The Tube Circuit
When I first set out to design the tube circuit, I had just received a remarkable piece of computer software courtesy of my business partner Mike Tseng. This software called “Multisym” by Electronics Workbench enables anyone to “build” complete circuits and connect them in operational configurations and to measure various quantities like voltage, current, noise, etc.
One can input a signal into the circuit and measure it at any point and in any way using a variety of test equipment like voltmeters, ammeters, oscilloscopes and a distortion analyzer. One can connect as many as five voltmeters, ammeters and oscilloscopes simultaneously and display their output readings on the computer display.
The actual hardware components are contained within the confines of the computer and the real beauty of this simulation is that the electrical characteristics of the components themselves are derived from measuring actual real components and incorporating them into the simulation software. The non-linear aspects of each component is included so the simulation is a “real world” measurement, and is indistinguishable from where the user would build a real live circuit and perform the same measurements and tests using real live test equipment.
I have verified the accuracy of this software many times by building an actual circuit and comparing the measurements I obtain with real test equipment and comparing the results with that of the software results. The advantage here is considerable time savings in that I don’t have assemble and solder together test circuits, but rather I can do the equivalent assembly and test using the keyboard and mouse. What would take days or weeks can now be done in hours or less. I can make adjustments to circuit values and changes very quickly and measure the results of these changes instantly. I was able to “build” and test and finalize the design of the tube circuit in about two hours. The only thing I cannot do is listen to the finished circuit. But I know for certain which circuit I will actually build when the time comes. I also went back and adjusted the solid state circuit.
The first attempt at a tube circuit was a throwback to the basic topology used in the 1950’s and 1960’s in “classic” equipment. This was a very simple arrangement using one tube for the entire channel of the line stage (a 12AX7).The circuit consists of two common cathode amplifiers cascaded together. The only significant variation on the classic circuit arrangement is I DC-coupled the two cascaded amplifiers together rather than using an interstage coupling capacitor. This eliminates any of the effects of the coupling capacitor.
Using the software simulation program, this circuit exhibited exemplary square wave performance and consisted of two tubes and six resistors. It is considerably simpler than the solid-state circuit with about equivalent bench performance. There are however some disadvantages to using this type of circuit.
First, since there is no follower stage, the output impedance of this circuit is 3000Ω and it can deliver only 1 ma into a load. This means sonic performance is significantly compromised when using a solid-state amplifier having an input impedance of 20000Ω or less. Since many solid-state amplifiers on the market today have an input impedance of 20000Ω or less, the circuit is incompatible with those amplifiers.
Since user versatility is an important design precedence, it was decided to embellish the basic circuit to add more current gain to drive any power amplifier or cable length. I decided then to add another 12AX7 tube to the line stage design. This additional tube consists of two independent triodes, the first triode is a common cathode amplifier stage followed by a common anode unity voltage follower stage (cathode follower) which then drives the load. As per the previous 12AX7 elements, the just-mentioned are DC coupled for “no series capacitor operation. In fact, the complete line stage does not use any series coupling capacitor at all, which eliminates any low frequency time constant and its accompanying low frequency response roll-off.
The completed design uses a total of four stages. To insure proper internal as well as output square wave response, the overall feedback time constant is matched with the compensation time constant. This overall feedback design feature results in a distortion spectrum that is consistently the same regardless of measured frequency.
Some may ask why not use the circuit without negative feedback, because negative feedback is a bad thing sonically, right? Not necessarily. In my own investigation of amplifier systems in general, negative feedback becomes a negative configuration when the criteria mentioned previously in the article are not adhered to. With the assistance of the computer simulation, it is possible to adjust component values easily to ensure proper square wave response.
All one needs to do is look at the waveform. No other analysis is needed. This tube line stage has a gain of 20dB and an input impedance of 50000Ω. For the power supply, I use a three section filter capacitor arrangement featuring 5000MFD total supply capacitance for a very “stiff” DC supply reference for the circuit. Each electrolytic capacitor is bypassed with 1MFD polypropylene film capacitor in parallel with a .01 MFD polystyrene capacitor. This bypassing arrangement assures a smooth high frequency transient response and ensures low power supply impedance is maintained at high frequencies also.
I chose a 12AX7 tube mostly because it is listed in the database of the computer simulation software, and therefore an exact determination of circuit behavior could be reached. Of course, nothing is finalized and when the actual circuit is built, and a lot of tube rolling will occur during listening evaluations to see which tube is most appropriate. A question for DAGOGO readers would be to solicit opinions and input regarding tube selection. I therefore am asking DAGOGO readers who are interested to share their own experience with favorite tube selection and I will try each and every one suggested. Since this is a line stage amplifier, DC is used on the tube filaments.
Putting It All Together
As of the present day, the current state of affairs pertaining to the pre-amp design are as follows: Two circuits, one solid-state and one vacuum tube are designed “on paper” in that neither circuit has been actually built or listened to. I have yet to receive supporting commitments from my two partners concerning whether or not a pre-amp will be in the offing as a product line component for Luminance.
Luminance Audio is still in a relative infancy stage in terms of building market share, product name, and product acceptance in the audiophile community. It is my partner Mike Tseng’s opinion that in order for the pre-amp to be brought to market, the KST-150 power amplifier first needs to become a successfully, viable product entity before a pre-amp can be realistically considered for production. In addition, there is the considerable financial investment required for such a project to reach fruition. The investment level required would be comparable to that of the KST-150 amplifier.
While these considerations are non-technical in nature, they are crucial and important when the gestalt of the entire pre-amp project is considered. There are several questions and considerations which I would like DAGOGO readers to consider and express as feedback. The first is should the pre-amp come equipped with a phono stage as standard or as an option. My inclination is to incorporate a phono stage as a standard feature to “invite” audiophiles to investigate analog. What do you think?
Secondly, should the pre-amp come equipped for remote control operation? Should the remote control be universal? Should the pre-amp have single-ended or balanced inputs, outputs or both? I have omitted my design considerations on these topics because I don’t as of yet have a consensus from prospective audiophile customers. Luminance Audio would like to design a pre-amp incorporating the features and performance that audiophiles want rather than introducing a product under the “here it is” basis. We want the end user to be in control of defining what kind of product the pre-amp should be. Luminance Audio endeavors to be as accommodating and responsive to the end users needs as possible.
In a future article, the remaining items pertinent to a fully functional pre-amp will be covered. But first, I am requesting some reader input and feedback on anything that I have covered in the articles. I am all ears and welcome constructive criticism. The same offer is extended to DAGOGO’s editor also. In addition if any reader would like further elaboration on any topic in the articles, please feel free to ask and I will expound as necessary. Meanwhile, I would like to respectfully acknowledge Constantine Soo’s gracious invitation and opportunity to write this article, and I look forward to the development of an interactive dialog forum this article could generate.
- (Page 1 of 1)