Blog posts of '2012' 'September'


Early tellies were beastly things with wobbly pictures and if you took the back off to see what you could twiddle  there could be up to 7kV flying about.  Now although this may have been lethal for the twiddler, please spare a thought for the line output valve which had to be a pretty tough cookie indeed.  

As usual, the Mullard boffins came up with the perfect device, capable of working from a 170V HT supply whilst handling a peak anode current of 350mA and robustly standing flyback pulses of up to 7kV all in a miniature envelope they gave us the PL81.  Although primarily designed for television use, PL81 sales were bolstered as Mullard also specified the PL81 as eminently suitable for series pass element duties in voltage regulators.






OK,  for today's blog entry I promise that I am not going to act like a  Chemist wonk (even though I am one!)  but rather instead, a thermionic wonk ( because I am one as well)!!!!!.     Yesterday, we took a look at Tungsten, an essential commodity used in wire for just about every valve ever made.  We left the story at the metallic powdered tungsten stage and picking up the tale again, here we have a nice ex-Mullard photo of a batch of the fine dark grey powdered tungsten leaving one of Mullard Blackburn's reduction furnaces: -

The first stage in the multi stage process of turning this amorphous granular substance into wire is to form it into a solid ingot.  This is done by the process of sintering in which the tungsten powder is loaded into moulds, heated to 100 degrees Centigrade in a hydrogen atmosphere and then is subjected to a 100 ton force to compress form an ingot.  The ingot is then clamped between platinum electrodes and has a current of 2500A @ 24V passed through it which raises it's temperature to 2700 degrees Centigrade. After ambient cooling, the ingot then posesses a crystal lattice structure that makes it ready for wire making.  Below we have a nice ex-Mullard photo of a filled mould being inserted into the ingot press - the ingot so produced would be approximately 35 x 2 x 2 cm in size: -

The next stage of wire making is swaging, where the ingot is reheated to 800 degrees Centigrade and pushed into swaging machine where sequenced pairs of mechanically operated hammers - 11 pairs in total -  pummel the ingot whilst it is being rotated such that  the tungsten ingot now becomes a tungsten rod some 600 x 0.25 cm in size.

The final stages of wire making is where the tungsten rod is then further reduced by drawing the rod through a sequence of circular tungsten carbide dies, each smaller than the preceding one, the rod is lubricated with a colloidal graphite solution which is sprayed onto the rod pre draw and then removed post draw by running the wire through a boiling caustic soda bath before winding the wire onto a spool.  This sequence, lubrication, drawing, cleaning and spooling is repeated until a final wire diameter of 0.5mm diameter is achieved.  Below we have a nice ex Mullard photo Tungsten wire being drawn on a chain bench diamond die stock drawing machine: -

From this point on, the dies used are made of industrial diamonds and the wire diameter is successively reduced until the wire achieves the required diameter and so it is that a 35 cm ingot is transformed into 20 miles worth of 8 micron diameter fine tungsten wire ready for use in winding valve grids - but that's another story for another blog entry.







We've seen so far in my blog series on valve types and manufacturing techniques that fine wire is an important material used in the valve making process. To recap, tungsten wire is used as filament wire with molybdenum being used as grid and filament supports. Just to give you an idea of HOW essential this material was, in the Mullard Blackburn plant heyday of the early 1950s over 2500 MILES, yes, MILES of wire was used every WEEK in valve manufacture.

So, where does the tungsten come from?, how is it formed into wire?, well paisan, read on and all your questions will be answered..............

Tungsten is derived from it's ore, in particular, the form known as Scheelite which is calcium tungstate having the chemical formula CaWO4.   This type of ore was originally found in Sweden in 1820 where it was first named Tungstan - which translated as 'heavy metal', however, in 1821 it was renamed Scheelite in honour of it's discoverer Carl Scheele.

As an ore, Scheelite is fairly unimpressive apart from being heavy, but it does fluoresce a bright blue when exposed to UV light at 254nm.  Most of the ore used in valve manufacture was mined in Australia, with indigenous deposits coming from mines in Cumbria and Cornwall.    The Scheelite arrived at Mullard's Blackburn plant as small pebbles packed in coal sacks. A giant ball mill, originally designed for the limestone quarrying industry was used to reduce the ore to a fine graded material having a particle size distribution of between 0.1 to 0.15mm diameter - a process that took 6 days!

