Hello All Receptors of Quadraphony,
I am back from the recording studio and the West Coast Songwriters yearly conference which my equipment company sponsors each year (this was the 12th). So let us get down to business!
watsontr,
The turntables and internal wiring do not cause problems with CD-4 reproduction. The biggest problem is the RCA interconnecting cable. Some older cables had high capacitance. At a termination resistance of 50,000 Ohms this can cause some roll off of the sub-channel frequencies (20 KHz to 45 KHz). Low capacity cable is available from companies like Digikey. I will post some part numbers in my next post for any one who wants to replace their interconnecting cables. This by the way is real cable and not the overpriced, under performing junk like "Monster Cable" and all of its clones.
kfbkfb, Lucanu, Doug G., ress4278, Quadro-Action, bmoura, proufo, Quadfather,
For the time being I will let the speculation about a CD-4 cutting system remain with you all. By the way all of the CD-4 patents including mine have run out. All of the current info is now public domain.
Now it is time for Technical Soap Box.
Todays subject is Tubes vs Transistors, the myths are exposed!
This controversy has been going on since semiconductors overtook and replaced thermionics. The statements of warm tube sound vs harsh transistor sound only have validity based on the operation of the devices. In this post I will provide the conditions, the facts, and the realities of this controversy, and set the technical record straight.
What is the real controversy? It is how the devices sound when they are not operating in the linear portion of their transfer characteristics (transfer characteristic are defined as the change to the output of a device for a given change to input to that device).
What is an amplifier? Many think that an amplifier is a box full of tubes or semiconductors ( I use semiconductors instead of transistors because today’s technologies use integrated circuits which have large amounts of transistors, resistors, diodes, etc on a single silicon chip) that takes a small signal at the input and makes this small signal a large one at the output. That is in fact the result of what the amplifier does but not quite how it does it.
A basic amplifier consists of two sections. They are the power source and the control circuit. The power source has the ability to deliver the maximum output energy (voltage, or power- tubes are voltage amplifiers) that the amplifier system is rated for. The control circuit varies the output energy based on the input signal variations. The British appropriately named the vacuum tube a “valve” which is a good description of what it does. As the small input signal varies the “valve” controls the flow of energy from the power source to the load (load being a resistor, headphone, speaker). The gain or amplification is the ratio of the small input variance to the large output variance. As a metaphor, a small varying stream of water on the control valve will create a large varying stream of water from the reservoir. Transistors work in the same way but are current, instead of voltage amplifiers.
If you have a tube amplifier and a semiconductor amplifier side by side, and both amplifiers are being driven in their linear regions, the resulting output from either amplifier will be indistinguishable. If, however, you force the amplifiers into the non-linear portion of their transfer characteristics, then there is a perceptible difference in the sound. Why is this you ask? It is the way each device handles non-linear distortion.
As a vacuum tube approaches the non-linear portion of it’s transfer characteristic, the output no longer changes in a linear fashion with the input. Here is an example. Let us say that you have a tube amplifier that has a gain of 10. That means that for every small change on the input voltage, the output voltage change will be 10 times larger. A 100 millivolt (.100 Volt) input will produce a 1 Volt output, 200 millivolts (.2 Volts) in = 2Volts out, 300 millivolts (.3 Volts) in = 3 Volts out etc. Now let us assume that the linear portion of this tube amplifier stops at 5 Volts output. The numbers would look like: 500 millivolts (.5 Volts) in = 5 volts out, 600 millivolts (.6 Volts) in = 5.4 Volts out, 700 millivolts (.7 Volts) in = 5.8 volts out and so on. You can see that the output change becomes gradually less as the input continues to increase. The power source still has more energy to deliver but the control valve does not have the control range to deliver it. This is a non-linear characteristic called saturation. Saturation is the point on the transfer characteristic of the tube where the output can no longer follow the input in a linear fashion. This condition creates a gradually larger amount of even harmonic distortion. The sound of this type of distortion is quite pleasing and the term “warm” has been given to it.
