Mullard valve answers

[Alan Stepney]

Yesterday I visited a guy who spent his working life at Mullard’s, Mitcham.
He started as an apprentice and ended up as a departmental manager, but for most of his time there was involved in testing and quality control of valves.


He is a mine of information, and a fascinating person to meet.
Unfortunately, time ran out before we were able to cover a fraction of the topics I would have liked to discuss.


To start, some here may recall that I had a large batch of new Mullard valves that had cracks in the bases. We discussed this here about a year ago.

To start with, he confirmed that there were genuine Mullard valves, made at Blackburn.
This particular fault was a known one and easy to diagnose, especially if you have spent your life doing so!

At one point, Philips Eindhoven produced a large quantity of bases that were faulty.
Both the gas jet position and temperature reached in making the base were incorrect. It is not known how the Dutch quality control didn’t detect the faults, but they didn’t and hence many thousands were sent to each of the Philips/Mullard valve factories, including Mitcham and Blackburn.

The fault was soon detected and everyone was instructed to destroy the bases concerned. Mitcham certainly did so.

It is surmised that someone at Blackburn, instead of destroying them, put some or all of the faulty bases in store, and later, perhaps towards the end of valve manufacture when parts were slow in coming through, someone put them back on the production line.
It is likely that these were sold in large batches, and those I have could have been a bulk buy, probably at a discount, right at the end of the “valve era”.

So, a manufacturing fault, and nothing to do with storage conditions nor anything else.

[Alan Stepney]

PRINT COLOUR.
Yellow print valves are higher quality, or specially selected, or for a special purpose, or for a special customer. It must be true as I read it on the Internet. Lol

The truth is far more simple.

Mullard was founded in 1920.
1970 was, therefore their golden anniversary and to commemorate the occasion, a new rose was bred and planted around the Mullard factory/offices. (Capt Mullard was a rose enthusiast.)

Also, all valves were to have the print in gold paint. However for technical and financial reasons, yellow was substituted.
The yellow was planned to be on all valves made during the anniversary year, with some production line starting to use it at the end of 1969. During the year, and especially towards the end of it, as lines ran out of paint, they would refill with whatever colour was to hand, and so you could have adjacent lines turning out valves marked in white and yellow.

So, just a marketing device and with no technical significance whatsoever.

MULLARD LOGO
The change from the word MULLARD to the stylised M on valves was part of a plan by Philips to drop all brand names and re-brand everything under the Philips name.
The M was a first stage and would presumably have been followed by dual Philips/Mullard packaging, with the “Mullard” then being dropped.
As we all know, changed within the industry meant that Mullard closed although the Mullard Research Laboratories was renamed Philips Research Laboratories.



(More later).

[thermionic]

Alan.

Thanks for sharing this with us. Information like this is priceless. Perhaps you could persuade you friend to write it all down, Or perhaps, talk at length and at his convenience into a tape recorder!

Information like this must not be lost.


SimonT.

[Peter.N.]

Thanks Alan, very interesting. We used hundreds or may even be thousands Mullard of valves in the '50s and '60s but as the TV trade really took off, as you will know valve manufacturers, or at least suppliers seeemd to be springing up everywhere and some strange names appeared. We had an old boy that drove down from north london to where we were in Kent, every week with a Jowett Bradford van stuffed full of cheap valves, I don't know what he was buying them for but obviously enough to make the trip worthwhile.

Peter

[Alan Stepney]

I was amazed at the knowledge that this guy has.
(The "disadvantage" is that the time went before I had a chance to ask all the questions mentioned in this post: .https://www.vintage-radio.net/forum/showthread.php?t=57716.)

I suggested some notes re his life and work at Mullard, either for some articles, or a book (he has more than enough to fill a book) or that I did a "question & answer" session, recorded and perhaps written up for the BVWS magazine, or for a website.
I even mooted the idea of him addressing a meeting of enthusiasts, either at the BVWs or one of the Ham clubs.
He declined the latter idea outright, and rejected most of the other suggestions.

However, I do hope that he may consider the "Q & A " suggestion.

