We are here to shine Indoor From future-proof modular solutions to innovative new formats, we deliver high quality LED Indoor solutions at an affordable price. Because it has always been crucial to untie production from natural light. As a substantial slice of the operation costs paid for electricity. View all Outdoor LED Outdoor solutions deliver a light closer to natural daylight than the traditional lamps of the past. The latest LED lighting solutions provide an ideal upgrade path for public bodies looking to reduce energy costs and environmental impact. LED can make striking aesthetic improvements to landscapes and cityscapes.
|Published (Last):||10 October 2015|
|PDF File Size:||18.32 Mb|
|ePub File Size:||20.69 Mb|
|Price:||Free* [*Free Regsitration Required]|
He has been active in electronics since the age of five watching his dad build a Heathkit Williamson. Prior to starting Ram Tube Works and Music Reference in , he opened a high-end retail store where he met and later worked for Harold Beveridge. His other interests are music theory, piano, singing, building musical instruments, solar energy, architecture, and understanding how things work.
This article attempts to correct a few misconceptions about some of the popular dual triodes used in audio amplification. I have developed circuits for both home audio and musical instruments. I keep a little collection of these classics to test whether my equipment works with both old and new tubes.
Remember that the solo player can use distortion to enhance that instrument. Yet the same distortion, added again, destroys the reproduction of such a recording. Preamplification Once upon a time there was only the 12AX7.
This is not entirely true. Textronics and others used them in oscilloscopes and other high performance test equipment. I used 12 of them for my first commerical design effort with Harold Beveridge, Inc.
The RM-1 preamp that resulted had, and still has, the quietest phono stage extant with tube input. Its design led me to investigate the physics behind a low-noise tube. I also found that the 6DJ8 had better linearity lower distortion , wider bandwidth, and better ability to drive the output jack with a strong low-impedance signal, which is preferable when you use a long or high-capacitance cable. But a triode with high gm is not necessarily a high voltage-gain triode; the 6DJ8 has almost 10x the gm of the 12AX7, but one-third the mu voltage gain.
Transconductance in triodes and pentodes increases quite dramatically with plate current, which is a source of distortion because it increases as the signal swings negative and decreases as it swings positive at the plate.
This is the basic distortion mechanism in all active devices, vacuum and solid state. Two Other Factors In typical triode circuits mu becomes the controlling amplification factor or gain and is easy to spot in the data books since it has not units. Mu will vary with these, but not much or so hope those of us who desire low-distortion, nonfeedback amplifiers. Instantaneous voltage gain mu must be constant to achieve low distortion. When an amplifier is linear, its voltage gain is constant at all levels and frequencies of interest.
In the data books the mu value also assumes an infinite external plate resistance, which is the only way to isolate the tube from the effects of the circuit and provide a useful value.
The RCA Receiving Tube Manual includes a section, "Resistance-coupled Amplifiers," which includes the values of gain, maximum output voltages, and other interesting data for the popular voltage amplifiers. The "current source" camp approaches that ideal gain given by mu. My first RM-4 headamp prototypes contained quiet transistor current sources, but I abandoned them for low-noise resistive plate loads.
When you consider the load that follows the plate which we have tried to please with an infinite-impedance current source , much of the "infinite impedance" is lost. A high plate supply voltage can usually allow a plate resistor x the internal plate resistance rp of the tube, which is generally high enough; beyond that figure little is gained. Predicting Distortion In a graphical representation of these three qualities, gm rp and mu, as they vary with plate current, gm rises markedly with rising plate current, rp falls proportionately, and the product remains virtually constant.
Once I became aware of this, I knew what to look for in a low-distortion triode. Yet, I was amazed how difficult it was to find those three curves, which are not included in the RC series RCA manual or any poplar American manual, except the commercial Tung-Sol.
They are often contained in the European manuals, particularly those by Philips and Telefunken. I first seriously considered this question of simple distortion in in a single triode stage and realized that distortion is due simply to the change of mu with signal.
Building them in typical circuits confirmed this. Most designers like to talk about the even spacing of the plate curves, as given in RCA, as a measure of linearity, probably because that is all they have to work with. I concluded that for the small signals we see in preamplifiers the curves lacked the resolution to accurately predict the distortion. Remember that in a typical stage the plate current swings only a few percent. I believe they were hand drawn from a curve tracer, or by tracing a photograph of the screen.
Their main function is to allow circuit designers to figure voltage and resistance values to establish an operating point. Be careful with this method of analysis for predicting distortion. Only a year or two before I secured a reliable line of supply through their US office. When that line of supply dried up, I was forced to return to general distributors, who buy direct from the factory or through other distributors.
Basically, the tube distributor has a bit of a hard lot finding the most reliable quality and authenticity at the best prices. Factories often do not store inventories, as they will build even a popular type such as the 12AX7 only once or twice a year due to the time required to set up a type in the factory and retrain workers.
