"Associated Semiconductor Manufacturers".
 

ASM was a 'company' created in the 1960s by the merger of Mullard/Philips UK semiconductor manufacturing with that of GEC (who were owners of Ediswan/Mazda/AEI/Thorn).

AEI-Thorn transferred manufacture of Mazda transistors from Briimsdown to AEI's Lincoln factory which already produced devices for professional use,

It didn't seem to go well: to quote from _The Setmakers_ "Arnold Weinstock of GEC did not want to make entertainment semiconductors and asked Thorn to remove the Mazda plant from Lincoln; it was in production for just eight months".

After which there seems very little trace of what happened.

https://www.gracesguide.co.uk/Associ..._Manufacturers

https://www.electronicsweekly.com/ne...ustry-2010-09/

The UK semiconductor industry's established players seemed really rather slow to move off Germanium and into Silicon. Some 'upstarts' like Plessey licensed US Silicon technology from Fairchild etc and made a go of it for a couple of decades.

What actually became of ASM in the end??

[And what, for that matter, became of Newmarket transistors? I've got fond memories of bunches of NKT404 TO-3 cased transistors in inverters/modulators of Pye two-way radios from the 1960s, and the off 'lozenge' transistors in the NKT2xx series in similar-vintage transistor radios]·

Re: "Associated Semiconductor Manufacturers".
 

Quite a convoluted chain, but they eventually became part of GE Aviation it seems:
https://www.pye-story.org/companies/...et-transistors
https://www.suffolknews.co.uk/newmar...ished-9076621/

Re: "Associated Semiconductor Manufacturers".
 

That presentation is a fascinating bit of history. Along with this:

http://www.wylie.org.uk/technology/s.../Newmarket.htm

The "NKT600" series look rather like Mullard's AF11x/OC17x series... I wonder if there was some expedient cross-branding going on?

In the 1960s there was a range of Germanium "GET" transistors like the GET115/120

https://www.langrex.co.uk/products/g...ansistor-x1pc/

that came with a built-in heatsink. Part of me wants to buy a few and experiment using them in transformer-coupled push-pull amplifiers. I remember there was even a 'matched pair' fitted to a double-wide heatsink of similar style.

My big memory though is of the Ediswan 'Top Hat' XA/XB/XC series of transistors - equivalent to Philips/Mullard's first-generation OCxx transistors that were rather-more-common. i remember building an audio-amp using a pair of XC131 which was just as good as the OC81-powered equivalent.

Re: "Associated Semiconductor Manufacturers".
 

When I was in the DC drives sector in the 70's the Lincoln plant was our supplier of choice for commutation SCR's (up to about 100A/ DO package)
Main devices upto 1000A + in capsule package mostly came from Westcode

Ed

Re: "Associated Semiconductor Manufacturers".
 

Ferranti didn't do Ge at all, silicon from the start in 1955. By 1962 they were making DTL ic's and by late 60's TTL by the million. There was some tie up with Fairchild, perhaps to make certain devices under licence
I applied for a job at " Mullard, Hazel Grove" in Stockport , I think it was 1977, I'm sure it was part of ASM then. Didn't get the job, but it looked a very modern manufactury then.
Is it still operating in another guise?*
* Yes,it is:"Nexperia,Greater Manchester" Looks like Chinese owned .

Re: "Associated Semiconductor Manufacturers".
 

Interesting you mention Fairchild: their commercialisation of silicon planar technology along with low-cost epoxy-resin encapsulation in the early-60s was a real shakeup to the industry: along with the little brown 'jellybean' transistors they did a "Micrologic" [uL900] series of DTL IC devices in the same timeframe. I've heard it said that the flatter-with-2-corners-rounded Ferranti "ZTX" series of plastic transistors which were everywhere in the 70s and 80s were a refinement over the Fairchild Jellybean, to better accomodate automatic-board-stuffing assembly methods.

Back when the first Fairchild jellybeans appeared, the likes of Mullard/Philips were still using metal cases [AF11x and AC1xx-series] or glass [OC44/45/70/72/81 series] which must have been far more expensive to manufacture; in consumer-gear every tenth-of-a-penny counts!

