[HBR] GC-HBR Project -- Was 'Navy guy ...'

[Walt Hutchens]

I said:
> ... I have yet to hear anything about why it won't work!    
> LOL ...  

Jim's excu ... I mean, reason:
> Two words: Field Day. 
> 
> FD #1 for me was 1967 IIRC and I haven't missed one since.

This is actually very cool.  I have occasionally worked the Electric 
Radio vintage field day but the last couple of years have been so 
busy I can barely keep up with the traffic on this list, let alone 
participate in an event.

> > Of course the fact is that they don't all work -- the batting average 
> > is around 50% I think -- not even that good if you count the paper 
> > design.   So your comments are always appreciated.
> > 
> 
> Thanks!  (where's my pencil?)
> 
> DEVIL'S ADVOCATE MODE = ON

SHOOT -- you mean yours has a switch?   I wondered what was 
supposed to be where that snap hole plug is.   The front panel on 
mine is stencilled "MFP June 1944" which I understand and "Auto 
operation only" which I didn't figure out until now.  

Ah well -- on to the serious stuff.  

> The plan is a bandswitch with 9 positions, used in three 
> groups.   Each group of three selects the same front end tuning 
> range, roughly 1.75-5.5-15-30.5 Mcs.   The mixer input and 
> synthesizer coils are switched in those ranges and must be 
> manually tuned.   

Jim comments:
> Interesting! But such a switch is going to have a *lot* of sections! 

Well, not THAT many -- not more than required for a typical ham 
transceiver of the time period.   'Parts only' FT-101's are pretty 
cheap when you look at all the stuff that's in there.   I haven't 
actually done the counting but I'm pretty sure the number of switch 
sections is managable -- something like 8, I think, counting 
switching of the crystals.   

> And getting the isolation between sections isn't easy. 

This is an issue all right, and I haven't worked out all the details 
yet.   But a word about the layout of the receiver is in order.

The 1st mixer and calibrator will be in their own 4" x 5" box at the 
left rear corner under the main chassis.   Everything will be in there 
-- tuning cap, bandswitch sections, tube sockets, coils, and the 
antenna connector. All signal grounds can go to the frame of the 
cap.   The tuning cap mounts to the front of the box rather than the 
chassis and since input is push-pull to the mixer grids (no RF 
stage) there's pretty good isolation from chassis currents.   Power 
goes in via feed through caps.

The synthesizer is directly in front of the mixer box.  It's an open 
layout.   Bandswitch, sockets for six internal and one external 
crystal, synthesizer tuning cap and coils, the two crystal 
oscillators, multipliers, and mixer are all there.   Syn. signal flow 
front to rear, starting with crystals near the panel and ending with 
the syn. mixer right in front of the front end box.   The syn. output 
signal goes into the box via a couple of turn link on the syn. mixer 
plate coil.   Drive push pull to the signal mixer cathodes.   

The two bandswitches share a shaft but there should be plenty of 
room for an insulated coupling to prevent the shaft taking 
synthesizer trash into the front end.   That will require using two 
bushings to support the shaft -- one on the panel and one on the 
front of the front end box.     

To the right of the synthesizer -- basically on the centerline of the 
chassis -- is a second box containing the tunable LO and 2nd 
mixer.   This gets the output of the first mixer via a coaxial cable 
between the boxes.   The 2nd mixer is similar to the first -- signal 
goes to the grids in push-pull, LO input is p-p on the cathodes.  I 
was pleased and a bit amazed to discover (with the R8040A 
design) that you can run a high-C oscillator right into a mixer 
cathode without a buffer AND without a trace of anything bad.   You 
can even apply AGC to the mixer.   It'll be a while before I do an LO 
for higher than 2 Mcs, there's such an advantage to being able to 
truly swamp the tube capacitances.  

