Hi John
In a message dated 96-04-05 10:17:26 EST, you write:
> By paralleling two tubes this was reduced to
>approximately 52 ohms which made a nice match for the
>exciter. But there is a penalty for eliminating this matching
>network. The RF ground for the tubes is now returned through
>the exciter, making cable length and tuning of the exciter an issue.
>The results are an increase in IMD and, sometimes, in unstable
>tuning.
Great point John! Most people don't understand that, and think SWR is the
only issue. The problem is, as you said, rooted in the long path for the
higher order harmonics. That can cause "rounding" of the switching transition
when the tube crosses into cutoff. The largest effect I noticed is efficiency
decreases greatly. The WORSE possible length of cable is a odd multiple of a
1/4 wl at the third harmonic of the desired band. To avoid that, the amp MUST
have a low pass type CLC or parallel LC network. High pass or low pass "T"
networks won't work!!!
> It is a much better design to include a return path for
>the cathode within the amplifier. Doesn't take much of
>a network, either L or pi will do and the Q can be as low as
>2, so there should not be an issue of tuning the input. Lower
>IMD will result, making for happier neighbors on the crowded
>bands during contests and it will be somewhat easier to drive
>the amplifier.
IMD results are a twenty page paper, but that is a good general rule. The
sure thing is efficiency will almost certainly suffer.
>----additional notes on grounded grid operation----
>
>The 4-1000 is a good tube in grounded grid but just doesn't
>have the gain of a tube designed for gg service.
The gain of any GG amp is a function of the output impedance of the tube over
the driving impedance of the tube (simplified). The higher the driving
impedance the lower the gain. The 4-100 has over 100 ohms of driving
impedance.
>But stability
>can be an issue on the higher HF bands due to the normal
>resonant frequency of the screen grid and its phase shift
>that results in a tendency for positive feedback near the
>bottom end of the FM broadcast band. This can be reduced
>by adding a little inductance in the ground lead of the screen,
>thereby reducing its frequency a bit and producing a phase
>shift that tends to reduce the tendency for oscillation.
That moves the self neutralizing and parasitic frequency lower. We have to be
careful since neutralizing and parasitics are complex subjects. The best
solution is a resonant suppressor system tuned to the FM band, where the tube
actually tries to oscillate.
By the way, I've yet to see an HF amplifier with high VHF or UHF parasitics
like ones a certain "guru" claims exist. The damaging stuff is all most
always between ten meters and the FM band. The closer to ten meters, the
harder the tube is to stabilize. 811's, 572's, and 3CX1200's are good
examples.
>Just grounding both the grid and screen in the 4-400, 4-1000
>series tubes resulted in the grid hogging the current and limiting
>the tubes capabilities. A technique to provide better division
>of current between the grid and screen in grounded grid is to
>tap the grid connection down on the filament choke. This results
>in some RF being applied to the grid as well as the cathode and
>the tube will have a better balance between grid and screen
That's a bad scheme to use in the general case. The grid is no longer
grounded, and this adds positive feedback to the stage. It decreases
stability and increases IMD. The grid ALL should be grounded for RF in a GG
amp. The correct way to "balance" the currents without adding unwanted
feedback is by the addition of regulated dc supply voltage to the heavily
bypassed screen and grid elements.
73 Tom
>From aa4lr@radio.org (Bill Coleman AA4LR) Fri Apr 5 17:31:38 1996
From: aa4lr@radio.org (Bill Coleman AA4LR) (Bill Coleman AA4LR)
Subject: A 15m observation WPX SSB
Message-ID: <v01540b02ad8b05944bb2@[206.28.194.40]>
>I played around a bit with the
>Iding portion of the running. Is it me or have US stations
>become even more impatient on the bands? I ID'ed after every QSO
>and never had someone ask me,"what's your call?".
If you did, I'd label that guy a goofball.
> I tried IDing
>every 2 qsos and was asked many times after my "QRZ", "what's
>your call?".
I think this goes back to my theory that pileups are a lot more dynamic
than people think, with listening stations coming and going very quickly.
But imagine what it is like if you are a station S & Ping (and trying to do
so as quickly as possible), and you run across:
DX: "... You are 59 08"
Station 1: "Thanks 59 03"
DX: "Thank you, QRZed?"