The graded ore was then reacted first with hydrochloric acid and then heated to 400 degrees Centigrade to form the bright yellow Tungsten (VI) Oxide which occurs in accordance with the following formulae: -

CaWO4 + 2 HCl → CaCl2 + H2WO4

H2WO4 + heat → c + WO3

Tungsten (VI) Oxide is in turn reduced to metallic Tungsten in powder form by reduction in a stream of Hydrogen at 650 degrees Centigrade as per the following formula: -

WO3 + H2 → 3H2O + W

Although the hydrogen was both expensive and dangerous it had to be used as a reducing agent as the cheaper and safer alternative Carbon, would react to produce Tungsten Carbide - good for sandpaper but no good for valves.

So there we have it, pure tungsten, if you are still awake, please 'tune in' tomorrow for the next scintillating episode in this story and I'll explain how the grey powdered Tungsten became wire at Mullards!

PS: I did tell you all that I used to be a Chemist before I turned into a valve salesman - didn't I???? 





The ECL80 valve was designed and introduced by Mullard during the early 1950s to facilitate the reduction in valve count in televisions by placing two valves in one envelope that could be closely associated in such a set.  

For example, wouldn't it be nice to have a frame blocking oscillator and frame timebase in one envelope - voila, the ECL80 comprises such a system with the pentode cage designed as a frame timebase and the triode as a frame or line blocking oscillator .. or.. as an AF amplifier....or.... as a grid detector. 

Another great application is for the ECL80 to act as a high sensitivity audio amplifier with the triode cage acting as voltage amplifier with the pentode cage acting as output valve and indeed this valve found such use in contemporary single ended record player applications as well as TV usage.    Lateral progression presents other convenient uses, how about the triode cage as a grid detector and pentode cage as an audio output valve.

Although a great valve, it wasn't without it's faults, differential heater rates could cause conflicting problems in earlier televisions and the electrode system was prone to microphonics which could lead to vertical hold judders.





Well, the great Mullard Magic roadshow has come to an end.  We started with an appearance at the Head Fi 2012 show where we had a great time, courtesy of our host Andy Smith where we had some top titters discussing valve related stuff with Richard, Chris & Lucy and Julian amongst many others.

Here are a few pics to whet your appetite for the 2013 meet: -

Amazingly, we continued our journey, taking in Norfolk, Suffolk and the Isle of Bute and saw many wondeful things - here's just a couple more photos of some of the more interesting and bizzarre things we saw: - 








 A filament current drain of 1A was a real problem and the search was on to achieve good electron emission from a reduced filament current.  Early efforts were based on the use of alkaline earth metalled filaments where tungsten was doped with thorium, barium or strontium oxides before drawing into filament wire.  Development progressed  until a final configuration of tungsten wire, overcoated with a thorium oxide layer was settled upon.     This greatly reduced filament current consumption and had the side effect of a valve loosing it's bright but homely glow.   As a consequence, these new low current valves were termed DULL EMITTERS and often, retailled valves would carry a paper label adhering to the envelope stating that the valve wouldn't glow - just like the example in our photo below: -

The greatly reduced filament consumption of the DULL EMITTER valve allowed the use of a comparatively small two volt accumulator , hitherto valves used a whopping great accumulator of high AH capacity to power their four volt filaments.

And so it was until the late 1940s where the standard voltage for battery valves was 2V.    Interestingly by the 1950s, Mullard introduced a new series of miniature battery valves - the D series (there will be a  future blog entry on these).  The D series battery valves were rated with a 1.4V filament consuming 50mA , progress indeed.



Back in the dim & distant past, in a land far away, shrounded in the mists of time there was a valve................... and what a valve, not like the ones we have today, oh no, it was very different indeed.    

At the time, it was very similar to all its contemporaries - not that there were so many -  and like them,  it was a general purpose triode having a four volt pure tungsten filament which operated at a temperature similar to that of an incandescent light bulb.  The electrode cage was mounted an an inverted pedestal sealed into the bottom of the envelope with a length of glass tube being welded onto the pedestal - just the same way in which an incandescent bulb was manufactured.   It had a bulbous envelope which had at it's apogee a pip of glass. - yes, it was a BRIGHT EMITTER VALVE.