Output from the 2nd mixer is again in parallel from the plates of the 
mixer tube; it goes over the shield box via a crystal ladder filter 
located on top of the chassis to the IF stages which run front to 
rear on the right side of the chassis.   The LO tuning cap (main 
tuning) is mounted on the chassis and again, all grounds can go to 
the capacitor frame.   That's important for this one, because it 
develops some serious power and chassis currents WILL lead to 
coupling to other places.   But again, the balanced design helps -- 
only the imbalance between the two halves leads to a current in the 
capacitor rotor contact and frame.   There is some 2nd harmonic 
current there.

The rear central part of the chassis contains the detectors, the two 
audio stages, and two 17H3 rectifier tubes.   Since I need two 
rectifiers anyhow I may use two completely separate HV supplies, 
one for the oscillators and other possibly voltage sensitive loads 
and the other for the IF stages and audio output that tend to make 
the voltage vary.   There's plenty of room for a couple of command 
set 3H chokes and associated filter caps.

Separate HV supplies is probably an unnecessary refinement.  The 
LO frequency variation in the R-8040A from S0 to meter pegged is 
barely detectable with the ear, so not enough to cause distortion of 
SSB signals.  The GC-HBR will have even higher C due to narrower 
tuning range and slightly lower operating frequency.

Enough of that.  There are large unwanted signals in that chassis 
and pretty good defenses (layout, circuit, and tuned circuit) against 
most of them.   Some will be detectable and possibly worse.  Let's 
consider the details.

> The problem I see is that the multiplier/mixer setup can generate
> a *lot* of spurious injection signals for the first mixer. Weak
> harmonics of the multiplied LO mix with the 50.7 and all sorts of
> hilarty ensues. 

> This may not seem a bother on paper ...

But of course in reality, it WOULD be a bother, size to be 
determined.  Let's try to see how big. 

The strongest spurious signals generated will be the harmonics of 
the MF crystal oscillator -- 5.825-8.675 Mcs, whichever crystal is in 
use for the band.   Since the bands are only 250 kcs wide but the 
crystals are all higher than 5.8 Mcs few bands will contain one of 
these harmonics but the exact sequence of harmonics depends on 
the band in use.   

The 2nd harmonic is above the band in use until 14.00-14.250 Mcs 
where the 7.050 crystal gives you a harmonic at 14.100 Mcs.  
That's relatively ugly and it cannot be moved out of the band with a 
simple jiggering of the numbers without either ending the receiver's 
coverage below 30 Mcs or getting into other kinds of trouble.   

For example dropping another two crystals to start with 5875 kcs 
(Channel 278) thus ending the receiver's coverage at 30 Mcs would 
use the 7100 crystal for the band 14.000-14.250 and place this 
signal at 14.200.  Going the other way with the crystals you could 
solve this problem but at the price of putting crystal fundamentals 
in the high (dial) end of EVERY band because those frequences 
overlap the tunable IF -- 5.550-5.800 Mcs. 

The spurious signal at 14.100 Mcs has to get through three tuned 
circuits to reach the mixer, but the 7.050 fundamental will be a volt 
or more so I think this one will be heard.   A trap will help, though it 
might require another bandswitch wafer.

(The band crystal frequency formula is: (lower band edge + 5.8 
Mcs + 50.7 Mcs) / 10  )

Above the 20M band the 2nd harmonic is below the tuning range for 
the band.   The 3rd harmonic isn't a problem below the band 24.00-
24.25 Mcs.   This band uses an 8.050 Mcs crystal and the 3rd 
harmonic is 24.15 Mcs.  The 4th and higher harmonics are above 
the receiver's range of coverage.  

No crystal frequency or harmonic falls within the tunable IF, 5.550-
5.800 Mcs -- we just went through fixing that design glitch.  

