solid state circuit for >> distortion free sound. > >The big difference is that those valve circuits generally achieve their >performance with far fewer active devices and far less feedback, which >was the point I was trying to make. Valves are not 'less ideal' as >amplifying devices; you can achieve good results with fewer of them than >if you tried to achieve the same thing with transistors. > >If you pile more and more transistors into a circuit with more and more >feedback you will get a lower THD figure than with any practical valve >amplifier. That is not because of the inherent 'betterness' of the >transistors themselves but because they are cheap enough and small >enough to allow circuit topologies that wouldn't be worthwhile to >produce with valves.. 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. John
From: Chris on 1 Apr 2010 12:13 On Mar 31, 5:16 pm, "George Jefferson" <Geo...(a)Jefferson.com> wrote: > "Adrian Tuddenham" <adr...(a)poppyrecords.invalid.invalid> wrote in message > > news:1jg8ngr.1g7xp3le9s3eyN%adrian(a)poppyrecords.invalid.invalid... > > > > > «Leo» <leo2...(a)gmail.com> wrote: > > > [...] > >> This happens so because the transistors are more ideal amplifying > >> devices than tubes are...so the signal that comes in is the one that > >> comes out, with as little distortion as possible. Since the tubes are > >> less ideal, they add different components to the original signal... > >> this have turned out to be more pleasant to the human ear. > > > I would disagree with this statement. The curvature of the transistor > > characteristic is much sharper than many valves and the harmonics > > generated are much nastier-sounding. Valve stages with no feedback are > > quite common and the distortion they generate with moderate signal > > voltages is quite tolerable, but transistor stages with no feedback are > > virtually unusable for audio. > > > Transistor amplifiers can contain more devices in less space for the > > same price and that allows them to employ much more feedback than valve > > ones, not just overall but within individual stages; this is what gives > > the improvement in overall linearity which you have erroneously > > attributed to the devices themselves. The down side is that when they > > eventually overload, they do so much more sharply and generate higher > > harmonics, which sound vile. > > Most amplifiers are not meant to distort. Distortion is generally a bad > thing in 99% of applications excluding intentional use for audio effects. If > a solid state amplifier is distorting and not suppose to then it is because > the circuit was designed improperly or it is not being used properly. > > A tube circuit can never beat a properly designed solid state circuit for > distortion free sound. For audio, this is not necessarily good as I > mentioned in my original reply. > > A tube amp emphases the even harmonics which are octaves of the > fundamental(except for one 5th and up to the 8th harmonic). The odd > harmonics emphasize a dom7th chord sonority which means that if you play any > chords you are effectively playing a collection of dom7th sonorities(almost > anyways). This means the distortion will sound harsh no matter what you play > as there will be a great amount of dissonance created. Good for metal but > not for a warm sound. > > Again, in a properly designed solid state amplifier one will not have > distortion(well, the THD will be extremely low to effectively be 0) and a > tube amp will always have a higher THD(but designed properly it will still > be low). The problem with tube amps is that in most basic amplifiers the > load line vs the plate characteristics are non-linear over a large range > which will introduce a compressive effect which creates distortion. > Transistors have a similar effect but much more linear. > > People generally prefer tubes because when they do distort they distort > "nicely". Again, the point is, in a properly designed system where > distortion is meant to be minimized you cannot beat solid state. The problem > is that a SS amp may do it's job too well resulting in a dry unmusical > sound. The tube amp will color the sound in a good way resulting in a more > musically pleasing experience. > > If you neglect the musical aspects of amplification SS amp's win hands down > in every aspect except possibly power handling(but not so much any more). > Again, it depends on the specifics but this is the general case. Tubes are > generally better for musical applications in most cases. > > The good thing is, we can use both types to get the