Q:Homebrew Wadley Loop HF RX
PSK
Can somebody please point me to the principle of operation of
the Barlow-Wadley Loop Receivers like FRG-7 .
Also, I would like to hear about possible Homebrew approcahes
using this Wadley loop design .
Please dont autoreply, and Please reply post to NG.
Best Regards !
PSK
Jim
PSK wrote in message <36963D2B...@exit109.com>...
>Hello everyone,
>
>Can somebody please point me to the principle of operation of
>the Barlow-Wadley Loop Receivers like FRG-7 .
>
>Also, I would like to hear about possible Homebrew approcahes
>using this Wadley loop design .
>
is fed into a comb generator, which will generate lots of harmonics at 1 MHz
spacings and is then filtered with a 30 MHz low pass filter. There is also
a VFO. For this example lets say it runs from 70 to 100 MHz. Lets say
there is also a narrow filter at 70 MHz. [It is called 'comb generator'
because it looks like the teeth of a comb when you look at it with a
spectrum analyzer.]
The VFO is mixed with the comb generator and fed to the 70 MHz filter, which
filters out all but one of the harmonics. The output of the 70 MHz filter
is then mixed again with the same VFO signal to bring it back down to the
original frequency (with another 30 MHz low pass filter).
The result is, allegedly, a nice clean signal that can be switched in 1 MHz
steps by tuning the VFO.
It sounds neat, but there must be some catch or it would be used more often.
I can think of a couple of possible problems. One is that the 70 MHz
filter, in my example, would have to be pretty darn good to eliminate the
unwanted parts of the comb signal. The second is that any crud on the VFO
signal will show up on the output, destroying the nice clean crystal purity.
Don't give up hope yet. I think that it still has some use for the
homebrewer if you are aware of the potenial problems.
Jim
N8EE
nicholas charles caruso
> PSK wrote in message <36963D2B...@exit109.com>...
> >Hello everyone,
> >
> >Can somebody please point me to the principle of operation of
> >the Barlow-Wadley Loop Receivers like FRG-7 .
> >
> >Also, I would like to hear about possible Homebrew approcahes
> >using this Wadley loop design .
> >
> As an example, it starts out with a crystal oscillator, say at 1 MHz. This
> is fed into a comb generator, which will generate lots of harmonics at 1 MHz
> spacings and is then filtered with a 30 MHz low pass filter. There is also
> a VFO. For this example lets say it runs from 70 to 100 MHz. Lets say
> there is also a narrow filter at 70 MHz. [It is called 'comb generator'
> because it looks like the teeth of a comb when you look at it with a
> spectrum analyzer.]
This is a cool idea.
Can you describe the operating principle of a comb generator? I've always been
curious about this
but haven't seen an explanation anywhere. Perhaps it's that the harmonics all
have an equal amount of power in them, as opposed to falling off as in a
multiplier?
OK, I'm curious now. Anyone have a circuit for a comb generator online?
73
nick caruso KB1DEJ
Joe McElvenney
Hi,
This was of course the basis of the famous Racal RA17 receiver
that was made in England and sold in its tens of thousands in the
60/70s (a large number went to or were made in Canada I seem to
remember). There was a version with U.S. equivalent tubes/valves and
with a meter calibrated in S-points for use by the well-heeled ham.
Anyone who has seen one of these sets will know that it used a
very substantial diecast chassis with lots of compartments to keep
potential spurs at bay. Legend has it that Racal nearly gave up on it
until someone thought of hacksawing through selected parts of the
chassis to isolate and ultimately eliminate the problem.
Wadley-Barlow (or was it Barlow-Wadley) was a South African
company if I remember correctly, as I think was Wadley himself. One
of his most striking inventions was the microwave Tellurometer which
could measure distance to a fraction of a metre over several
kilometres.
73 de Joe, G3LLV
Muaskar Al Murtafa'a
Sultanate of Oman
BOB DUCKWORTH
Barlow-Wadley, and the FRG-7. If I recall, one selling point
for the BW and the FRG7 was that by using the same osc any
drift cancels in the two conversions and a mess of crystal references
could be eliminated.
-bob
c.c.t...@qxhomebrew.killspam.us.com
: Re: Wadley Loop
: This was of course the basis of the famous Racal RA17 receiver
: that was made in England and sold in its tens of thousands in the
: 60/70s (a large number went to or were made in Canada I seem to
: remember). There was a version with U.S. equivalent tubes/valves and
: with a meter calibrated in S-points for use by the well-heeled ham.
:
: Anyone who has seen one of these sets will know that it used a
: very substantial diecast chassis with lots of compartments to keep
: potential spurs at bay. Legend has it that Racal nearly gave up on it
: until someone thought of hacksawing through selected parts of the
: chassis to isolate and ultimately eliminate the problem.
