Chapter 3:
ON SPECIFICATIONS AND DATA SHEETS
- 1: Data Book Specifications
No manufacturer ever put out bogey values that are comprehensible. Or practical. Or in any fashion remotely useful. Specification sheets were clearly designed to give you no useful information. They are exercises in mystification, obscuration and deceit. Generally the only useful information is the pin out diagram. But even there they cannot resist using all manner of differing abbreviations to confuse the reader further. If they could lie about that they would. But surely the reader will say that the heater supply requirement is a given: its either 6.3V or 12.6 volts. WRONG! Principally tubes were designed to run off a car battery that was 6.3V and later 12.6V. But here comes the rub: a cell puts out 2.1V when drawn down which equals 12.6V if you stack 6 cells together, but in order to keep the battery in a constant state of charge it requires 13.2V from the generator. In effect all tubes were designed to handle a 12.6V A supply, while running at 13.2V. Some of the older tube sheets clearly indicate a minimum and maximum heater voltage. Again, we fail to realize even the most basic bogey value.
To summarize: no spec sheet conveys anything useful. To claim that a tube must be compared to the parametric specification is false. What some honest manufacturers do is say “in this circuit this tube should behave in this manner, with final values in this range depending on circuit configuration.…”.
The reason for this is simple. If manufacturers were to provide specifications then customers would be able to return tubes “because they are bad”. Worse still people would be able to say “my circuit works within the parameters established in your specification sheet, my circuit does not work, so obviously the tube batch must be bad”.
This is not only the case with older tubes, its also the case with brand new tubes. Nobody wants to be held responsible for objective standards. Even if consumers could be educated to understand that there is no such thing as a bad tube, I doubt manufacturers would provide any object way to test the tubes, even parametrically in direct current. The best we can hope with new fresh from the factory tubes is that they are consistently tested to the same specs so that we can take a trade mark as a given set of variables.
The reader also needs to understand that “tube resellers” (mostly mail order nowadays) do their best to keep their clients happy. Most are perfectly honest business people. Most will not knowingly resell tubes they know are off bogey, or shorted or microphonic. Most tube resellers do their best to only sell what they know will not come back. Nobody is giving them a warranty. And a return is a costly proposition, both in goods and in freight. If they make 3$ on a 10$ tube and it costs them 6.90$ in shipping and handling to send you a new one they will loose money (ignoring the time it takes to do so). Most online resellers live off their brand equity, the knowledge that happy customers return to buy more and recommend them to others.
The problem with the market for tubes is that manufacturers ship tubes without any markings on them (semiconductor manufacturers do the same). They ship the “on spec tubes” to companies that private brand them and offer some sort of guarantee of quality, and ship the rest further categorized from good to bad. The parts that fail probably get crushed and recycled, although its hard to get a totally dead tube. Its when “B Grade” parts hit the market, that problems develop. If you have just paid $2.00 for a tube that retails for 12$ with XYZ trade mark on them, will you counterfeit the mark? Another problem with the tube supply chain is that some brands will specify materials and others not. Then you have two physically identical parts with completely different emission profiles. Some materials conduct better than others, some emit better than others and some filaments last allot longer than others. But you would never know looking at the tube. The last problem is one of channel control, namely that the managers in charge of “brands” are control freaks. I remember that before the internet a very heavy amount of sucking up was required just to get the annual parametric databook. Getting a manufacturer to sell to you is hard. This ties in with the fact that the same reps are also lazy. They would rather sell 1000 units at 6.00$ then sell ten lots of 100 at 9.00$. What that means is that one dealer will buy the 1000 in order to get the price and then sell the 800 he does not need in lots of 100 at 9.00$. At that point all channel control goes out the window, and counterfeits magically appear through “the usual channels”. One must remember that it is not a crime to sell a counterfeited electronic device- its simply “reversed engineered”- which is legal. The only illegal part is counterfeiting the trade mark. Actually, the worse crime is reprinting the spec sheet- that is a copyright violation. I’m serious on this last point, US Customs has no way to know if something is original or not- they can only stop goods on trade mark and copyright violations. And the only way to know if the trade mark is genuine is if the trade mark owner sold you the part directly.
If you go to, for example the Electro Harmonix website (http://www.electro-harmonix.com) you will not find any spec sheets. You can find them “floating around the internet”, but there is no specific sheet that can be correlated to the tube you just bought. It might just be an older design or built at a different factory. That is not to knock their products, I like their tubes and have them installed in some of my personal equipment.