Lets now look at a semiconductor amplifier that has a gain of 10. For every small change in the input current, the output current change will be 10 times larger. A 100 milliamp (.100 Amp) input will produce a 1 Amp output, 200 milliamps (.2 Amps) in = 2Amps out, 300 milliamps (.3 Amps) in = 3 Amps out etc. Now let us assume that the linear portion of this semiconductor amplifier stops at 5 Amps output. The numbers would look like: 500 milliamps (.5 Amps) in = 5 Amps out, 600 milliamps(.6 Amps) in = 5 Amps out, 700 milliamps (.7 Amps) in = 5 Amps out and so on.
As you can see this non-linear distortion is different. Any current from 500 milliamps and above yield the same output current. This non-linear characteristic is called clipping. The power source energy limit has been reached but the control valve could control more energy if it was available. This condition creates an immediate large amount of odd harmonic distortion. The sound of this type of distortion is very irritating and the term “steely” has been given to it. It is almost like fingers on a chalkboard.
Now all of this information should tell you one thing. Don’t run your amplifiers, tube or semiconductor in their non-linear transfer characteristics. After all of this which type of amplifier is actually better. When it comes to linear operation semiconductors have much better performance then do tubes. High fidelity means sound reproduction that is transparent and with out coloration from the electronics used to record and playback the sound. Today’s semiconductors are capable of at least 4 orders of magnitude (10,000 times) lower distortions (Total Harmonic Distortion, THD, Inter-modulation Distortion, IM, and Transient Inter-modulation distortion, TIM). Noise levels of these same semiconductors can be 6 orders of magnitude (1,000,000 times) lower then the best tubes. The key to obtaining this real time performance is how to operate your amplifier.
One characteristic of music that should be understood by anyone interested in the best high fidelity playback is dynamics. The most important is crest factor which is a fancy name for Root Mean Square (RMS) to Peak ratio. While a conventional Sine wave has a crest factor of 1.414 to 1 or 3 dB, music has a maximum crest factor of 10 to 1 or 20 dB.
So what does this mean to you? If you want to reproduce your music at a power level of 20 Watts and do it in linear portion of the amplifiers transfer characteristic, you need to have a 200 Watt amplifier. This is because the music peaks at 20 Watts average will be 200 Watts. Any peaks that exceed 200 Watts will be in the non-linear portion of the amplifiers transfer characteristic and will distort those peaks. I like to have a little headroom to make sure I stay in the linear portion. For this example I would use a 250 Watt amplifier for 20 Watts linear.
So what is happening in the amplifier real world? First let me divide the amplifier real world into two segments. Segment one is high fidelity playback monitoring and recording, while segment two is music instrumental amplification. As shown in the previous paragraph, the highest fidelity performance is obtained by operating your playback monitoring amplifier in the linear range. If you maintain this type of operation, semiconductor amplifiers are the best performing especially phono cartridge playback, microphone pre-amplifiers, mixers, and power amplifiers.
There are some audio “experts” who cheat. They do not maintain the music crest factor of 20 dB and hide it by using tube amplifiers. As pointed out earlier, tube amplifiers saturate gradually, which means the non-linear distortion rises slower then semiconductor amplifiers which don’t saturate but clip. Even with this cheating the linear performance is no where as good as the semiconductor amplifier.
What about musical amplifiers? Let me put on my recording engineer hat for a moment. Tubes prove to be better in this application because the distortion they create is part of the sound effect for the instrument. When I record an amplified music instrument, I put a pickup microphone in front of the amplifier cabinet so that I capture the total effect of the instrument, the effects pod, and the amplifier and speakers. The tube saturation is a large part of the instrument sound.
The conclusion is that for high fidelity monitor playback or recording, use a semiconductor preamp and amplifier operating in the linear portion of the transfer characteristic and observe the crest factor to maintain ultra high fidelity. For amplified music instruments, tube amplifiers provide a great effect.
kfbkfb
Kirk, CD-4 cartridges that worked well had no resonance in the sub-channel frequencies but above (50 to 60 KHz) with amplitude variance of 3dB in the sub-channel band and a Bessel phase response. To achieve this the cantilevers were made smaller in diameter and shorter to reduce the mass. Shure and Satin were far from the state of the art cartridges at the time. High Fidelity Magazine got that information from me during and AES session I gave on CD-4 demodulator playback.
Well folks that is all for now,
Lou Dorren