Now we get to the 10X(crystal frequency)-50.7 Mcs mixer.   The 
50.7 Mcs signal is from an overtone crystal at that frequency so it's 
pure except for multiples of that -- 101.40, etc.   The other signal -- 
58.250 to 86.750 Mcs by 250 kcs steps -- contains all the 
harmonics of whatever is the chosen crystal frequency, with the 
desired harmonic probably 20 db or so above the strongest of the 
others.  Thus for the lowest band (1.75-2.00) we intend 58.250-50.7 
= 7.550 Mcs but we will also produce significant signals at 7.550+/-
 5.825, namely at 13.375 Mcs and 1.725 Mcs.   Whew ... dodged 
THAT bullet, didn't we?  25 kcs below the end of the band is pretty 
close. 

On the next band up (2.0-2.25 Mcs) the frequencies are 7.800 
desired and that +/- 5.850 = 1.95 and 13.65 Mcs, unwanted.  For 
2.25-2.50, spurious frequencies are 8.050+/- 5.875 = 13.925 and 
2.175 Mcs. It doesn't appear that these spurious signals will be an 
issue, as they flank the band all the way up.  Pure luck ...

Yet another sequence of unwanted signals is produced by 101.4 
Mcs (second harmonic of 50.7) - 10X crystal frequency.   For 
example at 29.75-30.0 Mcs, that would be 101.4-86.25 = 15.15 
Mcs.  That one is way out of the band, but we dodge another bullet 
at 22.25-22.50 where the same calculation leads to a spur at 22.65 
and 22.50-22.75 where it's 22.4 Mcs.  In this case the spurious 
signal jumps right past the band in question and will not be an 
issue.

There are yet other beats that could possibly matter:  101.4 - 9X 
and 11X crystal frequency for example, and all the other such 
multiples.  These are of course much weaker than the 10X one.  I'd 
have to guess that some of these will fall in a working band but 
whether they'll be audible or not I can't say.  All of these unwanted 
signals can be reduced by cleaning up the 50.7. Mcs signal with a 
second tuned circuit or 101.4 Mcs trap; since this is a fixed 
frequency that's not a big problem.   

Harmonics of the LO at 1.354-1.704 are a potential source of 
spurs.  They are the reason both the LO and front end are in 
boxes.  About the only way for these signals to be heard is via 
coupling from the LO/2nd mixer box back to the 1st mixer box 
along the coax between the two.   Careful tweeking of the oscillator 
circuit to minimize harmonic production can make 10's of dB 
difference in the strength of these signals.   If audible at all, I 
believe these will be very weak, they tune backward, the higher 
harmonics are going fast, and they don't fall in any of the real ham 
bands.

Wait -- there's even more!  These were just the spurious SIGNALS 
coming from the synthesizer.  But spurious frequencies that are 
not heard can hetrodyne with strong wanted signals from the 
antenna to produce yet another whole set of unwanted responses.   
A receiver that is silent when tuned end to end with the antenna 
shorted may be very far from 'clean' when the antenna pours in 100 
mV signals from 41M religious broadcast stations.   

(Gotta love 'World Harvest Radio.'   At least if you're serious about 
testing receivers ...  I swear there are winter nights when you could 
run a lightbulb from your feedline by cutting an antenna for that 
one.)

The worst offender here is likely to be unwanted multiples of the 
band crystal reaching the 1st mixer and the defense is better 
purification of that signal.   But that's tougher than it may seem 
when 2x, 3x, 4x (band crystal) is close to the synthesizer output 
frequency, 7.550-35.550 Mcs.  

For example on the band 6.75-7.00, the synthesizer output is 
12.55 Mcs and the crystal frequency is 6.325 Mcs with a second 
harmonic at 12.65 Mcs.   On 7.00-7.25, the synthesizer output is 
12.80 Mcs and the crystal is 6.350 Mcs, second harmonic 12.70 
Mcs.  Now that is UGLY.  Rejecting a frequency that's <1% from 
the wanted frequency isn't going to happen in one or two practical 
tuned circuits.  It's especially ugly because the band is 41M -- 
home of some of the strongest unwanted signals ever heard on the 
planet.   