This is true, Racal used to have the original on show (maybe they still
do). What's not so commonly known is that before Ray and Cal got hold of
it, the design was previously offered to Eddystone, who turned it down
thereby missing a lot of business and a bigger place in history.
Stephen
: Wadley-Barlow (or was it Barlow-Wadley) was a South African
:
c.c.t...@qxhomebrew.killspam.us.com
<A confusing and not very correct description of the Wadley loop>
The RX can be separated into two functional blocks;
The simple part is the back end which is a conventional tunable IF, in
other words a superhet tuning a fixed 1 MHz range. Tuned by the second
(KHz) VFO which needs to be high stability. There is no drift cancellation
for this VFO.
The front end converts the HF band down to the tunable IF range in 1 MHz
steps, controlled by the first (MHz) VFO. The first VFO runs above signal
frequency and is used in 2 conversions;
a) By itself as the local oscillator in an up conversion to a wide band (>
1MHz) high frequency IF (45 MHz in the RA 17) first IF.
b) Mixed with harmonics generated from a 1 MHz reference (the comb
generator) and filtered to provide the local oscillator for the second
conversion down the tunable IF range.
Mixing the first VFO up and down cancels any drift in this section of the
RX.
Here how it works;
Suppose the first IF is 44 to 45 MHz.
The first VFO will tune 46 to 74 MHz for a 1 to 30 MHz signal
range.
Suppose the second IF tunes 4 to 5 MHz
The local oscillator needed for the first to second coversion will be 40
Mhz.
This 40 MHz comes from the first VFO mixed with harmonics from a 1MHz
xtal and fed to a 40 MHz filter. This combination can only produce an
output at 40 MHz when the VFO is at some multiple of 1 MHz.
Eg VFO 60 MHz, - 20th harmonic of xtal = 40 MHz
VFO 61 MHz, - 21st harmonic of xtal = 40 MHz
If the VFO is at say 60.5 MHz there's no mixing product at 40 MHz and
therefore no injection to the second mixer. As you tune the MHz VFO you
can hear increases in noise as it passes thru each MHz point and the
second mixer gets a local oscillator signal.
To cover 14 to 15 MHz, the MHz VFO is set at 59 MHz, this mixes with 19th
harmonic of the xtal to generate the 40 MHz injection signal.
Suppose the signal is at 14.2 MHz, mixed with 59 MHz gives a 1st IF of
44.8, mixed again with 40 MHz gives the second IF of 4.8 MHz.
If the MHz VFO drifts to 59.05, the 1st IF will become 44.85, but the LO
injection will also change, from 40 to 40.05. 44.85 - 40.05 = 4.8. Bingo!
the second IF frequency is unchanged, the drift has been cancelled out.
Stephen
Michael Black
The HRO-500 used a PLL to get the injection to the first mixer, At first
glance it might appear similar to the Wadley loop because the phase
comparator took it's inputs from the VCO and a crystal oscillator with
high harmonic contents, in effect locking the VCO to a harmonic of the
crystal. And you did tune the VCO manually to get it to the next lock
point. But if the hardware had been up to it at the time (early sixties),
the synthesizer would have looked like what we are used to now, with a VCO
feeding a divider chain and then the phase comparator.
Radio Shack had a shortwave receiver that used the Wadley loop, though I
can't remember the model number.
And Barlow-Wadley themselves had a transistorized portable that used the
loop, in the early seventies. Undoubtedly it was alot cheaper than the
Racal, and so perhaps it made the loop more prominent. I'm still not sure
if it's the Wadley loop or the Barlow-Wadley loop, and I think perhaps
this confusion comes from the populariztion of the loop in that portable.
Michael VE2BVW
Graeme Zimmer
>I believe the National HRO500 employes the same scheme as RA-17,
>Barlow-Wadley, and the FRG-7.
And "Electronics Australia" magazine had a Wadley loop RX design
called the "Deltahet".
About 20 years ago........
Cheers ............................ Zim
Paul Keinanen
: As an example, it starts out with a crystal oscillator, say at 1 MHz. This
: is fed into a comb generator, which will generate lots of harmonics at 1 MHz
: spacings and is then filtered with a 30 MHz low pass filter. There is also
: a VFO. For this example lets say it runs from 70 to 100 MHz. Lets say
: there is also a narrow filter at 70 MHz. [It is called 'comb generator'
: because it looks like the teeth of a comb when you look at it with a
: spectrum analyzer.]
: The VFO is mixed with the comb generator and fed to the 70 MHz filter, which
: filters out all but one of the harmonics. The output of the 70 MHz filter
: is then mixed again with the same VFO signal to bring it back down to the
: original frequency (with another 30 MHz low pass filter).
: The result is, allegedly, a nice clean signal that can be switched in 1 MHz
: steps by tuning the VFO.