It would be a great pity if his knowledge should be lost. Even worse, he has massive collection of paperwork, documents etc, from and relating to Mullard, ( a double-garage full) and I dread to think what will happens if he ever wishes to dispose of it, or he dies.

[G8HQP Dave]

Thanks for this, very helpful.

Regarding the yellow print, I noticed just in the last hour a well-known valve dealer offering yellow print ECC83 as specially selected for audio! One wonders what other ECC83 were intended for?

[Alan Stepney]

Wire ended valves, storage problems.

Wire ended valves were used by many users, but demand gradually fell to the point that it was uneconomic.
One user was the GPO (CV466 for in-line repeaters), who wanted supplies to continue for some considerable time.
It was decided to make sufficient to meet their requirements for the future.
Thus extended lifetime and extended storage tests were undertaken to ensure that they would still be “fit for purpose” when taken out of store some years in the future.

Early examples of these valves were taken from the Mullard store[*] and then it was found that some early samples had the leads broken just where the lead entered the glass.
Tests were undertaken to discover the reason.


In manufacture, the valves were manufactured and then the leads dipped in flux before going into a tinning bath. They were then cleaned to remove any residual flux.
It was found that, in many cases, the cleaning was less than effective and some flux remained on the leads, just where they enter the glass seals.
The flux is slightly corrosive.
The effect is that after a long storage period, the leads can and will snap off just where they enter the glass.

A more effective and prolonged cleaning regime was introduced to ensure they had the expected life.

He said:
“It is worth checking any wire ended valves to see if there is corrosion near the seals.
If there is, a good wash, preferably in an ultrasonic cleaner should leave the leads clean and prevent further corrosion. The only problem is that it is likely to remove the valve markings, so will need a label to identify them in the future”.
[*] I understand that samples of every valve, including development samples, were kept so that production could be compared with originals, or others at all stages of the valves production life.

[TuningIndicator]

Many thanks for this.
Mods, could this thread be pinned please?

[Silicon]

I also had the chance to talk for a few minutes with an ex Mullard (Mitcham) employee.

One of the tests involved taking a sample of the valves off the production line at the start of a run and immerse them in a tub of boiling water for several minutes.
They would then take them out and drop them in ice cold water.
After that they were powered up on the test bed to measure their parameters.
They were all expected to pass.
I can't remember if these were military valves or domestic.

I hope that this kind of industrial history is not lost.

Brendan

[AlanBeckett]

Could I recomend that anyone who worked for the valve manufacturers should contact the valve museum and pass on their knowledge.
http://www.r-type.org/
Alan

[Alan Stepney]

Brendan,

This was also mentioned on Wednesday.
It applied to the first sample off each production line at the start of each day.

The hot water had to be above 85C and below boiling point, and the valve had to stay in it for 10 minutes.
It was then plunged into the cold water, which was as close to freezing as possible.

The thermal shock was designed to reveal incorrect temperatures for the sealing and for the glass tube-to-base joint.
(The bases should already have been tested for glas-to-metal seal.)

If any one failed, the entire batch had to be scrapped and aother sample tested.


As an aside, the standard line produced a "carousel" of 36 valves, at roughly the rate of one every 4 seconds. This did vary with the complexity of the valve and also for other reasons, but it is a good general guide to production speed.

(More later).

[kalee20]

Great thread, Alan!

[Alan Stepney]

EF91 GLASS COLOUR.

Some users required extremely extended operating lives. The GPO being one example, and for some purposes, they expected valves to operate for 10 or 20 years.
This was partly achieved by under-running them, but also, considerable attention was paid to the causes of failure over prolonged operating times.

Looking at the EF91 (and similar) the anode only runs round part of the cathode/grid structure, and so a considerable number of electrons pass through the grids and through the gaps in the anode, and into the glass envelope.

It was found that this gradually caused gas to be released from the glass.
Nothing that had any effect under normal use, but over 10’s of thousands of hours, it was sufficient to impair the efficiency of the valve and to shorten its life.

Therefore, the centre part of the glass envelope was coated (by vacuum deposition) with a chemical. <I was told what the chemical was, but forgot to write it down.>

This kept the electrons from the glass and thus reduced the gas released, and thus lengthened the working life.

