[HBR] The long, SLOW HBR project

[donald haworth]

Thanks Walt!  Lots of information to digest.
The receiver I am currently building is based on the R392.  It uses 26a6's
for the IF, and 26FZ6's for the RF stage.  The detector is solid state,
however, using FET's-low distortion-very good audio...
I built the thing in 'modular' form so I could make changes as I went
along.  My inductors are all toroidal-silver plated wire-highest Q I could
manage-bandswitched,  I built the LO and mixer stage separately.  The LO has
a digital frequency readout (N3ZI kit), so tracking is not an issue..dial in
the wanted LO frequency, and peak the RF.....one more knob to twist, but the
results are more than worth it!  BTW, this is a single conversion radio, IF
at 455.  The RF stages tune sharp enough (high Q toroids-cores matched to
tuning range-per Amidon's charts) that images and spurs can't be found...of
course you can TUNE to an image, so good dial calibration is a must for the
RF stage.  The RF stages are agc controlled from a separate agc detector, as
well as adjustable bias to set the threshold.  With adjustable gain on the
IF, and a gain control pot for the RF, I can find the best mix for signal to
noise---REALLY makes a difference!  And proves your point about stage gain.
As to the 26FZ6, this was a tube made for and used in the R392-only!  Hard
to find, but what a performer!  Extremely sensitive, yet not easy to
overload.....and I live about 15 miles from WWCR--also WSM drops  .780
voltp/p at the output of my tuned loop antenna, so intermod is always a
problem, but with the 26FZ6, all is as it should be.  Really a remarkable
tube.
The IF strip and agc is lifted from the R392 manual....all SIX stages!
Bandwidth is about 7kcs-best I could do with the IF transformers I had-but
the audio quality is great.
--Somehow it all works- tuning is from 520kc to about 16mc--the RF coils are
bandswitched and the LO coils are bandswitched, separately (yes, yet another
knob!)
And the highest voltage is only 28vdc.  The LO is very stable. at 15 mc (osc
at 15.455), 1kc drift from cold turn on to 30 min-then it will sit there all
day long-  BTW the LO/mixer is the same circuit as the LO in the R392-a
26d6.  Looks like a 6be6.......
--Short comings of the above:  You need the skill of a safe-cracker to tune
the RF stage.
--The thing weighs in at 40 pounds-and counting...and, as mentioned, lots of
knobs to twist!
--So, with it all working....it's time to 'make it work better', or fix it
tell its' broke.....
--Rebuilding the RF stage.  Changing to a four gang cap 10-150pf.  This to
spread-out the tuning rate...also means rewinding the coils w/ more
inductance......change the turns ratio of the coils.....put the forth LC
stage at the front of the RF amp (it will now have 2-tuned circuits into the
grid)..........and that's were I'm at right now.......
--As you can see, I look forward to your posts.  Nice to know I'm not the
only one out there!
--One of the stranger things I ran into was a loud hash at about
780...wouldtake out WBBM as well as a local on 790.......Did the transistor
radio-walk around the house-thing...turned circuit breakers
on/off.............turned out to be the power supply for the receiver
itself...the bridge rectifier was outputing quite a signal...my power
transformer is a 200va toroid---so all the chopped noise was getting right
back into the 120 v side....my radio was picking it up as a real
signal...not from the PS itself.....four .01's across the bridge made it all
go away!
-This has been and continues to be a great learning experience--
My apologies for the long post!
Don in Nashville
On Tue, Sep 20, 2011 at 2:18 PM, Walt Hutchens <waltah at earthlink.net> wrote:

> Donald Haworth posted:
>
> > Thanks for the update, Walt.....Kinda reminds me of an old Republic
> Pictures
> > serial...."next week....Chapter ten..."The Deadly 10 Meter
> > Drift..........",  BTW, I am a HUGH fan of Republic Pictures......
>
> Half hour, black and white ... matinee admission only a quarter.   Ran
> before the feature?   I forget -- that was a LONG time ago!
>
> > Anyway, how about some pics of your radio?  Would love to see it!  I feel
> > like its an old friend, after following its' progress.
>
> I will try to get pics up this week.
>
> > -I have a question for you-or anyone else, for that matter.  Given a
> typical
> > rf amp-6ba6-what would be the best l/c ratio for the (input ) tuned
> > circuit?  For example, an inductance of -say 280uh with a capacitance of
> > 365...versus a cap 0f 150pf and a much higher L of 650.....both will
> > resonate at the wanted frequency...but which combination will give the
> best
> > transfer of energy to the tube?
>
> Almost no energy is transferred to the tube.   A bit is dissipated in the
> grid circuit including the tank circuit itself.
