Why are most semiconductor cases black ?

[Miyuki]

Hi folks,
I just saw some vintage NEC transistors and they are in green epoxy mold (TO220)

I newer saw it before, just common black/gray ones

Have it some reason or why we have no colorful semiconductors ? 

[radar_macgyver]

Semiconductors are generally light sensitive, so black epoxy will prevent them from reacting to ambient light.

https://www.raspberrypi.org/blog/xenon-death-flash-a-free-physics-lesson/

[Zero999]

Semiconductors are generally light sensitive, so black epoxy will prevent them from reacting to ambient light.

https://www.raspberrypi.org/blog/xenon-death-flash-a-free-physics-lesson/

Presumably they rules out the EMP generated by the capacitor in the xenon flash gun discharging, which I would have thought would be much more likely to cause this.

[RoGeorge]

Semiconductors are generally light sensitive, so black epoxy will prevent them from reacting to ambient light.

https://www.raspberrypi.org/blog/xenon-death-flash-a-free-physics-lesson/

Presumably they rules out the EMP generated by the capacitor in the xenon flash gun discharging, which I would have thought would be much more likely to cause this.

The EMP itself might be enough, but I tested myself by pointing a pen size green laser on the RPi's LDO. It resets the RPi instantly, every time.

[Synthtech]

I’m sure that some university “professor” will soon publish a paper on the “Patriarchy of semiconductor case colouring and it’s implications for climate change”.

[coppice]

I think a more interesting question is why are cases sometimes not black. Black is cheap, and cheap is good. However, some parts, like opto-couplers, are made in large quantities in other colours. Why is that?

[mariush]

Optocouplers use infrared or other wavelengths, and other colors than black may be better for that purpose.   

Also have some vague memories about something to do with heat (retaining, absorption, fluctuations etc)... black is worse than white or other colors when it comes to heat.

[BrianHG]

 

[BrianHG]

Now for Orange:

[CopperCone]

replaceable black parts chained down by copper doing all the work on green boards someones gotta change that

[wraper]

Also have some vague memories about something to do with heat (retaining, absorption, fluctuations etc)...

Black is good for heat radiation and absorption, yes it works both ways. Although what is black in visible spectrum is not necessarily in IR and vice versa.

black is worse than white or other colors when it comes to heat.

Define "worse".

[jaromir]

Russian semiconductors often did have different than usual black epoxy bodies, like
https://commons.wikimedia.org/wiki/File:USSR_K155LN1_(%3D7404).jpg (though they had different specialties, like this 8086 clone - http://www.cpu-zone.com/8086/DSCF1285.JPG )
or jelly-bean transistor KT315
https://en.wikipedia.org/wiki/KT315#/media/File:Kt315b.jpg
though typical KT315 package has orange-red color http://dragmet.netdo.ru/uploads/images/73/3f/733fae540f99fff0d6d6542c0101e64a.jpg

The brown color of IC packages earned them a lot of unflattering nicknames.

[BrianHG]

Darn, cannot find a photo of those really old blue TRIACS you used to be able to get in TO-220 package.

[wraper]

Darn, cannot find a photo of those really old blue TRIACS you used to be able to get in TO-220 package.

This?

[T3sl4co1l]

Both pictures showcase something else unusual: the full breadth of the TO-220 package standard.  The leads are allowed to be round or rectangular, the body dimensions can be all over, and the tab thickness can be as thin as the leads, or as thick as the more familiar TO-220AB (which is what most people know today, and what most think of, when thinking of TO-220!).

Tim

[T3sl4co1l]

Regarding color, I've seen green PNP power transistors more often than NPN.

There are also white packages typically for optos, which usually improves CTR slightly.

Package color is irrelevant at radiation wavelengths, in the long IR.  The resin completely dominates.

Tim

[BrianHG]

Darn, cannot find a photo of those really old blue TRIACS you used to be able to get in TO-220 package.

This?

Yes.  Here are blue-ish pnp transistors.

I don't remember seeing white TO-220, but I have seen a true light grey case.
We are missing brown, yellow and purple needed to finish the rainbow.

[Nauris]

Both pictures showcase something else unusual: the full breadth of the TO-220 package standard.  The leads are allowed to be round or rectangular, the body dimensions can be all over, and the tab thickness can be as thin as the leads, or as thick as the more familiar TO-220AB (which is what most people know today, and what most think of, when thinking of TO-220!).

Tim

But ain't that a TO-202 that that SC136BI is in?

[IanMacdonald]

What I can't understand is why they print the numbers on a black case with black ink. 

[wraper]

What I can't understand is why they print the numbers on a black case with black ink. 

I don't know any of such components.

[BrianHG]

What I can't understand is why they print the numbers on a black case with black ink. 

Are you talking about laser etched/burned printing on components.  Laser printing uses no ink, cant be easily rubbed off, 10x finer print, 10x faster printing, and the printer lasts almost indefinitely compared to paintjet or silkscreen print, but,,, there is that annoyance that if you want to read the numbers, your source light need to be at an off angle to easily see the burned depressions.

[T3sl4co1l]

But ain't that a TO-202 that that SC136BI is in?

"Both pictures" as in, the ones picturing TO-220s.

Tim

[BrianHG]

  Surprise, look at what I found!
Though I believe that's ceramic, not plastic.

[SeanB]

Purple ceramic on a RF power device is almost always a very strong indicator the ceramic is based on Beryllium oxide, a very good thermal conductor, a very good electrical insulator with very low RF losses, but also incredibly toxic in the form of dust if the package is broken or handled roughly.

Purple or pink is dangerous, a white ceramic is almost always alumina, which is safe to handle.

[CJay]

Regarding color, I've seen green PNP power transistors more often than NPN.

Tim

Similar with MOSFETs, I've a pile of complimentary Hitachi parts, the 2SJ P type parts are green, the 2SK N types are black.