The actual colour of the coating varied with the temperature of the firing, and has no significance.

Under test it was found that Mullard valves with the coating had a longer life than other makers uncoated valves of the same type. However, the coating added a stage to the production process and therefore cost more.

[dominicbeesley]

Keep this up, very informative!

Dom

[kalee20]

Fascinating - I do love the blue colour of EF91 envelopes.

(As an aside, it's surprising that this has never been a feature of an audiophile fad. No doubt if it had been EF86 or ECC83 it would have).

I wonder if other methods for protecting the envelope against electron bombardment had been researched - an outer shield, to be taken positive, would capture the electrons; an outer shield, to be taken negative (like the EF80 has) might repel them back to the anode. Or was there just not enough room in the EF91?

[Alan Stepney]

Good questions, to which I dont know the answer.

(Apparently the largest quantity of "EF91's" were the CV version with the domestic version only accounting for a relatively small part of production, so perhaps it made sense to treat them al the same. )

[Leon Crampin]

A shield will always increase anode to ground capacitance - so an optional external shield makes the valve more versatile.

The explanation of the blue colour is fascinating - but the principles involved are not new. Many beam tetrode output valves had an internal graphite coating to catch stray electrons which were apt to escape the anode in these valves - possibly due to the aligned grids. Some KT66s and many of the early Mazda tetrodes were made like this.

Presumably the "Mullard blue" in the EF91 is non conductive and does not add capacitance.

Who is going to start the rumour that triode connected blue EF91s sound better than "third day of the month hairpin getter Blackburn rovers" ECC83s?

Leon.

[OErjan_S]

Leon, do we really want the price to go up that much?

[Alan Stepney]

Looking back at some of the questions that were asked in the previous thread:

Getter shapes evolved, and the circular (or similar) is the most effective. It also has a more lasting effect as the getter is spread more evenly on the glass.

Yes, there were different qualities of valve for special customers, such as the military. Often they had their own type number, such as the CV numbers which may or may not, be exactly the same as the "standard" type, upon which they are based. Extra or stronger electrode supports to make them less prone to vibration, for example.

Mullar supplied all Ever Ready valves.

Philips owned (all or part) of Tungsram, and many of their valves came from within the group.

Mullard put a lot of time and effort into designing audio circuits that would give good perfomance (with their valves) and made all the techincal information freely available, and even encouraged manufacturers to use their designs. (The well known Audio Circuits.) They included enough technical information to enable manufactures to work with them, rather than just giving a basic circuit.
This was all to promote sales of their valves.
The implication is that such circuits wouldnt work as well with other makers valves.

This has, I was told, been shown as, for example, a matched pair of xxxxx might have similar IA and Gm figures but the "knee" might be in a slightly different position on the graph.
Perhaps (and this is my surmise) Mullard designed to make use of specific characteristics of their valves that were not immediately obvious from the "short form" data that most people look at.

We did briefly discuss type numbering.
I gather that Philips issued lists of available type numbers and each factory (that did design work) should have contacted Philips to get the next available number.
Apparently that took so long that some used to just "give" the valve a number (that wasnt already in use) and wait for the system to catch up.

In a few cases, identical, or virtually identical valves were developed at two places simultaneously, causing some confusion!

In other cases, where a factory had designed and released a valve, it might have been hard to back0track and change its number so it stuck, even if not in sequence.

I gather that all this happened rarely , but did occur.

[PJL]

I am looking at two yellow print Mullard valves that have come from a mono kit built tape recorder with a colllaro studio deck. The valves are both marked B4...

This would make them 1974 according to your rule that nothing earlier than 1969 used the yellow markings which seems late for the date of the tape recorder.

Then there are other claims like this

"12AU7 Mullard YELLOW shield logo, some with BVA, Great Britain.

New Old Stock White Box. Beautiful old Mullard ECC82 types direct from a radar installation, in England. These have the rare yellow label that was limited to British domestic production in the late 1950s, early 1960s. Some say these are the best 12AU7 ever. Get some while you can and hear the quality yourself! MATCHED PAIRS ARE JUST $160.00!"

I am wondering if yellow ink had been used earlier as well?