Greetings to all,
I have just finished about six weeks experimenting with PIN diode
QSK and would like to share my findings and thoughts. I have been working
with the Ameritron QSK-5, QSK-5PC, and the original HeathKit board that
Ameritron, basically, copied for their current board.
These boards seam to either work, or not, with no apparent reason
why. When they work, they are amazing, better, quieter, faster (resulting
in more readable space between dits) than any relay unit I have used. More
on speed (switching) later.
HeathKit used the same small PIN diode in all three switching
sections of their board. This seams to work fine up to the 600 to 700 watt
level. This board is easily upgraded to the full legal limit, plus a bit
more. The new boards use the small pins only in the receive section and
have eight, instead of four, diodes in series. Also, the new board uses
four of the large PINS in the Hi-power transmit switch and one, large PIN in
the Lo-power switch.
The PIN diodes, as used in the new boards, are very reliable. I had
one, on a brand new board fail in the first few minutes of operation. I
replaced it and have had no further problem with PIN diode failure. I
contribute this failure to infant mortality. So, if the PINS are not the
problem, what is causing so many boards to fail? The Hi-voltage switching
transistors are really the problem. Some of these switching transistors,
and their NTE replacements, simply will not handle the rapid switching of
the Hi-voltage reverse bias. They work, fairly reliably, in the PC model ,
with its, normally, lower reverse bias. However, they fail rather often in
the stand alone model, with its higher reverse bias. The factory must
hand select the transistors in the stand alone units. I have found that
2SC5305 transistors make an excellent replacement for the original
switchers. They are rated at 1,200 volts. I have had no failures with
these transistors. Be sure to replace the 1N4003's (or 4's) with 1N4007's.
While waiting for some PIN diodes to come in, I experimented with
1N4007's in the receive section. They worked so well that I installed
6A10's, four in the Hi-power section and one in the Lo-power section. I
was, again, amazed with the results. These inexpensive silicon diodes seam
work just as well as the expensive PIN's. In fact, measuring the
temperature of the silicon diodes while running 1,500 watts +, for 10
minutes, key down, shows that the silicon diodes run about 3 degrees C
cooler than the PINS. I have had no switching, or sequencing, problems at
the maximum speed of the internal keyer in my FT-1000D.
I asked Richard Measures, AG6K, why the silicon diodes appear to
work as good as PINS. He said that 1,000 volt silicon diodes achieve their
extra voltage rating by an increased size of their "intrinsic region"
(spacing between "P" material and "N" material--thus the "I" in PIN). If
you remember, some data sheets used to state that 1N4007's were actually PIN
diodes. Whether they are, or not, my experiments show that they work very
similar.
A good relay driven QSK unit switches in approximately 3 to 5
milliseconds (I said, a good one). That is fast enough for just about any
practical CW application. PINS switch in microseconds. I really have no
idea what the switching time of silicon diodes is, but, I think it is safe
to say, that it is magnitudes faster than a fast relay. A full set of PINS
cost approximately $155.00, plus shipping. A full set of silicon diodes
cost about $6.50. It dose not make since to me to spend all this extra
money on PIN diodes to achieve a meaningless amount of extra switching
speed. Decide for yourself.
Now we come to the real question. Is diode QSK worth all the extra
cost, trouble, and loss of versatility? That is a question that you will
have to answer for yourself. Have fun.
73 de Lon, K5JV
281-358-4207
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|