I have seen this wave analyzer type of circuit in some microwave
systems and also for selecting the desired harmonic of a pseudo-random
noise generators that produces a thick forest of discrete frequency
lines.
However, I do not think this is known as the Wadley loop described in
some other postings in this thread.
Since both the wave analyzer circuit and the Wadley loop require about
the same amount of hardware, why is the Wadley loop used instead of
the wave analyzer circuit in receivers ?
: It sounds neat, but there must be some catch or it would be used more often.
: I can think of a couple of possible problems. One is that the 70 MHz
: filter, in my example, would have to be pretty darn good to eliminate the
: unwanted parts of the comb signal.
The first unwanted signals are at 69 and 71 MHz, so fixed notches at
those frequencies should go a long way without using crystal filters
(expensive or not available for these frequencies in those days).
: The second is that any crud on the VFO
: signal will show up on the output, destroying the nice clean crystal purity.
The narrow filter between the mixers will produce a delay
(microseconds in this case) into the signal and if the VFO frequency
is rapidly changing (phase noise etc.), the signal is not remixed with
exactly the same frequency it was mixed before the filter, thus
producing a frequency error at the output. However, if the VFO signal
going to the second mixer could be delayed by the same amount as the
wanted signal through the filter, this error could be reduced.
Paul OH3LWR
abcd
It seems that cleaner and easier way to generate the second IF
conversion oscilator at 40 MHZ would be to use a analog translation
PLL-VCO for filtering and translating it, instead of the
mixer + sharp 40 MHZ filter .
U would still need the COMB generator ,unless,
U decide to replace that with another PLL.
1St Mixer 2ndMixer
--------------X1--------X2--------(4-5 Mhz Variable IF )----------
| |
VCO1 VCO2 <-LPF
| APLL |
| | |
---------- X3 ---P1 <<<< APLL Mixer
|
|
1Mhz Comb----------------------|
X1 = 1st Mixer
X2 = 2nd Mixer
X3 = Translation Mixer
P1 = APLL phase detector
LPF= Low Pass filter that effectively replaces the 40 Mhz BPF
Thanks again !
Niko G Shepherd
abcd wrote in message <36993731...@efgh.com>...
>Thanks for the very clear explanation of the Wadley Loop,
>It seems that cleaner and easier way to generate the second IF
>conversion oscilator at 40 MHZ would be to use a analog translation
>PLL-VCO for filtering and translating it, instead of the
>mixer + sharp 40 MHZ filter .
True today, but remember it was devised almost 50 years ago.
Stephen
Alan Peake
><A confusing and not very correct description of the Wadley loop>
>Stephen
That's basically what's in my Drake SSR-1 receiver. The mfrs blurb is a
classic case of marketing however. The 1MHz comb generator put out harmonics
at all multiples of 1 MHz which of course, get back into the front end. The
designers apparently couldn't get rid of them so marketing proudly announce
that they have included a 1MHz marker signal at no extra cost!!!
Despite that, it's not a bad little HF rx.
Alan
Jim
nicholas charles caruso wrote in message <36965965...@certco.com>...
>This is a cool idea.
>
>Can you describe the operating principle of a comb generator? I've always
been
>curious about this
>but haven't seen an explanation anywhere. Perhaps it's that the harmonics
all
>have an equal amount of power in them, as opposed to falling off as in a
>multiplier?
It is partly physics and partly careful design.
If you have a 'square' pulse you will have lots of harmonics and the
amplitude of the harmonics will fall off as sin(x)/x, with x being
proportional to the width of the square wave.
When you make the pulse very narrow the harmonics will be relatively flat to
a higher frequency. You can never get them completely flat for all
frequencies (unless you use zero pulse width!), but you can get them 'good
enough' over a limited frequency range, particularly if you use some type of
filtering (the 'careful design' part).
Jim
dennis W. Ostrowski
> It is partly physics and partly careful design.
>
> If you have a 'square' pulse you will have lots of harmonics and the
> amplitude of the harmonics will fall off as sin(x)/x, with x being
> proportional to the width of the square wave.
>
> When you make the pulse very narrow the harmonics will be relatively flat to
> a higher frequency. You can never get them completely flat for all
> frequencies (unless you use zero pulse width!), but you can get them 'good
> enough' over a limited frequency range, particularly if you use some type of
> filtering (the 'careful design' part).
>
> Jim
Using a square wave, wouldn't you only get the ODD harmonics?
Dennis
#12
Jim
dennis W. Ostrowski wrote in message <369AA5F9...@lucent.com>...
>
> Using a square wave, wouldn't you only get the ODD harmonics?
>
>Dennis
>
>#12
That is true, you will get only odd harmonics with a _square_ wave, but if
you make the pulse width smaller you no longer have a square wave and you
will (at zero pulse width) get all of the harmonics.
