Rick at al,
My comments in line
On 2/28/2021 12:31, Richard (Rick) Karlquist wrote:
On 2/28/2021 7:40 AM, Grant Saviers wrote:
The nice thing about switched serial caps (if same values) for tuning
upwards from a low resonance is the voltages are all the same. Plus
caps and appropriate relays are cheap, easy, and small.
Rick, I think you might have suggested this to me, and my
implementation was published in QEX May/June 2019.
Grant KZ1W
Grant, Dave etc:
I hadn't seen this article: the DVD was still shrink wrapped :-)
Anyway, nice write up, it's similar to what I did, so maybe
I suggested it. A couple of discussion points (not criticisms):
1. I use binary weighted cap values. In the article you say
you going to do that but a comment on this reflector changed
your mind to equal values. You mention in your posting today
that this is nice because all the voltages are the same. Is
there any other reason for this decision? The advantage of
the binary weighting of course is that only two caps and only
two relays are needed for four segments (or 3 and 3 for 8 segments).
For my T, 85ft up and 2x 33ft arms, 8x 125ft 10ft up radials, resonant
at 1825+/-, the 40 KHz almost equal steps modeled with one value cap
4000pf, so the series values were 4000, 2000, 1666, and short. Works
per the modeling. That yielded the swr curves shown. There is an error
in the schematic as wire numbers were mixed up and ground doesn't
connect control to relay box properly. Other ones don't matter since
all caps are the same value.
2. Figure 7 shows 6 relays to get "all decodes"? Perhaps this
got messed up in Newington. Can you explain it better?
The control box PCB was a general purpose diode decoder, up to 6
outputs. only 3 used in this application.
3. Figure 8 shows your 50 ohm t0 25 ohm matching transformer.
The text credits K9YC and then goes on to say it is a transformer,
a balun, and a common mode choke, all in one. I know K9YC
very well and he correctly rails against the misapplication of the
word "balun". Maybe more QRM from Newington?
Choke and TLT/unun are separate. Choke about 5k ohms RG142 on FT240-31,
TLT 50:22 Balun Designs. Pre K9YC designs with the 4" cores.
Anyway, if you were going from 50 ohms to 12.5 ohms, this loose
talk could almost be true. But AFAIK, going from 50 ohms to
22.2 ohms (3:2 turns ratio) is simply an unun, which is an
autotransformer. At least that's what I'm doing. It could
never be confused with a transformer, balun or common mode choke.
4. I like the Schrack relays. I see that DigiKey still has
these available, but they identify them as "TE Connectivity
Potter & Brumfield" but they also say "Series RZ, SCHRACK."
I may use those going forward in place of the ones I
used before. You mentioned the CDE mica caps I used,
and then said you used cheap Ukrainian ones. Do
you by any chance have a pointer to them on ebay? I realize
that 4 years is a long time in ebay years, so maybe not.
I looked on ebay for some more of the Tx caps a few months ago - no luck.
BTW, an additional advantage of using capacitors instead of
a tapped inductor is that the inductor will lose a lot of Q
if a metal enclosure is used, unless it is huge. A plastic
enclosure is clumsy and self destructs in sunlight.
Plus sub one ohm inductors are pretty hard to make vs the Q of silver micas.
Finally: have you gotten a lot of response to the article,
especially in terms of people actually building it?
Maybe I need to jump on the frequency agility bandwagon,
if there is a critical mass of other stations to work
who are on board.
One builder who pointed out the wire number error. Besides agility for
SSB, I built it to operate in the now defunct JA band plan.
Keep the good ideas coming guys!
73
Rick N6RK
_________________
Searchable Archives: http://www.contesting.com/_topband - Topband Reflector
|