On the other hand if you do go to the Tung-Sol website (http://www.tungsol.com/tungsol/) and download the spec sheet on the 12AX7, you will see at the top of the sheet, the heater spec and the pin out. In the middle you will see some capacitance and resistor values. What interests us is what is at the bottom under the heading:
TYPICAL OPERATING CONDITIONS AND CHARACTERISTICS, Class A1 Amplifer, each Triode Unit.
Heater Voltage 12.6 6.3 12.6 6.3 Volts
Heater Current 150 300 150 300 mA.
Plate Voltage 100 250 Volts
Grid Voltage -1 -2 Volts
Plate Current 0.5 1.2 mA.
Plate resistance 80 000 62 500 Ohms
Transconductance 1 250 1 250 ,umhos
Amplification Factor 100 100
Ultimately that is all that we need for our purposes: The cathode voltage (Vk) which is 100-250V, the grid voltage (Vg1) range -1 to -2V and the expected plate current (Ia) of between 0.5 and 1.2mA.
The best part of the above spec sheet is that it is giving us a full range on the three variables we need. The most important one is of course the plate current value: don’t exceed the higher of the two and the tube cannot be damaged. The limiting value of the current at the anode (Ia) is the most important value- every thing else you can deduct.
So this the skinny on tube data sheets:
1) Some manufacturers will not even give you a data sheet to work from (the tubes being sourced from multiple factories).
2) Some manufacturers will give only that which you need (product sourced from captive factory).
3) Some manufacturers will frustrate you by using complex circuit centric mathematics to try to make you feel stupid and abandon the endeavor.
4) Some will give you everything you need to test and nothing more (www.jj-electronics.sk)
Once you have familiarized your self with data sheets pull some old Telefunken spec sheets from the internet (say an ECC81, ECC82, ECC83 and EL34).
They are incredibly well written and comprehensive.
In the next section we will try to explore all the different abbreviations and symbols that have been used and are currently in use.
I bring these old examples up not because one should tube-roll older tubes but because some manufacturers still use the same terminology. A case in point is JJ that uses German, or better, the Telefunken system of abbreviations. Similarly the Tung Sol example above is straight Americana.
- 2: Abbreviations and Symbols
First and foremost what we use.
Let us start by at least defining some of the more common terms encountered.
The Heaters if separate from the Cathode, are heaters, namely H. When cathode and heater were one, they were referred to as Filament, “F”.
Thus the “F” designation stuck. Its also technically a filament, so it’s a valid name for it. For the sake of convenience we will call them “H”.
Most all heaters don’t care which side is positive or negative, so they remain “H” not H+ or H-.
In the case of a double triode like the 12AX7 it can be run “off two 6.3V secondaries” in which case the common is “Fc”.
Since the A Supply is capable of 12.6V run it at 12.6V.
The cathode is usually referred to as a “cathode” which is what it is. Abbreviations might vary but its usually indicated as “K”
The cathode is the emitting element, and its connected to the B Supply at the Negative “-“ pole.
The Anode is usually abbreviated as “A”. It is sometimes referred to as the Plate “P” which its not- it’s an anode.
The anode is the receiving element, and its connected to the B Supply at the Positive “+” pole.
The grid is known by many names, some functional.
As an emitting device, it follows that it is ”cathodic”, and on the C Supply would tie in to the negative polarity.
Grid 1 is usually known in the USA as the “control grid”, which is what it does so its valid.
Grid 2 is usually known in the USA as the “Screen grid”. The reason for that is that cathodic current will repel emission, while anodic current will attract.
Grid 2 can be run positively in relation to the anode itself, thus attracting or accelerating emission.
On the A and B Supply the Positive lead Floats to Ground.
Thus on an A-B-C circuit the cathode is grounded to the positive lead on the C supply in order to complete the circuit.
The Control Grid is usually abbreviated as “G”. The number alongside it refers to which one of 3 possible or more grids are available: G1, G2, G3.
Grid 2 is usually tied to the Anode and G3 to the cathode, and they derive their purpose from them not from the power supplies.
Abbreviations:
V = voltage, expressed in volts “V”, sometimes directly described as “DC” = Direct Current and “AC” alternating current
I = current, expressed in amps “A”, or its register shift “mA”. Thus 500mA would also be 0.5A.
USA uses commas for legibility, the rest of the world uses commas for register shifts: USA 0.5A = Europe 0,5A. So 1 Dollar in US would be $1.00, one Euro E1,00.
The denominating unit takes precedence over the tube element, therefore it precedes it: as in Vh: voltage at the heater, Vk: voltage at the cathode, Ia: current at the anode, Vg1: voltage at the grid.
