A brief introduction to tubes

What I am presenting here is a quick look into vacuum tubes (valves) and is certainly not the authoritative text on the subject.  This is meant as a quick way to get a basic understanding of tubes in audio before purchasing that first tube amp.  I am simplifying a good bit for the sake of brevity.


Tubes are electrical devices that are designed as regulators.  That is, they control the flow of electrons through a circuit and thus influence voltage, current, or both.      All tubes operate on the principal of thermionic emission.  In plain language, when some metals are heated, electrons will break loose and can be pulled away from the metal.   In order to keep those electrons from being absorbed by random dust particles etc, this is done in a vacuum so electrons have a clear path between end-points (Thus the need for the sealed glass tube).   Once that metal is up to heat, a current is applied to supply electrons, and a positively charged Anode is placed some distance away.   Electrons flow from the heated cathode to the anode.


This is a diode tube (2 nodes) and works as a regulator as it forces electron flow in a single direction.     In audio, we commonly see use of diodes as rectifiers in power supplies.     Diodes are not amplifiers as they have no mechanism for controlling the flow of electrons within the tube so simply act as a one way valve (now you know where the name came from).


In order to amplify a signal, we need to introduce a 3rd electrode, the grid.  This creates a tube called a triode (3 node). The grid is a wire mesh that wraps around the cathode and is negatively charged.   With a large negative charge on the grid, electrons are repelled and effectively are trapped between the cathode and grid, in this state, we have no output.  When the signal we wish to amplify is introduced to the grid, it causes the grid’s charge to vary proportional to the input signal, and thus allows electrons to flow from Cathode to anode in direct proportion to the input signal on the grid.   This can be thought of as behaving a bit like an overhead projector in that you place the original on the grid, and the output signal is produced by the electron flow from Cathode to Anode.  In that regard we are not truly amplifying the original signal, we are re-creating it at a higher amplitude.    This is the simplest form of amplification tube and is very common in low power circuits.


The triode has its limits, chief among them is internal capacitance between the cathode and grid.   As current increases, this becomes a larger issue so usually triode are seen in low power applications that avoid it by simply staying below the threshold.   In order to produce greater output power, several varieties of tubes exist that correct for the triode’s shortcomings in one way or another.


The simplest is the Screen-grid Tetrode (4 nodes) which introduces a screen between the cathode and grid that acts to counter capacitance between the two.  This works by basically prescreening the signal so it passes through both the screen and the grid to reach the Anode.    The advantage of the tetrode is that it can generate a much larger current than a triode and we often see beam-tetrodes used in the power stages of amplifiers as a result.

Another problem can crop up with tetrode tubes, this time at the anode.  When electrons hit the anode, they can jar other electrons loose which can then flow back to the grid (negative resistance) and cause oscillation (This is bad).    Usually, if used below certain thresholds, the tetrode works fine and this is a non-issue, but if pushed too hard, this can present itself.   So engineers came up with a Pentode (5 nodes) that now adds a 3rd grid called the suppressor grid.    So now we have cathode, screen, grid, suppressor, and anode in that order.   The screen and suppressor effectively sandwich the grid and prevent capacitance from the cathode side, and negative resistance from the anode side while allowing the 3 original components to operate in exactly the same way they do in the triode.   Since the pentode design compensates for the two most common failings of the triode, it is often seen in high power applications where those problems were originally discovered.



So if pentodes are better than triodes, why don’t we just use them for everything?   The pragmatic answer is because they cost more to make than a triode so telephone companies (the original consumer) and military installations (their biggest customer) don’t want to spend millions on more expensive tubes if they don’t have to.    For small signal amplification tasks, the triode worked just fine and it saved cost.   For those places where larger outputs were required, Tetrodes and Pentodes were used.


