[Collins] FD100 Diode data sheet

[Dr. Gerald N. Johnson]

On Fri, 2009-02-06 at 20:30 -0800, jefferis at antelecom.net wrote:
> Once again, thanks for the info Jerry. I have 3 versions of
> 30L-1 manual. In the 8th edition, 15 July 1966, CR17,18, and
> 19 are still shown as 1N458. In the 11th edition, 15 June
> 1973, these are all FD100s. The update/change page in
> section 7 of the '73 edition says: "Diodes CR17,CR18, and
> CR19 were changed from type 1N458 to type FD100 to improve
> rf detection. If any, or all, of these diodes are replaced,
> it is recommended that type FD100 be used". 

I'd have to go to the attic to look, but I think the FD100s were
obscenely expensive in the 1960s, like $4 or $5 each. In one of the
service bulletins that recommends replacing the 1N252 with a 1N458 the
'458 is quoted at $1.68.
> 
> I agree, the 1N458s were not great diodes for the
> application. The FD100 is interesting from a historical
> perspective because it was the first "planar" diode,
> introduced in 1960. Fairchild billed it as a fast recovery
> diode with "normal" conductance. They also offered an FD300
> that was their "ultra-fast recovery", low conductance diode.

FD300: Bv = 150. Ir 1.0 na @ 125 v. Low leakage. If 200 ma @ 1v. High
conductance. C 6 pf. Trr not given. The high If and C would hint at it
being slower than the FD100.

I haven't put the catalog away because I discovered data on the FA4311E
and its parts used in the 32S-3A balanced modulator.

> I tossed a complete 1970 Fairchild semiconductor data book
> some years ago. At the time I thought I would never have a
> use for it. Oh well. I have not been able to find much cross
> reference data, and the Federal Stock system references to
> the FD100 do not show that there were any competitive
> substitutes ordered by the Feds.
> 
> Based on your input, it looks like the 1N4149 might be a
> better match to the FD100 based on junction capacitance at
> the test condition. Off hand, I don't think Schottky diodes
> would be right in in the tuning bridge, but might be fine in
> the  ALC circuit. It would be interesting to get the FD100
> characteristic curves.

Schottky may not handle the reverse voltage too well, like the 1N252 in
SB2.

Within the limits of the condensed catalog data, it would appear that
the 1N4149/1N916 is better than the FD100, lower leakage; while the
1N3064/1N4454 matches all the specifications of the FD100 except package
size.

I suspect that they will all function adequately in the 30L-1 and if
there is any difference it will show up at the highest frequency of RF
operation in slightly poorer rectification efficiency for the higher C
diodes. CR19, the ALC detector has an adjustable output so any improved
rectification can be accommodated, though if there is a problem it will
be frequency dependent, e.g. poorer ALC on 10m, where the tube gains are
already down some. CR17 and CR18 see the same voltage levels and
impedances for the ratio of the tuning meter, so it looks like as long
as they work the same the tuning meter will work. And that tuning ratio
has a variable capacitor to trim it for unmatched diodes.

I think we are splitting hares (much to their detriment) excessively.
That any of the 1N914, 1N3064, 1N4148, or 1N4454 families will work at
least as good as the FD100, and better than the 1N252 and 1N458. That
the FD100 remained the diode of choice simply because it was easier to
keep it in the documents, than to suggest the 1N4454 was as good or
better at a time when there was no project funding for manual or parts
engineering. With adjustments available for the ALC level to the exciter
and for the tuning balance, the quality or lack thereof can be adjusted
out. Of course, not doing those adjustments, it might be found that the
IMD on 29.6975 MHz is greater than on 3.50123 MHz by 0.73 dB, if one can
trust the IMD equipment to be that precise AND accurate and the exciter
better than that.
> 
> 73, Bob KF6BC
> 
> 

-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer