KLUDGE@AGCB1.LARC.NASA.GOV (12/06/90)
I can feel the beginning of a religous war here. - There is no reason for tube amps to be better than transistor amps. - There is also no reason for transistor amps to be better than tube amps. You can't make generalizations like this, because you are comparing not only different active devices, but completely different designs. What you can do is make statements like: - Some tube amps sound better than some transistor amps. - Some transistor amps sound better than other tube amps. Even this is not general enough. My Citation I amp (tube) sounds fantastic driving my Magnepans, but when I try to drive a friend's pair of Infinities, it doesn't sound as good as his Rowland Research amp. But on the other hand, his amp sounds pretty limp driving my speakers, when compared with mine. - Tube amps have output transformers. This makes the output stage more isolated from the load. They can drive very reactive loads without difficulty, and if equipped with the proper taps, can drive the very low impedance speakers that are becoming popular today with ease. - Transistor amps have no output transformers. Their bass response is usually much more solid because there is no transformer to roll it off. Also the overall distortion tends to be much less because the transformer nonlinearities are avoided. (There are a few exceptions to this... transistor amps with transformers and tube amps without them, which have their own tradeoffs). Buy a tube amp. Or buy a transistor amp. I bought my amp because it was very good and very cheap, and picked a pair of speakers to match it. You might want to do the opposite. Just whatever you buy, make sure it sounds good. --scott
bill@vrdxhq.verdix.com (William Spencer) (12/20/90)
Sure, I have $.02 to offer on this subject. First, something should be pointed out: there are really only 3 kinds of active devices used (typically) in audio electronics. They are tubes and the two transistor families, FET and bipolar. That is all. When you realize the choice is that limited you also may realize they all must be considered. There are, of course, variations of the major device types. Transistors can be made in positive and negative polarities -- one advantage. Tubes have different types according to number of terminals -- triodes, pentodes, etc. And they all have countless variations of geometry and power handling. FET's and bipolar transistors do deserve seperate consideration as major device types because their principles of operation are very different. Tubes have a relatively linear voltage transfer characteristic. This comparison is slightly unfair however. With tubes the output current is reduced when the output voltage swings (in a voltage amplifier configuration). This is essentially a negative feedback effect of sorts. All the devices have a nonlinear input voltage to output current response. The tubes require less external negative feedback for linear response. Remaining distortion will be a gradual nonlinearity spread over the full voltage swing. This translates to larger amounts of lower order (frequency) distortion components and little higher order harmonics. Transistors are easier to use in more complex circuits with more open loop gain and more negative feedback. It is intuitive that the more complex circuit is more likely to muck up complex audio information although it is hard to explain exactly why this happens. With proper design, that is. Bipolar transistors have one troublesome characteristic that niether FET's nor tubes have: signifigant and nonlinear input current in normal operation. In other words, the transistor can distort its own input. Complex. Tubes run at high voltage (typically) and produce high impedance designs. Tube preamps have lots of headroom typically. Input capacitance, for example, is very low with tubes, but so is output current, so comparison is not trivial. Tube-transistor hybrids are tricky. It's hard to find a MOSFET with high voltage handling but designed for low power (power MOSFET's are oriented toward switchmode power supplies more than audio). Even a low power MOSFET will have far more input capacitance than a tube (typically a few pF). And tubes are quite rugged even in low power sizes while you have to be careful with FET's. While the power to input capacitance ratio of MOSFET's is comparable to tube triodes niether compares to that of pentodes. Not all the differences between amplifiers can be attributed to device type. The circuit configurations generally used are very different. Due to high impedance and voltage transformer coupling is used (usually)when tubes must drive low impedance loads (speakers). Since only one device polarity is availiable a balanced transformer is used when push-pull operation is desired. The transformer is a major disadvantage (especially for bass) and expense. However, the balanced configuration can be much more precisely balanced than a push-pull transformerless design. Also, power supply ripple cancels. A transformer is an ultimate DC blocking device so biasing problems are much less critical. Servo loops and speaker protection are not used. Also tube amps are generally biased closer to class A than the majority of transistor amps. This only touches on some of the main differences. Technical differences definitely exist and the comparison is not at all one sided. Part of the choice depends on application. All things equal though, you might prefer transistors to tubes due to reliabilty and maintainance. The lifetime of tubes is good, however. BTW, some of the reports of trouble with VTL's are probably due to tubes damaged during shipping. Tube sockets can also be as much trouble as the tubes themselves. bill S.