The unwanted second harmonic has to be killed before it reaches 
the synthesizer mixer output because it cannot be eliminated 
there.  Well ... Uh, I was going to use a balanced mixer for the 
synthesizer anyway.   Yeah, balanced mixer.   And probably an 
extra tuned circuit ahead of that mixer to hold down the unwanted 
harmonics.  Yeah, I was going to do that.

I'll consider the design of those circuits specifically to minimize 
second harmonic generation.  A trap might also have a place in the 
scheme.   

> Big coffee can for the xtals. 
Actually, that's a drawer under the receiver in the usual plywood 
cabinet.   The cabinet will differ only in detail from the 'One month 
HBR' that used the command receiver plug in tuning units and 
stored the spares in a drawer under.

I have the crystals in original military cases with felt blocks 
including a junk case that I don't mind sacrificing.  I haven't worked 
out the details but I plan to transfer the felt blocks to the drawer.   
120 = 10 x 12, so that should work out okay.

> Also you have the problem that when you are tuning a range right
> next to or on top of the tunable IF, the feedthrough can be a real
> headache. 
Not necessarily when the mixer input is push-pull and the IF is 
single ended.   I have that scheme in the R8040A and it's perfectly 
stable and quiet when you tune the front end through the IF. The 
GC project will have somewhat more gain:  It is possible that 
there'll be instability at or near the 1st IF, but that's a very common 
failing of somewhat simplified designs.  Even the S-line has a 
forbidden zone, does it not?  The frequences 5.550-5.800 (tuning in 
the reverse direction) aren't that big a deal in any case.

> Seems to me that what you want is a general-coverage (160 through
> 10 with all the spaces in between) rx that will make best use of
> that barrel of FT243s in the 4-9 Mc range. 

> The simplest way to do this I know of is to build a basic receiver
> and put a converter in front of it. Say we built a basic 1.75 to
> 2.25 receiver with IF in the 455 range (actually more like 465, to
> use some of those bazillion FT-241s). Such a receiver could use
> the tuning gang from a BC-454, so we'd have a slow tuning rate
> (10-12 kHz per turn or so), integral dial and ganged sections for
> tracking. 

Followed by a design for a receiver using harmonics of a smaller 
number of crystals, mostly from the same set but WITHOUT 
mixing with another crystal frequency, covering that range in 500 
kcs segments.  

This would have to be run through the same wringer as above and 
this is already a very long post.  The 500 kcs bands could lead to 
problems.  One of the main reasons I chose 250 kcs bands was 
that I found it much harder to avoid in-band spurious frequencies 
with the wider range.   However ethis is a differnt set of numbers 
and the details would be different.

Time again to nod in the direction of the W6TC designs:  His 
genius was in popularizing a design that delivered good usable 
performance while being home-buildable.   It has plug-in coils and 
drift is an issue on the higher bands, but any ham with modest 
construction skills could build it and beat the selectivity, crossmod 
performance, and sensitivity of anything he could buy for the same 
money six ways from Sunday.  In other words, he delivered a good 
COMPROMISE design.   

That's also my goal with the GC-HBR, but for the somewhat 
different function of 'station general coverage receiver.'  No 
compromise very quickly takes you out of 'you can build it' territory. 
 The very first thing that would happen is you'd need specialized 
tuning caps -- four gangs instead of two in the front end, six rather 
than four for the LO, maybe more than three for the synthesizer.   I 
have no idea where these parts would come from.  

IF this paper set ever becomes solder and aluminum and IF it 
works, then I hope others will improve on it.  But the designs I've 
found most successful (both mine and those of others) were also 
those that were the frankest about their own limitations.   This one 
will doubtless have a few spurious internally generated signals and 
be less than 100% bulletproof against spurious responses to the 
very strongest unwanted signals.

Thanks Jim!

Walt
KJ4KV