At this point, you have no idea what the DX might be, and you've already
heard him fail to ID after a contact. You've already sat through one
exchange, you might have to sit through 10 more. (this is because goofballs
at the OTHER end of the microphone that insist that they shouldn't ID
except every 10 minutes if they feel like it....) You really have no idea
how long it will take for the DX station to ID.
> I remember being able to run 3 stations straight
>before IDing and never having someone ask me that question.This
>was back in the 80s.
Contesting has gotten a lot more intense since then. Computer logging has
really picked up the pace (not quite as much pencil and paper housekeeping
to do).
> I don't understand why a station cannot
>wait for at least one transmission to go by to see if a station
>will ID or not.
In my above example, he did. If you ID every Q, then a station never has to
sit through more than one exchange to get a DX callsign. (Unless they
miscopy it)
>Serious in the contest or not,asking that
>question without giving your own call is illegal and usually
>winds up interfering with another station who is calling.
I'm now aware of any FCC regulations that state that you cannot ask a
question without giving your call. The requirements state that the
identification must be given every 10 minutes and at the END of
communications. By my interpretation, staying on that frequency means that
communications hasn't ended. In fact, it is that SAME interpretation that
allows you, as DX on KP4, to only ID every other Q.
I'll agree with you that asking the question creates unnecessary QRM. (I
think we will disagree as to whose "fault" it is....)
> I
>suspect that the culprits are mostly serious contesters who
>don't want to waste precious seconds waiting for an op to
>identify on the second time around. Anyway, I ID'ed after each
>QSO 98% of the time which worked fine for me.
I think the intemperance is due to intolerance of really bad operaters
failing to ID after many, many QSOs (then arguing when they are asked to
identify -- sheesh).
I used to be a hard-liner about IDing after every Q. Now I take a more
relaxed attitude about it. 98% of the time is probably about right (49 IDs
in 50 Qs).
Your experiment tells me that ID to Q ratios close to 1 are a really good idea.
Bill Coleman, AA4LR Mail: aa4lr@radio.org
Quote: "Not in a thousand years will man ever fly!"
-- Wilbur Wright, 1901
>From Bruce (AA8U)" <aa8u@voyager.net Fri Apr 5 18:51:58 1996
From: Bruce (AA8U)" <aa8u@voyager.net (Bruce (AA8U))
Subject: Announcing HamVention Contest
Message-ID: <199604051851.NAA22471@vixa.voyager.net>
>To: Spike Lazar <slazar19@sgi.net>
>From: "Bruce (AA8U)" <aa8u@voyager.net>
>Subject: Re: Announcing HamVention Contest
>
>>
>> Info: For your pseudo checksum # contact AA8U.
>>
>I have found your posting at times humorous, but more often superfluous and
annoying. However, I stand by your right to post them regardless. I
interpret this latest posting as not so thinly veiled sarcasm. Like I said,
I have very good duck feathers. Guess this one will require their use.
>73,
>Ugly
>
>From mraz@rdxsunhost.aud.alcatel.com (Kris I. Mraz) Fri Apr 5 19:10:15 1996
From: mraz@rdxsunhost.aud.alcatel.com (Kris I. Mraz) (Kris I. Mraz)
Subject: Tower Detuning
Message-ID: <9604051910.AA09055@maverick.aud.alcatel.com>
Bruce, AA8U, said:
> Make sure you towers are detuned and suspend the 4sq on one of them.
Others have also mentioned tower detuning. I assume this means to make
the tower system (tower, guywires, yagis, etc) non-resident at some
selected frequency of interest (80m and/or 160m). Is there any science
to this or is it basically a cut and try proposition on a case by case basis?
What is the usual methodology? What detuning techniques are used?
Have there been any articles written addressing this subject?
73
Kris AA5UO
mraz@aud.alcatel.com
>From Larry Tyree <n6tr@akorn.net> Fri Apr 5 19:28:15 1996
From: Larry Tyree <n6tr@akorn.net> (Larry Tyree)
Subject: SprINT rules
Message-ID: <199604051928.OAA14601@paris.akorn.net>
TENTH INTERNET CW SPRINT CONTEST
Contest period: 01:00:00Z to 03:00:00Z on Sunday April 28th UTC. This
is Saturday evening in the USA.