best of both worlds. > When we require absolute accuracy in amplification we use solid state. When > we want the musical effect we use tubes. We might, for example, choose to > use solid state amplification for a mixing desk's inputs and it's outputs > are tube based. Or simply have one final tube for the output(PA or > whatever), which offers less control over the individual tracks but requires > far less tubes. > > An analogy is incandescent and LED's. In fact, this is much more than an > analogy as they both operate from the same physics. I play guitar and mix sound. All of the tube amps I have tried for the guitar have a more pleasing sound by a long shot, it is nice to learn about the Fourier structure that you have shared. It quantifies and solidifies what my music ear has been hearing this whole time. Moreover, the tube amps that I have tried and own have virtually no noise coming from them when the guitar is quiet. SS amps are hissing like a pissed off snake. Regards, Chris Maness
From: Adrian Tuddenham on 1 Apr 2010 12:43 John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: > On Thu, 1 Apr 2010 09:33:06 +0100, adrian(a)poppyrecords.invalid.invalid > (Adrian Tuddenham) wrote: > > >George Jefferson <George(a)Jefferson.com> wrote: > > > >> "Adrian Tuddenham" <adrian(a)poppyrecords.invalid.invalid> wrote in message > >> news:1jg8ngr.1g7xp3le9s3eyN%adrian(a)poppyrecords.invalid.invalid... > >> > �Leo� <leo2100(a)gmail.com> wrote: > >> > > >> > [...] > >> >> This happens so because the transistors are more ideal amplifying > >> >> devices than tubes are...so the signal that comes in is the one that > >> >> comes out, with as little distortion as possible. Since the tubes are > >> >> less ideal, they add different components to the original signal... > >> >> this have turned out to be more pleasant to the human ear. > >> > > >> > I would disagree with this statement. The curvature of the transistor > >> > characteristic is much sharper than many valves and the harmonics > >> > generated are much nastier-sounding. Valve stages with no feedback are > >> > quite common and the distortion they generate with moderate signal > >> > voltages is quite tolerable, but transistor stages with no feedback are > >> > virtually unusable for audio. > >> > > >> > Transistor amplifiers can contain more devices in less space for the > >> > same price and that allows them to employ much more feedback than valve > >> > ones, not just overall but within individual stages; this is what gives > >> > the improvement in overall linearity which you have erroneously > >> > attributed to the devices themselves. The down side is that when they > >> > eventually overload, they do so much more sharply and generate higher > >> > harmonics, which sound vile. > >> > >[...] > > > >> A tube circuit can never beat a properly designed solid state circuit for > >> distortion free sound. > > > >The big difference is that those valve circuits generally achieve their > >performance with far fewer active devices and far less feedback, which > >was the point I was trying to make. Valves are not 'less ideal' as > >amplifying devices; you can achieve good results with fewer of them than > >if you tried to achieve the same thing with transistors. > > > >If you pile more and more transistors into a circuit with more and more > >feedback you will get a lower THD figure than with any practical valve > >amplifier. That is not because of the inherent 'betterness' of the > >transistors themselves but because they are cheap enough and small > >enough to allow circuit topologies that wouldn't be worthwhile to > >produce with valves.. > > 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). -- ~ Adrian Tuddenham ~ (Remove the ".invalid"s and add ".co.uk" to reply) www.poppyrecords.co.uk
From: Jim Thompson on 1 Apr 2010 12:51 On Thu, 1 Apr 2010 17:43:19 +0100, adrian(a)poppyrecords.invalid.invalid (Adrian Tuddenham) wrote: >John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: > >> On Thu, 1 Apr 2010 09:33:06 +0100, adrian(a)poppyrecords.invalid.invalid >> (Adrian Tuddenham) wrote: >> >> >George Jefferson <George(a)Jefferson.com> wrote: >> > >> >> "Adrian Tuddenham" <adrian(a)poppyrecords.invalid.invalid> wrote in message >> >> news:1jg8ngr.1g7xp3le9s3eyN%adrian(a)poppyrecords.invalid.invalid... >> >> > �Leo� <leo2100(a)gmail.com> wrote: >> >> > >> >> > [...] >> >> >> This happens so because the transistors are more ideal amplifying >> >> >> devices than tubes are...so the signal that comes in is the one that >> >> >> comes out, with as little distortion