The bright emitter valve was hampered by two disadvantages.  Firstly, the intrinsic characteristics were not best suited to each stage of a radio receiver and compromises in design and performance were the key to garner adequate performance. Secondly, the filament had a whopping current consumption at 1A.

There had to be a better way if this new fangled thermionic electrickery was to catch on.................



In a previous blog, I described the introduction of a the Mullard World Series range of valves, now I should like to describe each one of them briefly.  Let's start with the EF80 which is an RF pentode.  This versatile valve can be used as an RF amplifier, a mixer, an IF amplifier and additionally a video output valve.

It was designed specifically for use in the 40 - 70 MHz frequency range to suit contemporary television receivers along with the ability to operate satisfactorily from a 170V HT supply which was the operating voltage of choice for contemporary transformerless TV receiver designs of the time.

It's peformance when presented with a 100uV signal was impressive - the minimum signal required for a watchable monochrome picture needed a gain of approx 100dB is required to achive satisfactory viewing - using a bandwidth of 3 - 6MHz centered at 65MHz in a transformer coupled circuit stage, a gain of approx 130dB was achievable, so a good picture with plenty of headroom was possible.  Well done to the EF80 I say!





Hah! what a way to start September with the last in my series on components that make up a valve, yes, we've saved the best till last with episode 14 dealing with..... the GETTER.   Having assembled the electrode cage and mounted it into the glass envelope, the last stage in valve manufacture is to remove the air within to form a high Torr vacuum immediately before the sealing-off process.

In order to achieve this vacuum, the valve is connected to a vacuum pump and whilst being evacuated is passd through a chamber where radiant heat and an RF eddy current is applied. This double whammy of heating - radiant & induced -  is an attempt to ensure that any occluded or interstitial gas adhering to the glass or trapped within the metal componentry is driven off.   Inevitably, some gas molecules will remain, in fact, the mean remaining gas level has been quantified as one part per one-hundred, thousand, million, so as you can appreciate, we are talking about a miniscule amount of residual gas! 

In order to remove this residue, a volatile transition metal is submlimated onto the glass envelope, typically barium but sometimes magnesium which can be seen as a bloomed silver mirror on part of the interior of the glass envelope. - this is the GETTER.    The GETTER is a sacrificial material which 'mops up' the slowly released residual oxygen.  Many of you reading this I am sure, will have seen high power beam tetrodes - KT88 are particularly prone -  on their last legs with the getter apearing a pinkish-brown shadow of it's former self as the shiny metallic bloom is consumed by the process of oxidation. 

The pre-charging and application of the GETTER is an elegant and beautifully engineered process which is not often discussed, but we will now rectify that!  The picture below shows a typical bar getter ring as used by Mullard and other valve manufacturers and these found use either as a pillar mounted trans mica assembly and as a pinch supported assembly. 

Many people mistakenly describe the getter ring as the GETTER, many people think that the entire assembly gives up parent metal to become the bloomed GETTER silver mirror, many people believe the getter ring to imbue valves with a particularly favourable sonic property. OK, let's take each of these points in turn: -

The only part of the getter ring that gives up parent metal is the part labelled B in the inset photo above.    The getter bar, B, comprises of an hollow iron sheath, C, which is filled with  a core, E, of 0.5 to 25mg of metallic Barium.   The content of the core is varied dependant upon the size of the glass envelope used.  The sheath, C,  has a thinner planar side, D, which allows sublimation of the core when an RF eddy pulse is applied.  

The nickel induction loop, A, attenuates the RF eddy current applied during evacuation to allow effective sublimation of the core.  The shape of the induction loop varies and is governed by the geometry of the getter ring mounting provision and the size is governed by using the least amount of metal to effectively attenuate the RF eddy curent pulse to ensure adequate heating to provide effective sublimation such that a complete GETTER silver mirror is formed.

The shape of the getter ring should have no effect on the sound quality of the valve, however, some valves may in the course of continuous development have used strengthened or lengthened support rods  in conjunction with different getter ring shapes which could effect overall cage rigidity which could have a consequent effect on micophonics and perceived sound quality.  These subtle changes would then be attributed to getter shape changes giving rise to today's urban myths.