On Thu, 3 Feb 2005 09:58:31 -0500, "Jon KB1HTW"
wrote:

More info, Richard:

I climbed up on my roof last night (from my 3rd floor balcony - always an
adventure!) with a flashlight to check my RF ground - sure enough the old
cable had a splice that was broken in one place (that I could find in the
dark). So I fixed that. I also verified that the internal fuse of my radio
(IC-706) wasn't blown (the tuner connector on the back of the radio still
put out +13.8Vdc.

When I plugged everything back in, I figured I'd at least be able to tune at
least a couple of the bands, but I was met with - nothing. Pressing the Tune
button on the radio for two seconds would do absolutely nothing - no
changing mode to CW and keying for transmit, not even a short beep saying
the tuner couldn't tune the long wire. Just dead.

By then my wife was dragging me out for dinner, and it was too late after
for any troubleshooting. And this weekend, I'll be out of town, so I won't
be able to do any serious troubleshooting during daylight hours for another
week and a half.

Man, this is frustrating...


"Jon KB1HTW" wrote in message
...
Thanks for the info, but I don't think it'll work.

From http://home.comcast.net/~hamlakemn/ah4/ah4.htm (great site for AH-4
info, by the way!)

Hi Jon,

I consulted the page you offered and given the timing diagrams I would
suggest there are no issues of risetime on leading edges of the logic.
The microprocessor appears to be a state machine looking at levels,
not edges. Anyway, the timings are extremely long in comparison to
any edge issues at these scales. Even with a mile of wire, the timing
is too simple to be upset by slope.

You should have (or now do) designed this and assembled it on the
bench for testing in a warm place (for your sake). Replace the cable
so there are NO splices anywhere (if you found one broken, that is
pilot error).

As for the size of wire. This will never be a problem unless you are
running in the 40s of gauge. You MUST however meet the peak current
demand. This is for the benefit of the 22 relays. You must confirm
the average current demand meets spec too for the benefit of the
control circuitry.

Skip the attention being placed on the ground lead down. As long as
it is long (and on the roof of a three floor home, length is
sufficient) it hardly matters, RF-wise, what it ties to. The only
point about it going to ground is for the sake of lightning protection
(which seems a matter of luck anyway in this configuration).

I see no discussion of choking either the transmission line nor the
control lines on this page. This is poor coverage (or I simply
scanned it too fast). This may/may not change your situation, but
using them is standard procedure.

Get an O'scope and tie it to the control lines and confirm the timing
diagrams illustrated at the link offered. If you have no O'scope,
then patch a meter across the lines and at least confirm a twitch of
the needle. This won't prove much - unless there is no twitch. The
timing suggests this would be long enough to overcome the ballistics
of needle mass. If you have a digital VOM, and it offers AC Peak hold
reading, this should resolve at least one change of state - it will
not distinguish between the fail indication timing waveform and normal
operation illustrated at the link however.

If you have the bench talent, you could assemble a simple pulse
stretcher from ICs to aid in this troubleshooting. If you are very
talented, you could assemble ICs to stretch, clock, gate, and resolve
the fail indication waveform. An O'scope is far simpler and by far
the surest method.

73's
Richard Clark, KB7QHC