Because of how these tubes were developed (remember we are talking 1910-1930 here), there is another reason we now see triodes and pentodes used in combination.    Because Tetrode and Pentode tubes were designed for much higher power handling, they generally require a large input signal in order to operate correctly.    Early audio was largely turntables with a very low output so in order to power inefficient speakers (remember here 1940s)  that required considerable power, the signal was run from the turntable to a pre-amplifier that worked to boost the signal from the level of the turntable to the minimum level needed for input to the tetrode or Pentode tubes that would then output the signal to the speakers.   It is very common to see triodes used as pre-amplifier tubes to feed tetrode or pentode tubes.


The other type of tube we see is more about space saving than about electrical nodes.  it is the dual or quad.  Dual tubes are quite common and combine two triodes, tetrodes, or pentodes into a single vacuum enclosure.   You can guess why we see a lot of these in stereo application where we have two signals to work with.    Quads are less common, but do exist and are occasionally seen.


That covers the most basic types of tubes you will see in a nutshell.  The Rectifier which keeps current flowing in one direction, the triode, tetrode, and pentode which amplify signal but have differing power handling capabilities.    Now the fun begins,  There are literally millions of tubes out there for various purposes.  All electronics up to the point where solid-state transistors become common place (early 1970s) were made using tubes.   That means, Power supplies, mono and early stereo reproduction equipment, TVs, radios, aircraft electronics, industrial control systems, telephone systems, and all types of military equipment utilized tubes, lots of them.    Tubes were made for about every purpose imaginable and some continued to serve after the advent of solid state equipment for a number of reasons.


So how do we know which is which and what to use them for?

Almost all tubes have either a number or a combination of numbers and letters printed on the tube that identifies it.   The problem is those numbers are not universal.  The most common names will be the American name which is often something like 12au7 or 5963 and the European name which will usually start with ECC (in the case of the 12ua7 the Euro name is ECC82).     In more recent years, a third Tube naming scheme has become more common as with the fall of communism and free(er) trade with the far east, Russian and Chinese tubes have become options.  So now we have at least three and maybe four names for the same tube.  To complicate matters more, often there are tubes that, while technically not exactly the same, are close enough to be swapped in certain applications.    So a 12au7 may not be exactly the same as 5963, but in certain applications, they may be interchangeable.

A word of caution here, to be interchangeable, all specs should be exactly alike or with a very tightly defined margin.  Ebay vendors are terrible for suggesting that tubes can be swapped which should never be, due to differences in plate voltages, heater currents etc.    If in doubt, go with the proper tube or consult a knowledgeable individual who isn’t attempting to sell you something.  It would be a shame to save $5 on a tube only to ruin a $1500 amp.    As an example, I recently saw an exchange between a friend of mine and a seller regarding a 6350 tube.  The Seller had it labeled as a hifi substitute for 12au7.  The 6350 is a computer tube and was not designed for audio, and it actually has the grid and plate pins reversed so is not a drop in replacement.   When called out on it, the seller’s response was that he didn’t say it was drop in or plug n’ play.   While true, he didn’t suggest anywhere in the description that it wasn’t a direct replacement either.   My friend forwarded the exchange to me since he knew I was writing on this topic.   I’ll stop short of calling this seller out, but it is one of the larger US based tube sellers and someone who certainly should know better.  Caveat Emptor! 


It can be difficult to know which tubes are related, and which are exactly alike so the best way to tell is to find their data sheets.  I have attached two RCA data sheets for the 12au7 and related 5963 tubes below.

Understand that data sheets are recommended specs for tubes and the actual voltages and currents used in various devices will vary and quite often are below what is listed on the data sheets as doing so yields longer tube life.   Some more exotic designs may push tubes over the listed specs in order to get enhanced performance at the expense of tube life.

The first key point to check is the base diagram to be certain the pin sequence is the same.  In our case, both are listed at the bottom of the first page of each data sheet.   If the sequence is not the same, at the very least some form of adapter must be used.  A few amplifiers have built in switches that allow for changing the pin order of the socket, but most will require an adapter between the tube and socket in order to be usable.   For our two models shown below, the pinouts are indeed identical.