Jim
Jim
PSK wrote in message <36963D2B...@exit109.com>...
There is another method that I think may give better results, and is a much
simpler circuit.
Start out with a NN MHz oscillator (NN being a number).
Have a variable frequency oscillator covering the range that you are
interested in. This oscillator can be controlled by a mechanically variable
capacitor or by a voltage controlled capacitor. In addition you will need
another voltage controlled capacitor that has a limited tuning range,
perhaps tens of KHz.
Take the output of the variable frequency oscillator and feed it into a D
flip flop (sorry, but can't remember which input---my books on this are at
home). Feed the NN MHz into the other input of the D flip flop.
Take the Q output of the D flip flop and feed it into a phase detector (as
in a phase lock loop). The reference signal for the phase detector will be
NN/NN, in other words 1 MHz.
As you tune the main tuning capacitor in the variable oscillator it will go
in and out of lock every 1 MHz.
As someone else in this thread pointed out, the Wadley loop was/is used as a
'block converter' or 'band select'. You can either convert a block of
frequencies and have a tunable IF, or mix the output of the Wadley loop with
a VFO and have a fixed tuned IF. The circuit I just described can be used
the same way.
Jim
Michael Black
On Thu, 14 Jan 1999, Jim wrote:
>
> PSK wrote in message <36963D2B...@exit109.com>...
>
>
> There is another method that I think may give better results, and is a much
> simpler circuit.
>
> Start out with a NN MHz oscillator (NN being a number).
>
> Have a variable frequency oscillator covering the range that you are
> interested in. This oscillator can be controlled by a mechanically variable
> capacitor or by a voltage controlled capacitor. In addition you will need
> another voltage controlled capacitor that has a limited tuning range,
> perhaps tens of KHz.
>
> Take the output of the variable frequency oscillator and feed it into a D
> flip flop (sorry, but can't remember which input---my books on this are at
> home). Feed the NN MHz into the other input of the D flip flop.
>
> Take the Q output of the D flip flop and feed it into a phase detector (as
> in a phase lock loop). The reference signal for the phase detector will be
> NN/NN, in other words 1 MHz.
>
> As you tune the main tuning capacitor in the variable oscillator it will go
> in and out of lock every 1 MHz.
>[stuff deleted]
You'd see the idea of using a type D flip flop as a mixer years ago, but
for some reason you don't see much mention of it in recent years. I can
picture articles that used them, but I'd have to do some searching to dig
them out and so I can't provide info on which pin is fed from what source.
But I think you've complicated something unnecessarily. There was a time
when you'd see PLL's for generating signals every 500kHz or 1Mhz for the
first oscillator in a receiver. I even mentioned one in this thread, when
I mentioned that such a scheme was used in the HRO-500. And you'd see
articles for building such a PLL in the ham magazines.
But they locked to a harmonic of the reference oscillator, without the
need for a mixer between the VCO and the phase detector. You'd tune the
oscillator manually till it was close to lock (often indicated by a beat
note in an audio amplfier fed from the output of the phase comparator),
and then the phase detector would take over. The oscillator would feed
the phase comparator, and the reference oscillator (with plenty of
harmonics) would feed the other input. Remember, the simplest phase
detectors are harmonic sensitive.
Some of the construction articles mentioned that it was harder to lock the
loop the higher the harmonic you were locking to, and also that the lower
the reference point the more trouble.
There is nothing particularly different between this and having the VCO go
through a programmable divider. One uses harmonics, the other uses
division. Functionally, I think there are trade-offs to either scheme.
If you lock to a harmonic, you'll need a dial for the oscillator tuning,
and you will be going in and out of lock as you tune through the harmonic
points. But since you are manually tuning the oscillator, the voltage
variable capacitor does not need to tune very much.
If you use a divider, the VCO should always be in lock, and the number of
the division will always tell you the correct frequency. But you'll have
to have a VCO that tunes the whole range (though you could use
pre-steering off the programmable divider input).
I think the locking to a harmonic of the reference was primarily used
because programmable dividers weren't available or were still hard to get.
The scheme was used in the HRO-500 and the Galaxy R-530 receiver, perhaps
others. One construction article that comes to mind was in QST for
January 1972 (with an excerpt in the ARRL Handbook for a nubmer of years).
The author had an updated design in Ham Radio about 1978 or so.
Michael VE2BVW
Jim
Forgot to mention....
The variable oscillator will have to tune from NN MHz upward, and will then
lock at every 1 MHz frequency starting at NN MHz.
Jim
dennis W. Ostrowski
Sounds cool!
I had already done something similar to run very low power FM radio
commercials on a lot of different channels simultaneously, like house
for sale, open house, etc...where someone driving by would hear it and
maybe stop in, but I couldn't cover every frequency, just every other
one!
Thanks again,
Dennis
#12