- 3: Non Circuit Conventions used:
Ih = Current at the heater expressed in amperes
Vh = Voltage at the heater expressed in volts
Ik = Current at the cathode expressed in amperes
Vk = Voltage at the cathode expressed in volts
Vg1 = Voltage at the grid expressed in volts
Ig1 = Current at the grid expressed in amperes
TUT = Tube Under Test in a parametric voltage driven system.
The Cathode (Vk) is always Negative
The Anode (Va) is always Positive
The Grid1 (Vg1) is always Negative
The Heaters (Vh) are polarity neutral.
G2 is tied to the Anode
G3 is often tied to the Cathode, sometimes internally some times externally.
A Voltage indication is an actionable command, which means the indicated Voltage requires you to set it to such value.
A Current indication is a verifiable observation, which means that the current indicated on the meter is a result of Voltages applied.
A Current limit is available and will inhibit the ability of the emitting tube element to reach set voltage value.
A voltage setting on a particular supply (A,B or C) requires a polarity indicator, positive or negative (+ or -).
A voltage setting on a particular tube (Vh, Vk, Vg) does NOT require a polarity indicator since it is assumed that the user understands that current moves in one direction only, namely from the cathode to the anode and from the cathode and grid to ground.
On Measurements:
A Supply:
The heaters are filaments. What you supply them is what you supply them. In our Constant Voltage model, you set the voltages and whatever current it draws it draws. Thus both measurements are upstream of the tube under test. Measurements are internal to the circuit producing the direct current.
B Supply:
You take the Voltage measurement before it enters “as supplied” to the cathode and you take the current measurement after it exits the anode. Thus the voltage measurement is upstream of the cathode, and the current measurement is downstream of the anode. Both measurements are in the test circuit.
C Supply
In our Constant Voltage model, you set the voltages and whatever current it draws it draws. Thus both measurements are upstream of the tube under test. Measurements are internal to the circuit producing the direct current.
- 4: Country Specific Databook Terminology
Every country has its own methodology to describe electricity. Its part of the culture war on civilization. Some attempts have been made to standardize it internationally. First there were pan-european standards for abbreviations, then international ones. The US Military (JAN) used its own set of abbreviations. The issue is this: the standards are derived from CIRCUIT DESIGN. And we are interested in TUBE DESIGN. The difference is substantial. The design criteria for circuits have changed radically in the last century. Its an ongoing evolutionary process. When a tube manufacturer uses a set of abbreviations its because it assumes its tubes will be used in a certain circuit topology as in “Pentode Connection”, or “Class A”, or “Class AB”, or “Pull Push” or “Fixed Bias”, or “Cathode Bias” or better yet as “Two Valves in Push Pull”. We could go on for ever… Suffice it to say that allot of the information is superfluous for our needs. Class A reference circuits are the closest to real life parametric variables.
Nonetheless if we wanted to design a new family of thermionic tubes we would have to use different abbreviations to describe them and possibly provide a “reference circuit design” to be able to use them. To test them we only need the basics. Class A reference circuits are the closest to real life parametric variables.
Since abbreviations are country specific lets look at some examples.
- 5: British Standardization:
Marconi, Mullard, Genalex:
Va = voltage at the anode
Vk = voltage at the cathode
Vg1 = voltage at the grid 1
Ia = expected current at the anode
Vh = voltage at the heater
Ih = current at the heater
A sample Marconi Specification Sheet.
Notice how in 1955 they made the switch from B329 to the more international 12AU7.
Notice how there is nothing to discuss about their terminology- it is what it is and what be be.
- 6: Dutch Standardization:
Phillips: (Find limiting Values for non circuit specific)
Va = voltage at the anode
Vb = voltage at the cathode
Ik = current at cathode
Ia = current at anode
Vg = voltage at the grid 1
Ia = expected current at the anode
Vh = voltage at the heater
Ih = current at the heater
- 7: German Standardization:
Covers most Telefunken, RTR, and JJ Electronics
Uf = Voltage at the heater
If = Current at the heater
Ua = Voltage as the anode
Ia = expected Current draw at the anode
Ug1 = Voltage at the grid (s) 1,2,3
Ra would be resistance at the anode and Ca would be capacitance at the anode et c. all pretty logical.
In the Latin Alfabet, a “W” is a double-U. And U’s were usually printed as V.
Lets be generous and leave it at that.
- 8: USA Standardizations:
Generally USA manufacturers do not use abbreviations, they spell the symbol out completely in consumer facing communications. But its not clear ever if it’s a consumer term, or a term given to a service technician or whether the spec sheet should be useful to a circuit designer. Usually, if the document is stand alone it probably for a designer, if its part of a collection of tube families, its directed to service technicians (sell tubes, don’t think).