chu@ims.alaska.edu (Chirk Chu) (12/26/90)
Oh, sometimes I wished I have my ADCOM 555 back! Power here in Fairbanks, Alaska is not the best in the world. The other day the power went down just for a fraction of a second. The resulting power surge promptly killed a driver tube (6201) and a output tube (EL34) in one of my VTL 225W monoblock. The VTLs were plugged into the wall as recommended by the seller. In fact a line conditioner was not recommended because it would limit power as the monoblocks demanded alot. Anyone has a suggestion for this problem other than spending $2K on a pair of Tice Power Blocks? Transistors are more reliable. Do not get tube amps unless one is willing to give plenty of TLC to your amps. Tubes also need replacement every 1-3 years and they are not cheap. On a unrelated subject, I've been told that MC cartridge sound best at 70+ degrees. What is the reason? It is hard to keep the interior temperature of a house at this level when it is 40 below outside.
caf@omen.UUCP (Chuck Forsberg WA7KGX) (12/26/90)
In article <8488@uwm.edu> bill@vrdxhq.verdix.com (William Spencer) writes:
-All things equal though, you might prefer transistors to tubes due to
-reliabilty and maintainance. The lifetime of tubes is good, however.
I'd rather say, the lifetime of *some* tubes if good. I still remember
the short lifetime of EL34 tubes in a Heathkit 55 watt amplifier. I
cheated by backing off on the idling current. That saved money for LPs.
Come to think of it, those tube amps had some pretty spectacular
failure modes. I have never seen a solid state audio amp put on
a light show to match that.jhess@orion.oac.uci.edu (James Hess) (01/04/91)
In article <8547@uwm.edu> chu@ims.alaska.edu (Chirk Chu) writes: > >Power here in Fairbanks, Alaska is not the best in the world. The other day >the power went down just for a fraction of a second. The resulting power >surge promptly killed a driver tube (6201) and a output tube (EL34) in one >of my VTL 225W monoblock. > >Transistors are more reliable. Do not get tube amps unless one is willing >to give plenty of TLC to your amps. Tubes also need replacement every >1-3 years and they are not cheap. > Must of been one heck of a power surge! While I can't speak authoritatively on this specific design, in general Chirk is wrong--tubes are more tolerant of overloads and surges than transistors. Faults that would take minutes to seriously damage tubes will eat transistors in milliseconds. As to replacement, while I imagine golden-eared audiophiles will dissent, I have seen tubes, particularly in low-power sections of the circutry, operate for years with no measurable deterioration beyond the tolerance ranges of new components. Of course, if your circut requires devices to be matched within +/- 1%, plan on spending lots of money.... J. Hess
bill@vrdxhq.verdix.com (William Spencer) (01/07/91)
in article <8687@uwm.edu>, jhess@orion.oac.uci.edu (James Hess) says: > In article <8547@uwm.edu> chu@ims.alaska.edu (Chirk Chu) writes: >>Power here in Fairbanks, Alaska is not the best in the world. The other day >>the power went down just for a fraction of a second. The resulting power >>surge promptly killed a driver tube (6201) and a output tube (EL34) in one >>of my VTL 225W monoblock. > While I can't speak authoritatively on this specific design, in general Chirk > is wrong--tubes are more tolerant of overloads and surges than transistors. > Faults that would take minutes to seriously damage tubes will eat transistors ^^^^^^ > in milliseconds. I think you said it: Faults. It's very possible that an overload condition would not have resulted in a different amp design. Especially if this was a result of a power dip without overvoltage. Or it could have been that the tubes were ready to go. It seems that these higher power tube amps have a lot more trouble than low powered ones. bill S.