The Radiokit QRP40 is on its way to the airport as I write this, with
guaranteed delivery to Doug Hendricks by Friday. (I intended to include
the adapter cables that I put together, but I forgot. Sorry, Doug. Did
you know that your zip code was one for which they will not promise to
deliver the next day? Where are you, in the middle of the desert?)
I have my fingers crossed that the rig will arrive in working order and
that the judges will be satisfied with the results. I got mail from
debuggers number 3 and 4, and I think they are hoping even harder!
As I believe I said in a previous post, this contest was more of a team
relay race than a competition between opponents. Larry did a great job on
the first leg of the race, I did the best I could on leg two, and came
close to having to hand off to debugger # 3. There was a lot of work to be
done to get the radio working. I would not have been able to finish in
seven days had I been the first contestant. However, both of us together
managed to achieve better results in two weeks than I think either of us
would have done alone.
It is perhaps interesting to compare styles. Larry attacked the problem by
checking and double-checking the details. Since the documentation was so
lacking, he gathered all the reference material he could about the design
and looked for errors and inconsistencies. He methodically checked each
and every part. He recalculated many of the design equations to find the
errors in parts values. He disassembled, tested, and reassembled. It
takes a great deal of patience to do this, but it was absolutely necessary
for this kit. I remember reading comments on the list wondering why he
didn't just plug it in and look for smoke. I have to admire his willpower
to get it right even when the urge to 'go for it' must have been very
strong. I'm very glad that Larry did this preliminary work, because it
permitted me to do the part that I enjoy the most. I turned the board on,
and worked through each stage, playing as I went. At each step, I tried to
predict what I would see, then confirm it. If it wasn't what I expected, I
had to create a hypothesis about why it would behave that way, then test
again. I played the game of "what happens if I do this?" over and over,
trying to understand what was happening in the circuit. Everywhere I
could, I tried to confirm correct operation in more than one way. In his
last few days, Larry started on this part of the debugging but ran out of
time. He would have had a time advantage at that point, because he had
already simulated all of the circuits in his mind and on paper. It was
clear to me after re-reading his notes that he had suspicions about many of
the problems that I encountered.
Debugging takes both sweat and inspiration.
Why was this kit so hard to get working? The Rick Littlefield design is
basically sound. The fault lies in the implementation. The board had no
silk-screen. The parts placement drawing did not exactly match the board
layout and most of the components on the drawing were identified only by
value, not reference designator (i.e. R1, L2, etc.). Different versions
of the kit work on different bands using different component values;
however, the documentation was not very clear about which versions needed
which modifications and which values. Some of the components supplied with
the kit were the wrong value. The kit builder had ample opportunity to
fail, and it required a lot of effort to figure out what was right and what
was wrong. Some of the problems I encountered could be blamed on the
design, but could also have been averted by some warnings in the documents,
like "do not use a socket for the MC1350 because this may tend to make the
IC oscillate". All kit suppliers could learn a little from Heathkit.
I don't know if you had as much enjoyment out of the exercise as I did, but
I would like to see the contest repeated. I would love to see how other
people attack problems and find solutions. Thanks Doug and Chuck. Maybe
you can make a QRPp article out of this.
Now, it's time to get back to organizing the 50/40/30 challenge.
Mike K1MG
[
QRP-L Archive |
]
[
1993 |
1994 |
1995 |
1996 |
1997 |
1998 |
1999 |
2000
]