On my visit to the Yugoslavian factory Ei , I learned that assemblers required as many as six weeks to reach full speed on a type they had built before. When a factory such as Ei or Tungsram manufactures a popular tube, it generally has contracts for the majority of its production run. These firm orders plus inventory requirements determine the size of the run and the quantities of raw materials needed for that tube. Many of the materials are type-specific, that is, the diameter of grid wire, size of cathode sleeve, width of steel on the roll that becomes formed plates, and glass tubing diameter which is cut to length at the factory.
After these tubes roll off the line they are sent to various distributors, who have estimated their annual sales and ordered accordingly. Sales are often guesswork, customers come and go; one is fat on a product and the other is lean by the end of the year. Sometimes everyone runs out of a particular brand and is forced to go hunting.
Imposter Revealed By this time I knew how to identify a Tungsram tube by looking beyond the glass to the internals to identify the maker. I had seen too many Russian tubes marked "made in W. Germany" or "made in England" to be fooled by the paint on the glass.
Only Tungsram tubes, to my knowledge, had an easily visible, shinny silver square with a two-digit embossed number attached to the getter post. So, my tubes really looked like 6DJ8s but I noticed they acted a bit schizophrenic. In some circuits they behaved very much like 6DJ8s, but not in others. I suspected the difference had to do with plate current, so I rigged up my chart recorder to plot mu versus plate current and gm versus plate current.
In a line-up of four samples of four manufacturers, the fake market F. Tungs on the graph showed itself without a doubt Fig.
This impostor had a decidedly variable mu curve, which we do not want at all for audio. I encourage you to study the chart carefully. The top four curves show mu versus plate current from zero to 6mA. Note that mu is very constant from 0. The F Tungs is noticeable for its very low mu that rises from less than 10 at 0. After seeing these curves, I was quite comfortable running a real 6DJ8 anywhere between 0. I later found it wise to keep the dissipation down to 0. The 1. The lower set of curves shows the actual transconductance versus plate current.
Note that the scales are log. But at 15mA I would have a plate voltage of only 33V to stay in my 0. Sylvania specifies 15mA at 90V, resulting in 1. I remember burning my fingers all too often when replacing 6DJ8s in color TV tuners.
Sometimes application engineers get carried away. In this case, they dealt with the "devil of reliability" for removing that last bit of snow in the picture. Now we must create the application note for an existing tube in an entirely new area. For me, this is where the fun starts. Remember that the cascode circuit for which the 6DJ8 was designed works with microvolt inputs. We employ volt inputs in line stages and drivers and so use higher supply voltage and with lower current.
The 6DJ8 has plenty of transconductance at 5mA and below, thus making it ideal for this new application. I also carefully measured several sample brands, including the Chinese-made Gold Dragon, that were unavailable in Tungs" was a poorly made 6DJ8 or something else. Reflecting on the days of my youth as a TV repairman, I recalled variable-mu tubes used for RF amplifiers in color TV tuners to control the gain of the first amplifier stage and accommodate a wide range of signal strengths received from distant and local stations.
The input signal is so small that nonlinearity is not as important as preventing overload of the following stages. In the manual I found that the 6ES8 had many of the same characteristics as the 6DJ8, and immediately went to my tube museum to find one.
The visible structure on the marked 6ES8 was identical to the suspect 6DJ8. After further research I learned that is a similar tube with a grid wound in a special way to create the variabl-mu characteristic. Tearing one apart to satisfy my curiosity, I discovered a grid purposely wound of nonuniform pitch. As I expected, the winding was closely spaced in the center and gradually opened up toward the ends.
Both the 6DJ8 and 6ES8 are "frame grid" tubes, one of the last major advances in the tube industry. Instead of winding a hair-thin wire on a pair of thicker side rods as in a 12AX7 and its family, a wire ten times finer is wound onto a two-rung ladder or "frame. Grids are wound as a part and are stored in trays until a skilled worker picks them up with tweezers and places them into the micas.
Without the frame, the fine wire structure would collapse upon insertion. These finer grid structures can then be placed closer to the cathode to achieve the high transconductance that is the basis for the wonderment of these tubes. A variable-mu tube is fine for its purpose, but neither I nor anyone else was making color TVs with tubes at that time.
No wonder these appeared; someone tried to turn a large stock of unwanted tubes into gold, dumping so many of these onto the American market that I was pulling them out of preamps for years to follow.
Stormy Weather If I am any judge of what is coming in the pages of the audio journals, I see a big storm for Glass Audio and maybe a side effect in Stereophile. He was a bit excited about the cover article in the subsequent issue "Is the 6DJ8 Suitable for Audio? When I read it, I was floored.
Preamp Tubes – 6922 / 6DJ8 / 7308 / PCC88
The Number 1 tube in my shootout! With They do some things I have not heard from any of the other tubes — at least not to their extent. First off they have the greatest detail of any tube in the review, and the highest amount of Clarity while remaining completely Musical, with a warm naturally rich tone. Superb balance of air and transparency with positive Euphonics.
Welcome to TubeDepot
Tungsram E88CC / 6922
6922 / E88CC Tubes