>
> The grid VOLTAGE resulting from a specific signal depends on the Q of the
> grid tank circuit and with modern materials and techniques, it's pretty
> easy
> to get as much as is useful.   The higher L tank can deliver higher
> voltage,
> but that's only needed in specialized cases.
>
> The problem in HF communications receivers is rarely one of having enough
> gain to hear a weak signal.   It is instead being able to pick that signal
> out of the noise.
>
> 1. Atmospheric noise -- worst on the low bands, most of it is gone by the
> time you get to 10M.
>
> 2. Other signals of all sorts.   This includes all the trash generated in a
> modern household by computers and computer chips in household devices and
> carried around the neighborhood on power and phone lines.
>
> 3. Noise generated in the radio itself.   This includes both the noise
> caused by electrons moving and noise resulting from signals (including
> unwanted signals) and external noise combining because the radio isn't
> perfectly linear.
>
> Assuming you have a decent outside antenna with a proper feed line so the
> in-house noise sources aren't a big deal then up through 40M the problem
> for
> weak signals and uncrowded band conditions will be atmospheric noise.   The
> best you can do is have the passband no wider than the bandwidth of the
> desired signal -- say 6 kcs for AM, 2.8 kcs or so for SSB, and as little as
> a few tens of CPS for CW.
>
> When these bands are crowded with strong signals (say evenings on 80 and
> 40)
> then the unwanted signals replace atmospheric noise as the main issue.
> Again, however, an appropriately narrow passband is the best you can do.
>
> Now things get more complicated.   The simple picture is you can cut your
> passband down (with a filter -- sharp IFs, a crystal filter, etc.) at any
> convenient stage.   HOWEVER the lower bands generally supply so many very
> strong signals -- I once lived where I could measure 0.5 VOLTS at the
> connector on my antenna! -- that when the gain you need to copy (say) 1 uV
> is applied, one or more stages in your receiver can be overloaded to the
> point of generating new unwanted signals.
>
> (If you have to amplify the 1 uV signal to 1 V at the speaker, that's a
> voltage gain of a million.   Applying the same gain to the broadband signal
> from the antenna mentioned above would deliver 500,000 volts at the speaker
> terminals ... talk about overload!)
>
> Think of how an overdriven linear amp sounds.   That's the wanted signal --
> the guy's voice -- plus a bunch of trash generated in his flat-topping
> amplifier.   If you want to hear signals clearly, your receiver must be a
> series of properly adjusted linear amplifiers, all the way from the antenna
> jack to the speaker or headphones!
>
> In the receiver situation, however, the overdriving signal can be anything
> that can get to the overloaded stage.   A poor quality receiver might be
> hitting the 1st IF amp with a 100 kcs bite of 40M at night.   If you are
> operating such a receiver in the upper part of the band here in Virginia,
> World Christian Radio with (is it?) 100 kw in the Nashville area beaming
> toward Europe is part of what is reaching that 1st IF and that stage IS
> being overloaded.   Similar conditions can occur on 80M due to the great
> number of phone stations in the upper part of the band and also on 160 if
> you have nearby BC stations in the upper end of that band.
>
> (You might think that AGC would help with this, but if the unwanted signals
> are filtered out anywhere before the AGC detector, then World Christian
> Radio may be invisible to the AGC.   Only if the IN-passband signals are
> strong enough to kick up the AGC does it help at all with the
> OUT-of-passband ones.   This is where manual gain (or AGC threshold)
> controls come in handy.
>
> Overloaded stages -- often the 1st IF but can also be the first or second
> mixer or even the RF stage -- can combine two signals that you think are
> out
> of passband to create new IN passband signals that will ride right along
> with the station you're trying to hear.   Once this happens no filter can
> separate the trash from the wanted signal.
>
> For example strong stations at 3820 and 3840 that aren't filtered out
> before
> they get so strong that they overload a stage will give you trash signals
> at
> 3800 and 3860.   If you are listening at either frequency, you will have ON
> FREQUENCY interference.
>
> The ideal way to beat this would be to put the sharp filter at the antenna
> but that can't be done since the combination of 'very sharp' and 'tunable'
> is not practical.
>
> The practical solution has four parts:
>
> 1. Don't couple more signal than needed from the antenna.   Once you have
> coupled enough signal to the first stage (usually the RF amp) to hide its
> internal noise, you don't want any more.   Some receivers fix the coupling
> at a fairly low level, others provide variable coupling or an attenuator of
> some kind.