as possible. Since the tubes are >> >> >> less ideal, they add different components to the original signal... >> >> >> this have turned out to be more pleasant to the human ear. >> >> > >> >> > I would disagree with this statement. The curvature of the transistor >> >> > characteristic is much sharper than many valves and the harmonics >> >> > generated are much nastier-sounding. Valve stages with no feedback are >> >> > quite common and the distortion they generate with moderate signal >> >> > voltages is quite tolerable, but transistor stages with no feedback are >> >> > virtually unusable for audio. >> >> > >> >> > Transistor amplifiers can contain more devices in less space for the >> >> > same price and that allows them to employ much more feedback than valve >> >> > ones, not just overall but within individual stages; this is what gives >> >> > the improvement in overall linearity which you have erroneously >> >> > attributed to the devices themselves. The down side is that when they >> >> > eventually overload, they do so much more sharply and generate higher >> >> > harmonics, which sound vile. >> >> >> >[...] >> > >> >> A tube circuit can never beat a properly designed solid state circuit for >> >> distortion free sound. >> > >> >The big difference is that those valve circuits generally achieve their >> >performance with far fewer active devices and far less feedback, which >> >was the point I was trying to make. Valves are not 'less ideal' as >> >amplifying devices; you can achieve good results with fewer of them than >> >if you tried to achieve the same thing with transistors. >> > >> >If you pile more and more transistors into a circuit with more and more >> >feedback you will get a lower THD figure than with any practical valve >> >amplifier. That is not because of the inherent 'betterness' of the >> >transistors themselves but because they are cheap enough and small >> >enough to allow circuit topologies that wouldn't be worthwhile to >> >produce with valves.. >> >> 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 :-) ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | The only thing bipartisan in this country is hypocrisy
From: Joerg on 1 Apr 2010 13:26
Tim Williams wrote: > "Joerg" <invalid(a)invalid.invalid> wrote in message > news:81hoh3Fb4tU1(a)mid.individual.net... >> UC3842? How did you get PFC and regulation done with just one chip and one >> inductor? > > Nope, but close, FAN7527. Basically a 3842 with a multiplier after the > error amp, with a beefier output and more protection circuitry (over/under > voltage, overshoot, etc.). > http://myweb.msoe.edu/williamstm/Images/Highamp_PFC.png > Pretty much the datasheet circuit. Haven't tested the aux supplies yet > though (notice D5 and C13 missing on the board, and only two windings on the > inductor). > > Pretty neat, worked first time, exactly as indicated. Overvoltage > protection works too -- kind of unnerving to hear it chatter for the first > ~100ms before the error amp pulls in. They really should've built a proper > control loop instead of adding that cheap hack. > > It's a little unstable, you can nudge the line or load with a transient and > see it wiggle for about a second (at ~10Hz, so a Q of 5ish?). It's not > unstable, the voltage is always 410V give or take ripple, but it's not the > perfectly damped response I like. I'll get in there and tweak the feedback > loop, maybe add a 'speedup' capacitor for some derivative action. > > Tonight I tested half the PWM board that it powers... > http://myweb.msoe.edu/williamstm/Images/EE409_PWM3.jpg > ...I discovered these little bastards sucking over 300mA while the TL598 is > running. It seems that, in the ~40ns it takes for the '598 to transition, > these things are each drawing a peak of 10A or so (the tantalum is specified > as ~2.5 ohms ESR, so most of the current is probably coming from the 0.1uF > MLCC over by the '598). Multiply by 4, then 2, then 100kHz, and you get: > lots of heat. > 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. But a tantalum? I can already picture it, a sharp popping sound, some hissing, an orange puff of smoke wafting through the room ... >>> Now I just need to get rid of that masking tape! ;) >> Whoops, did it bake on? > > No, or not yet... it's kind of ugly though. > > Need to replace the hookup wire winding with proper litz, too. > I have one transformer here where the winding is square copper, wrapped in lacquer-drenched cloth, then hammered in place around the bobbin. Lots of amps. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM. |