Semiconductor device aging is not a yes or no question. To pick nits,
Yes they all do, but at vastly different rate when used within their
ratings.
As has been noted, The certainly do last far longer than tubes. Compare
the maintenance of tube type radios, TVs, and rigs of the 50s, 60s, and
70s to today's equivalent. Tube replacement in the early top end
Collins and Hallicrafters rigs was common. Today, transistor failure is
rare. Finals are a special issue.
But to the specifics of semiconductor aging. It would take a long
article to address transistor aging properly, but most of today's
transistors and devices as discrete transistors are the minority of the
semiconductor devices used and age from an entirely different reason
than early devices known as the 3 leg fuse.
In the early days the problems were primarily from impurities in the
semiconductor material. Purity levels in those "early days" (1950s into
the late 1970s) for production materials were on the order of several
parts per billion, which improved to around tenths of a part per
billion. Today I understand they are into the few parts per Trillion.
I'll try to keep this on the typical ham level of the news groups, which
means I'll need to leave out a lot of the details. I'm speaking
strictly about bi-polar transistors. A search can give far more detail.
Typically semiconductor devices consist of "N" and "P" type materials.
An NPN transistor can be thought of as a low voltage triode tube without
the vacuum. A material, called a "dopant" (
https://en.wikipedia.org/wiki/Dopant ) , is added to pure Silicon to
make it either N or P type. But all pure silicon is not "pure". Whether
N, or P type a tiny amount of the other material exists in there as
well. In those early transistors the impurities were considerable
compared to today's materials.
NOTE: Tubes are voltage devices and transistors are current devices.
Exceptions can be found.
The junction where the N and P type materials meet is tiny. It's on the
molecular level. The P type "dopant" will tend to migrate across the
junction to the N-type material, contaminating the N-type. This occurs
with age and is bi-directional, but is accelerated with increased
temperature.
This cross contamination was the primary reason for aging in early
semiconductors.
Technically the cross contamination still happens in junction type
transistors, but with today's purity levels the rate is so low that it
can normally be ignored with transistors lasting far longer than we do.
There are other reasons for aging, but an explanation gets complicated
and their contribution is normally miniscule to devices with very long
lifetimes.
There is a new type of aging in some devices. With some devices
operating well into the GHz range, the device may consist of very small
components, measured in the micron range. They are so tiny that
electron travel actually causes wear. Computer CPUs are now typically
running into the GHz range. That "chip" may contain many millions of
transistors so small it would take a microscope to see them.
This certainly must have an effect on ham communications up in the GHz
range. Who ever thought we'd see circuits so small that electrons going
around a corner could cause a wear problem
For a background, I worked in the semiconductor industry for over 26
years. I have a Bachelors degree in Computer Science with a Math minor.
Please excuse my one handed typos and spelling errors. I make my own
worst proof reader.
73
Roger (K8RI)
On 12/17/2015 Thursday 7:59 PM, Charles H wrote:
On this reflector, I have poo-pooed the idea of baby-ing your radio especially
by turning it off often, lowering xmit power, etc.
However, I read that with ordinary tubes, the chemistry of the cathodes and filaments are affected
by life-of-use, so that tubes do "wear out" with use. Tubes can be rated in terms of
"hours of use" with the type in a KWM-2 rated at around 2,000 hours of useful function.
Thorium in the chemical make-up of the metal parts in a tube, maybe a metal or ceramic tube
(usually added to higher power tubes) can extend tube life to about 100,000 hours.
The conclusion is that "hours of use" of course do matter for hams intending to use their
tube-type radios for thousands of hours. The hours of use can be limited by simply turning the
tube off when having only stand-by function. However, there is some suggestion in the literature
indicating that every "turn on" of a tube (or an incandescent light bulb) involves rapid
heating of the elements inside which could, immediately or over time, deform elements and causing
sudden failure (given that the metal elements do heat unevenly--due to resistance in the metal--if
only for a very short duration of time). Repetition of on-and-off cycles are thus a factor in
lower tube life.
Other than these factors, it appears in the literature that operating a tube,
particularly high power transmitting tubes, at its rated values can extend life
IF it is otherwise not sometimes operated below or above its ratings. It
appears that, for example, driving an RF amplifying tube at its full rated
power does not materially decrease its life. Reducing drive appears not to
extend such tubes' life.
It is all a balancing act. 73, Charly
P.S., next question is "do transistors 'wear out'?"
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|