Talk:BC548

Split off series sections

The series sections should really be separate articles, perhaps with this redirecting to the BC series article.

• BC series silicon transistors
• OC series germanium transistors

65.93.12.101 (talk) 03:52, 27 March 2011 (UTC)

It's a reasonable direction to go in, but those two groups are too broad. OC71, OC81, OC91 and the OCP71 could well be covered by one article (probably just called OC71), but the OC28 is quite a different beast, even though germanium and a contemporary. Likewise BC is even broader - the BC107, BC108 & BC109 go together, but the BC548 et al. is a few decades later, with quite a bit changed in the intervening. There's scope for "series" articles, even bundling the complementary PNP & NPN together, but doing it just from the naming prefix is too wide. Andy Dingley (talk) 11:07, 27 March 2011 (UTC)

We have an article called Transistor and even an article called Bipolar junction transistor - is there any reason the history of development of these devices can't go into existing articles? At least, until the development and references get so big that it needs to stand on its own. --Wtshymanski (talk) 16:08, 27 March 2011 (UTC)

Transistor is much too general, BJT is also too general. These are two different classes of BJTs, silicon based and germanium based. The semiconductor industry revolves around the substrate... germanium, silicon, gallium arsenide, etc; these should all be separate articles since they have different characteristics. 65.93.12.101 (talk) 05:48, 28 March 2011 (UTC)

BJT is about the theory, these are about the practical instances that became commercially significant.

2N2926 is another one that warrants mention, because of the practice in the mid-70s of selling them as colour-coded post-manufacture sorted examples according to their measured gain. Manufacturing of silicon transistors at that time had become cheap and reliable, but couldn't make parts to a consistent specification and so they needed to be sorted afterwards. Andy Dingley (talk) 10:20, 28 March 2011 (UTC)

Merge "BC548" specific material to 2N3904

A merge was suggested at the wikiproject, that BC548 be merged to 2N3904. This should obviously involve BC548 specific material. The series/family material would probably be better in a different article. 65.93.12.101 (talk) 05:53, 28 March 2011 (UTC)

I'd want to check the specs carefully before deciding (although I suspect not). As "benchtop parts bin" transistors, these two are regularly swapped around (I'm in Europe. I've probably never used a 2N3904, but I bulk-buy BC548). However although they're substitutable for one another, they're not identical to one another (pinouts are, I think, different). Nor are their manufacturers commodity-manufacturing them to a common spec. For that reason they're not close enough to merge. Andy Dingley (talk) 10:25, 28 March 2011 (UTC)

You might want to check out what was done at 2N7000 then... 65.93.12.101 (talk) 18:51, 28 March 2011 (UTC)

Is it the same device sold under both the JEDEC and Pro Electron numbers? Or is it just "close enough" for writers of TAB books? If it's got different pinout, ir surely can't be registered under both numbers. --Wtshymanski (talk) 19:36, 28 March 2011 (UTC)

Two different transistors, with power ratings where one is more than twice the other, are obviously not the same device. 8-( Andy Dingley (talk) 19:55, 28 March 2011 (UTC)

Oh cool. You can apparently get the JEDEC standard .PDFs for free on registering with their site ( which hasn't worked for me yet). Doesn't sound like a good candidate for a merge. --Wtshymanski (talk) 20:04, 28 March 2011 (UTC)

Package

Why do Pro Electron devices come in JEDEC packages? Surely they would at least have come up with their own metric packages. --Wtshymanski (talk) 03:27, 4 April 2011 (UTC)

What time period are you talking about? There's a lot more use of common package formats today than there was in the '70s. As JEDEC packages are workable, the obvious encouragement is to use them for nearly everything and only use a non-"mainstream"_(sic) package if there's a real need to, leaving the metal TO18 cans to those devices that really need them. The increased shift from hand-assembly to pick-and-place is also going to favour standard packages and in-line leads, rather than triangular pin spacings (easier to hand place, but harder to machine place). There's also the cost aspect, as plastic packs are cheaper than cans.

For this particular transistor, there's a widespread view (dating back to the '80s) that it's just the old favourite BC107 (metal TO18 can) in a TO92 plastic package with flat leads. I don't know how true this ever was "under the hood", but it was a widely held pragmatic viewpoint that built an awful lot of circuits. Andy Dingley (talk) 09:04, 4 April 2011 (UTC)

Independent of time period; surely the European manufacturers couldn't bear to adopt American standards? The Chinese, of course, will build anything you like to any price point you like. And the above discussion (with citations) would be interesting in an article about Transistor packages which could merge in the rat-bag listing of random TOXXX articles we have now. Organizations such as National Semiconductor regularly put in the process and die used to make various transistors, this nformation can't be too super secret if National is disclosing it. There's got to be something written about this that's more authoritative than "Build this code practice oscillator in only 13 separate articles" that the hobby magazines print. --Wtshymanski (talk) 13:35, 4 April 2011 (UTC)

Notability

This part number is not notable and is not a suitable topic for an encyclopedia article. Even history books about the semiconductor business don't spend any time on individual part numbers, because they aren't important to understanding semiconductors and thier roles. --Wtshymanski (talk) 13:38, 5 April 2011 (UTC)