The 40823, 40822, 40821, 3N140, 3N142, and 40673 are virtually
indistinguishable from their data sheets. The 4082x were pushed more as
industrial transistors while the 3N14x and 40673 were pushed more as RF
parts. Specifically VHF RF parts. Their parameters are very close to the
same. Some were marketed as mixers, others as RF stages. Possibly just
to cater to the radio designer used to having different tubes for
NTE has a sub, its pricey. Its pricey for a reason. The world production
of these double gate MOSFETS was based on 2" diameter silicon wafers.
The modern suppliers of silicon wafers can't make anything that small. A
foot or more is the standard and the equipment for making the 2" wafers
wore out. Rather than retool the MOSFET production lines for larger
wafers or cutting 2" wafers from larger wafers, all the US makers of
double gate MOSFETs abandoned the products. Siemens has had some in
silicon, but they may have abandoned them for surface mount GaAs parts
which are hard to use as replacements because of the need for different
biases and circuit impedances to say nothing of lead stretching to reach
the pads for the to-18 case. The BF-982 is a very nice silicon double
gate MOSFET if you can find it, but it has so much higher gain than the
4082x family that it surely would need to be biased differently to cut
the gain to something reasonable.
The NTE number is NTE454.
A couple years ago Electronic Supply of Ames had a sack of 40673 that he
was wanting to be rid of, not realizing their great value of a part no
longer made. I bought a handful. I don't know if he still has some. He
has a web page (phone number 515-232-7020).
The reason I know about the demise of silicon double gate MOSFETs is
that a few years back (about a year before the last sunspot die off) I
took on a design job for a single frequency pocket HF radio. My first
thought was a gang of double gate MOSFETS for RF, mixer, and IFs. Then I
started checking for parts and they weren't to be found. So I asked.
Then I found an AM radio IC characterized only through the BC band but
sold in quantity in RS AM Flavor radios. I found that chip with an
external crystal oscillator received fine at 15.420 MHz, so I designed
it in. Had to import the chip through a broker because Toshiba didn't
support analog chips in the US. Oh well, after a few days at the bench I
know more about that chip than their data sheet admits anyway.
One could always inquire with any electronics surplus store and with
TenTec while being flexible about the device used so long its not one of
the more recent (like the BF-982 or 3N200 family) versions that have
excessive gain for fitting in 4082x RF circuits.
73, Jerry, K0CQ
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.