>
> Coupling the maximum signal you possibly can from the antenna to the first
> receiver stage would only be a good idea with a poor quality receiver --
> one
> that doesn't have enough gain and thus wouldn't hear much of anything,
> otherwise.   (Some of those designs with a 6SA7 or 6BE6 mixer as the first
> stage might need all the sig they can get.)
>
> 2.  Until you get past the sharp filter don't use more amplification than
> needed to hide the noise of the next stage.   Use too much amplification
> and
> that next stage may be overdriven by unwanted signals that you haven't yet
> filtered out.   That's how 3820 + 3840 -> 3800 & 3860.
>
> Ever notice that top quality receivers are easier to listen to under
> thunderstorm conditions?   If your front end isn't linear, then 3810 signal
> + crash energy at 3820 -> noise at 3800 and 3830.   3820 signal + crash
> energy at 3840 -> noise at 3800 and 3860 ... and so on, for the entire band
> that can get to the overdriven stage.   A cheap receiver is getting more of
> the thunderstorm energy on the frequency you're trying to copy via this
> 'reciprocal mixing' than a quality one.
>
> (Worse yet, crash energy at 3820 and 3840 yields crash signals at 3800 and
> 3860 ... and this is true for EVERY pair of frequencies ... 3805 and 3810,
> 3806 and 3812 ... t'storm noise conditions are an excellent test of front
> end bandwidth and linearity!)
>
> Some very good receivers with low noise 1st mixers use no RF amplifier at
> all.   (The SS 1R with its beam tube mixer is one example but some triode
> mixers can be used in the same way.)   The noise of a quiet mixer will be
> much below the atmospheric noise on all the HF bands.   Other receivers use
> an RF stage with a gain of less than one:  Its only job is to serve as an
> AGC controlled attenuator to help protect the mixer from overload.
>
> Part of 'not too much amplification' is a combination of AGC and manual
> gain
> control that will let you reduce the amplification when you're copying a
> moderate to strong signal in the presence of extreme noise or unwanted
> signals.
>
> 3.  Put as much selectivity as possible, as close to the antenna as
> possible.   It's easier to build a good receiver with single conversion and
> a crystal filter at 1700 kcs in the mixer plate circuit than to get equal
> performance with dual conversion and sharp filtering in the 2nd IF (100 kcs
> or 85 kcs ...) because the dual conversion set has at least one and often
> two more stages ahead of the filtering.
>
> AND use the highest Q coils you can in the RF and Mixer grid circuits.
> High Q here may only cut the bandwidth seen at the mixer from 500 kcs to
> 150
> kcs, but every bit helps!
>
> Here we come to the answer to the original question:  Other things being
> equal the 150 pf / 650 uH combination will be the better choice because
> it's
> likely to do a better job limiting the bandwidth at the RF amplifier.
> That
> is IF you use loose enough coupling to the antenna that you get only enough
> signal to be sure of overcoming receiver noise in the RF amp.
>
> Just as for a transmitter, the antenna coil in a receiver is loaded by the
> antenna.   Take a coil with a Q of 200, couple it tightly to a resonant
> antenna, and you're prob'ly getting an effective Q around 10-30.   Yes,
> with
> the much greater bandwidth that the lower Q implies.   In receivers that do
> this, the details of the antenna coil hardly matter -- you could about wind
> it with nichrome wire on a wet toilet paper tube and get equal performance.
>
> To minimize this effect disconnect the antenna and adjust your receiver so
> you can hear the internal noise of the RF amp.  (Confirm that the noise you
> hear comes from the RF amp by detuning the mixer coil; if the receiver
> doesn't go quiet you have another problem!)  Then connect the signal source
> and set the antenna coupling to the point where the weakest signals just
> hide the receiver noise.   This is most easily done with a good signal
> generator but with patience you can do okay with an antenna.
>
> On HBR-type designs this is easy to do by tweaking the spacing between the
> antenna winding and the grid winding on the antenna coil.
>
> 4.  Particularly in the RF, mixer, and 1st IF sockets use tubes that are
> good at handling large signals.   The 6EH7 is about the best RF/IF tube out
> there, although the 6BA6 is pretty good.   However this isn't as simple as
> swapping a better tube for a worse one:  All the AGC controlled stages need
> to have similar AGC characteristics, otherwise one of them will be cut off
> while the others are still amplifying and this can lead to serious
> distortion.   Generally speaking this means that all AGC controlled tubes
> --
> usually the RF and IF amps -- should be the same.
>
> Walt
> KJ4KV
>
>
>
>
>
>
>
>
>
> ______________________________________________________________
> HBR mailing list
> Home: http://mailman.qth.net/mailman/listinfo/hbr
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:HBR at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>