Mullard/Philips responded with Lockfits - which history has shown to have been a Bad Idea.

I must get out and re-read my copy of "The Setmakers" and see what more is mentioned about ASM and the UK semiconductor industry.

Re: "Associated Semiconductor Manufacturers".
 

I'm still not convinced all lockfits (or lockfits from all factories) were bad, but they must have been vulnerable to manufacturing defects since they did get that reputation at least in the UK.

Also, the leadfree versions of TO-1 and TO-7 encapsulations might have been an even worse idea.

Re: "Associated Semiconductor Manufacturers".
 

Ferranti plastic transistors, "ZTX", then known as "E-line" once in mass production, c1972?,were made for a penny each. The cheapest, lowest spec device ,ie just about working , sold for 4p each. I doubt Mullard came anywhere near that with their metal and glass constructions but they may have seen the light by then and be producing BC10x etc plastics.

Re: "Associated Semiconductor Manufacturers".
 

What is a lockfit tranny? How do you spot one? I searched but could find no pics.

Andy.

Re: "Associated Semiconductor Manufacturers".
 

They were a square cross section but with two corners significantly chamfered. and the legs has a couple of barbs on them so that when pushed into the correct diameter hole they stayed put while you turned the board upside down to make the solder joints.

Our own Mark Hennesy has some images on his website here http://markhennessy.uk/articles/vintage_transistors.htm about two thirds of the way down

Craig

Re: "Associated Semiconductor Manufacturers".
 

A plastic-encapsulated small-signal transistor. The package is appreciably larger than TO-92 and has a recognisable shape once you've seen one.

The lockfit naming refers to the shaping of the leads. The leads are pressed from flat material and have little side lugs to prevent the transistor being inserted too far into the board as well as little lugs to act as barbs to stop it falling back out before wave soldering. A first generation of parts designed for auto-insertion.

Since then, they've acquired a fully-justified reputation for failure. Whether it's internal fracturing or corrosion from the plastic compound on the bond wires, I don't know. They are on most people's 'change on sight' lists. Not an if-it-fails part, but a when.

Most of them carried die from well known types previously in TO18 packages, so equivalents are easy.

David

Re: "Associated Semiconductor Manufacturers".
 

One of these.

Re: "Associated Semiconductor Manufacturers".
 

The Philips group lockfits failed my water ingress tests when I was working in the automotive business in the '70s. The problem appeared to be due to the stamped out ribbon "pins" of the device not being fully sealed to the encapsulant.

I discussed this with the Mullard rep at the time without reaching any satisfactory conclusions. The TO-92 devices which I used as a control in my experiments were, I think from Motorola and of course had round section leads. These were generally OK and received lab approval for limited (non engine mounted) applications. T0-18s were the package of choice for critical jobs.

I was sorry to see the lockfits used generally and largely I think in the UK. Another "Hunts" saga, brought about by inadequate testing - often the result of blindly following an existing (often UK military) spec as opposed to designing a test to reveal real weaknesses.

It grieved me at the time to see that a Hong Kong transistor radio bought on a market stall often contained components of better quality than those specified for a Racal RA17. Carbon comp resistors anyone?

Leon.

Re: "Associated Semiconductor Manufacturers".
 

It would make more sense to replace them with their TO92 (SOT54) equivalents as they are cheaper and in various aspects just a little bit closer equivalent (plastic can with similar properties instead of a metal can).

For example BC107-BC147-BC237-BC407-BC547-BC847 were produced using the exact same crystals, so replacing a BC147 with a BC547 would be the most logical choice.

Same goes for BF194-BF494, I think.

While the lockfit types seem to have a well deserved bad reputation in the UK, I still haven't figured out why they don't have that reputation on the continent. Lots of stereo and TV sets equipped with those transistors that haven't given any trouble. There might have been a difference in production process between Mullard (I think they used the letter G in the date code from around 1971 on) and other factories. Or maybe there was a change somewhere during production.

Re: "Associated Semiconductor Manufacturers".
 

That goes some way in explaining this "Lincoln" reference.