With RCA its all pretty nebulous. A Telefunken spec sheet is not nebulous, an RCA spec sheet is.
You can buy online a complete RCA Receiving Tube Manual for a given year. I have on my desk “RC-30”, its useful for pin out and heater voltage but little else.
I will take a stab at guessing: their treatment of abbreviations is circuit centric. But its also never specified if the circuit spec is AC or DC. They interchange the terminology which means that you should know a-priori if they are referencing a circuit, a vacuum tube, in AC or DC. In any event US abbreviations assume that current moves forward from the positive terminal. Similarly, Ep or electricity at the plate is only useful when using a tube to convert AC current into DC current. Also the current drawn by a grid is so small that it can barely be measured. So if the RCA spec says “screen or control current” its hard to understand what that could be useful for. What they are saying is “we will use the colloquial term current when what we mean is voltage”. If on the other hand they do mean current as in “I” then I would have to assume that it’s a throw back to when the grid was controlled by a battery, and current draw at the grid was an important consideration.
What is mind boggling is that RCA would be so sloppy with their numbers, and yet, the first one hundred pages of the Receiving Tube Manual (RC-30) is dedicated to essays on every aspect of the tubes, and the sections are exceedingly well written in plain English. I guess they do not want to be held to any objective criteria. In any event the best the RCA spec sheet can provide is nothing: look up a 12AX7 tube in RCA RC-30, there is absolutely nothing there, never mind any variable or fixed measurement of any use. There are allot of RCA tubes still floating around, and empirically they sound very good and are my personal favorites. They “test” very badly though, but within the bounds of an audio application it does not matter. I own hundreds of RCA receiving tubes. I will outlive my inventory of RCA tubes. Then they can bury me with them. I do not foresee anybody using RCA tubes in a high frequency high power transmitter; nobody should be that unkind to an RCA tube.
RCA:
Heater Voltage (ac/dc) 6.3 volts = heater voltage
Heater Current 0.9 ampere = heater current
Plate Voltage = cathode to anode voltage
Ep = Electricity at the plate = Anode Voltage
Ip = Current at the plate = Expected Current at the Anode
IG2 = Screen Current = Current at the grid2
IG = Control Current or Bias Current = Grid 1
G1 vdc = Grid 1 control Voltage
Tung-Sol
Plate = Anode
They will specify current type as in:
“Plate Voltage, DC 600 VOLTS”
“Grid 1 Voltage, DC -30 VOLTS”
Svetlana (English spec sheets)
Generally as indicated with Plate referring to the Anode.
Cathode is cathode and grids as G1, G2, G3
- 9: JAN Tubes
Joint Army Navy: USA
Ep = PLATE VOLTAGE Voltage DC Value
Ebb = PLATE SUPPLY VOLTAGE DC
Ec = GRID BIAS VOLTAGE, DC VALUE
Ecc = GRID BIAS SUPPLY VOLTAGE
eb = INSTANTANEOUS PLATE VOLTAGE
ec = INSTANTANEOUS GRID VOLTAGE
eg = AC COMPONENT OF GRID VOLTAGE
ep = AC COMPONENT OF PLATE VOLTAGE (ANODE)
Ip = DC PLATE CURRENT
Rp = DC PLATE RESISTANCE
Rg = GRID RESISTANCE
Rk = CATHODE RESISTANCE
Ri = LOAD RESISTANCE
The above is obviously circuit centric. Their understanding of vacuum tubes is entirely application/equipment driven.
- 10 Others:
France, Italy, Japan and the Soviet Union all produced excellent tubes.
Only Phillips, Telefunken and the Brits produced tubes for export with USA standardized sockets. Regardless, you should be able to view a non English spec sheet and make out cathode, anode, filament etc.
- 11: WATTS??
This tester does not express anything in watts.
If you know what they are use a calculator to make the conversions. We will make the (erroneous) assumption that if you are expressing anything in watts its either because you are operating in alternating current or discussing a circuit topology. Also, watts are mostly used when the voltage is a given, as in a household mains. Watts are units at a fixed voltage- and we are using the voltage(s) as the variable. All you need to know is that Power is Amps times Volts which means on your double triode that if you double the volts you halve the amps- 150mA at 12.6V or 300mA at 6.3V (with 2 secondaries).
In any event: 1.5A x 6.3V = 9.45 Watts, but then again even if the heater was employed as a light bulb that watt number would not give you a lumen value.
Watts are like horsepower in a combustion engine- its useless information unless the torque matches the weight it needs to move with the appropriate ratio.
Phillips gave certain specifications in watts, and they are certainly entitled to, if you know the voltage you can deduct the amperage.