So we have cleared the first hurdle to using a 5963 in a 12au7 socket.   Next we need to compare heater current to be sure using the 5963 won’t tax the power supply beyond what is expected in the circuit.   The 12au7 lists heater voltage and current as 12.6V at .15A.  The 5963 lists the same values so hear again, the two tubes are well matched.

Now we get into the actual cathode, grid, and anode characteristics and we begin to find differences in the two tubes.   Plate voltages are different although maximums are closer (250V for 5963 and 300V for 12au7), and plate current is different at 8.5 (5963) vs 11.8 ma (12au7).  This is where it can get tricky.  In our case, the 5963 has lower specs than the 12au7 so we need to know more about the circuit it will be used in before deciding if it is an acceptable substitute.     If the design doesn’t push the tube to very close to its upper limits, the 5963 will work.  If on the other hand the amp pushes the 12au7 to near its maximum, it may push the 5963 beyond its comfort zone.  At the very least, this will likely shorten the life of the tube.

Also in amps where a pre-amp tube feeds a power tube, changing the pre-amp tube will often require changing the bias of the power tubes as they will now have either more or less current than expected on the input because of the change in pre-amp tubes.   Some amps have pots in-line that allow for adjusting the bias, others will require that additional resistors be soldered into the circuit.

One other concern exists for the 5963, and that is it was designed as an industrial/computer tube.   This means that the designers basically designed the tube to be a switch where they were concerned with the off position (cutoff) and the on position (saturated) but very little in-between.    Tubes designed for computer use were often not tested for microphonics or noise as they were only concerned with saturation and cutoff which means there are no guaranties that they will be function in an analog amplifier.    Early on, many of the 12au7 type tubes were used in computer applications, but because computer tubes would sit for long periods in the off state, they developed a condition known as cathode interface resistance.   In most tubes the cathode interface is made of nickel and when charged for a long period with no electron movement (cutoff or off position) the interface would have an increasingly higher resistance that eventually put the cathode to “sleep” and kills the tube.   Computer tubes use a specially treated cathode in order to prevent cathode interface resistance but what impact this has between cutoff and saturation was not of primary concern so it could adversely effect its use in non-computer circuits.   In some cases, computer tubes are not suitable for use in audio amplifiers as the microphonics are audible and tubes can hum or howl as a result.    Some computer tubes seem to work fine, others not so this takes a bit of trial and error to find the ones that work in your circuit.

The good news is plenty of variety exists for the more popular tube types without having to substitute in a different type.   Since the 1930s, tubes have been produced by companies like Western Electric, Telefunken,  Mullard,  Svetlana, Phillips, Ryazan, Sylvania, Raytheon, General Electric, Valvo, RFT, Voshkod, Amperex, Positron, National, Siemens, Reflector, Novosibirsk,  JJ-electronics, and Psvane to name a few.    Often tubes from different years or even different months coming off a single production line will sound slightly different so with 50 years of production samples and multiple manufacturers to pick from, one can usually find an option that fits their preferences and budget.

It is worth noting that brands written on the box and even on the tube don’t necessarily tell you who produced it or where.  For example, some RCA branded tubes sold in Europe were actually made by Telefunken, others by Mullard, and still others by Phillips. The big American makers (RCA, Sylvania, GE) usually rebadged European made tubes for sales in Europe and Asia as it cut down on transport costs.      Also tubes made for the military often do not bear a makers logo (by law in England they could not) so many times the military ID and lot # has to be used to determine who made the tube and when.     There is a ton of information on the web about tube history but again, be careful of your source especially if they sell tubes and even more so where the brand demands a premium price.    An average RCA tube that was US made in the 1960s may be worth $10-$15 where a Telefunken made RCA marked tube might command $25-30 since that is the price of average Telefunken made tubes of the same model.   Unscrupulous vendors see a way to double their money and sometimes those US made tubes get new labels to appear as their more expensive cousins.   Internal details are harder to fake and usually a visual inspection by a tube collector or reputable dealer will provide a quick answer as to who actually made the tube.