Bands: 40 and 20 meters only (this is a real radio contest, no internet).
Suggested frequencies are 7030-7050 and 14030-14050.
Max power output: 150 watts at transmitter output connector.
Exchange: Consecutive QSO number (starting with one), name and state
or province or DXCC country (if outside W/VE). The name for
the first QSO is your name. For every QSO afterwards, the name
you send is the name you received in the previous QSO.
Call: CQ INT
The standard sprint QSY rule must be followed. This means that if you
solict a QSO (ie: with CQ or QRZ), after completing the QSO, you must
QSY at least 1 kHz before calling another station, or 5 kHz before
solicting another QSO.
Both callsigns must be sent during the exchange. Only one signal at a
time please and all QSOs are to take place on CW. All information
submitted must have been decoded during the contest. The use of post
contest detection or verification techniques or systems is not allowed.
Also, do not make round robin type QSOs. It will be very easy to spot
these with the names floating around. A round robin QSO is one where
you should QSY, but instead hang around to work the station who is
QSOing the station you gave the frequency to.
You may work the same station multiple times provided they are separated
by at least 3 other QSOs in both logs (regardless of band). For example,
if WN4KKN works N6TR, KKN must work at least 3 other stations before he
can work TR again. TR must also work 3 stations before working KKN
again. Changing bands does not eliminate the three QSO requirement.
The three QSOs must not be dupes themselves.
You must not work the same station or stations using any kind of schedule
or system. It is the intent of the dupe rule to make sure we don't run out
of stations to work. It is NOT the intent of this rule for you to change how
you would operate the contest if dupes were not allowed. If, in the log
checkers opinion, you have not lived up to the intent of this rule, your
log will be disqualified!!
Total score is the number of contacts you make. Any QSO found to be
defective in anyway will be removed from both logs (yes, if someone
miscopies your exchange, you won't get credit for the QSO, so QRS a
bit!!).
Please refrain from using vulgar or inappropriate names. If you receive
one of these names, feel free to either edit it or replace it with your
starting name. Make sure to make a note in your log so we know what you did.
Injecting the contest with an inappropriate name (in the log checker's
opinion) will result in a 1000 point penalty per occurrence. Examples of
inappropriate names may be found on MTV and generally start with the
letter "B".
Additional penalties will be assessed to people who work a significant number
of QSOs, but don't turn a log in. They will be given minus one point
for each QSO that we can verify actually occurred.
Logs must be sent in ASCII format via internet to n6tr@contesting.com within
72 hours of the end of the contest. Figuring out how to send in your
log on the internet is PART OF THE CONTEST. If you need help, we will
try to assist the best we can.
Logs must show the band, time, station worked, number sent, number received,
name received and QTH received for each QSO. Also, please tell me the
name you start the contest with. We will assume the name you send is the
name received on your previous QSO, so you don't have to show that.
Results will be publised on CQ-CONTEST within 2 weeks of the contest. Logs
are checked using the K2MM LogZap software system. All checked logs will
be made available by FTP except for those requested by the submitter to
be kept private. Decisions of the judging committe are final and arbitrary.
Good luck, tell a friend and HAVE FUN!!
Tree N6TR
tree@contesting.com
>From Spike Lazar <slazar19@sgi.net> Fri Apr 5 19:32:26 1996
From: Spike Lazar <slazar19@sgi.net> (Spike Lazar)
Subject: Announcing HamVention Contest
Message-ID: <199604051932.OAA00375@orion.bv.sgi.net>
At 01:49 PM 4/5/96 -0500, you wrote:
>>
>> Info: For your pseudo checksum # contact AA8U.
>>
>I have found your posting at times humorous, but more often superfluous and
>annoying. However, I stand by your right to post them regardless. I
>interpret this latest posting as not so thinly veiled sarcasm. Like I said,
>I have very good duck feathers. Guess this one will require their use.
>73,
>ugly
>
Hello Ugly,
Please save your duck feathers, I assure you, the loss of one or two
little feathers over what Bafoofnik writes is not worth it.
Please do not waste too many brain cells on whether any sarcam was
directed your way, because at our age those cells won't regenerate!
I will be careful in the future on what I post. I certainly will
rectumfy any and all future postings!
Sincerely,
dr. Bafoofnik
|