Crystal controlled super regen
Bill Higdon
have been replaced in most new designs. But I have a idea that would be
easier to do with a super regen as opposed to the new studd.
Thanks,
bILL hIGDON
Michael Black
a circuit for a crystal controlled superregen receiver. The same column
had that then novel transmitter in an IC, the LP-2000.
It was simply a cross-coupled multivibrator made up of two transistors,
running at 20KHz, with a tank circuit with a crystal added in. He said
the one he built at 10MHz worked fine. He pulled the circuit out of
an issue of EDN, but no date is given.
Charles Kitchin had an article in the May 1997 issue of QST about a
crystal-controlled regen receiver (for receiving 80meter W1AW code practice
runs).
In the Q&A column in CQ for May 1960, there is a circuit for a crystal
controlled Q-Multiplier, which is basically a pierce oscillator. Apparently,
the circuit was originally in the RSGB Bulletin for February 1960.
In October 1961, there is a simple portable transceiver described in CQ,
and it too uses a crystal controlled regen detector.
Likely there are others, but those are the ones I know about.
Admittedly three of those four are regen and not superregen, but that's
hardly important since a superregen is just a regen detector which
is oscilalting and modulated with an above audible frequency. (And take
note that a Q-Multiplier is just a specific application of regeneration.)
Indeed, a regen is just an oscillator with controlled feedback, or
to look at it another way, a regen is just an amplifier with positive
feedback.
So you can take just about any oscillator circuit, and by adding some
sort of control so you can reduce things so it's on the verge of oscillation,
you've got a regen detector. Coupling into it may be easier with some
circuits than others. And once you've got a regen circuit, adding
some scheme to modulate it at an ultrasonic frequency will make it
a superregen.
Since a crystal controlled superregen is relatively rare (who knows
why), you might want to start things by going through this sequence
and figuring out the best setup yourself.
Someone once pointed out here that for a lot of uses you don't need ultra
selectivity, and so a crystal in a regen is out of place. I'm not sure
how this translates to a superregen.
Be advised that while there has been some work done on using better
tuned circuits in a superregen, it seems more likely that the reason
for it's passband is because it is in effect an oscillator that is
being modulated. Working on the quench circuit, both amplitude and
waveform, would seem to be a better path. Charles Kitchin has done
some work on this, and he claims he has gotten better results by
fiddling with this. I've never seen the article, but he wrote
about superregens in the Fall 1994 issue of Communications Quarterly.
His article about regens the following year was an interesting mix
of history and new circuits, and I suspect the superregen article is
similar. He's also had a few superregen circuits in QST (and in
the Handbook for an issue or two), but I don't have a bibliography handy.
There was a superregen receiver article of his on the ARRL site,
http://www.arrl.org at one point, and doing a search there would hopefully
find it.
Take note that most superregens use the same active element as both
the RF oscillator and as the quench oscillator. Separating them,
has certain advantages in trying to get the best quench, and these
days it pretty much costs nothing to add the extra circuitry, unlike
in the 30s when the superregen was pretty hot.
I'd be curious about why you think you need a crystal controlled
superregen. Most superregens have been at VHF (indeed, if you go
by some rules of thumb of making the quench frequency a certain percentage
of the operating frequency the quench will be at audio once you get below
a certain frequency), and getting a crystal to oscillate up
there may start being a problem especially the higher you go. Though,
I think some work has been done with SAW controlled superregens.
And whether it adds to selectivity is under question. Stability might
be a factor, but then the traditional wide bandwidth has always offset
any issue of frequency stability. Who cares about drift when a superregen's
bandwidth would basically cover most of the 2meter band?
Michael VE2BVW
The Eternal Squire
an upconversion stage as a simplified IF-amp/detector
with (I believe) outstanding sensitivity.
I'm interested in a crystal controlled regen also. Seeing
as I have no back issues, can you point me to a web
page?
The Eternal Squire
Michael Black <et...@FreeNet.Carleton.CA> wrote in message
news:bf4kt1$ajh$1...@freenet9.carleton.ca...
Bill Higdon
Bill Higdon
Tom Bruhns
...
> Take note that most superregens use the same active element as both
> the RF oscillator and as the quench oscillator. Separating them,
> has certain advantages in trying to get the best quench, and these
> days it pretty much costs nothing to add the extra circuitry, unlike
> in the 30s when the superregen was pretty hot.
...
First of all, many thanks, Michael, for quite a nice summary about
regen and superregen!
A distinction between regen and superregen is that in regen, the
feedback is kept (just) below the point of oscillation, and in a
superregen, it's above, but with quenching, so that the circuit
multiplies the energy in the tank when it's gated on by some factor
depending on the feedback and the time it's left on by the end of the
gating time. A jitter-free gating time and stable feedback should
result in a low output noise level.
I believe the quench frequency will limit how narrow the detector will
be, and if one uses a high Q tank (such as a crystal) with relatively
low feedback, that will limit the gain available. During the off
time, the tank energy must decay sufficiently to track the input
signal amplitude. Is there any value in having a short time just
after the "on" period when the tank is actively quenched?
An example of a commercial use of an externally-quenched superregen is
in a radar altimiter...I think it is/was something like APN-141. It
applied the gate just once per transmitted pulse, at a time which
locked to the return pulse, as I recall. I wish I could remember now
just how they tracked changes in the return time. I do recall that it
would scan over a range of delay times from very short to relatively
long, and could track from just a couple of meters up to at least a
few thousand meters. I also don't recall for sure, but suspect they
changed the gate duration depending on the delay time, since close-in
returns are much stronger than distant ones. Shorter gates mean less
gain. That's one way to control the gain of an externally-gated
superregen.
It should be possible to build a pretty accurate model of an
externally-gated superregen in SPICE...though to limit the time it
takes to run the simulation, it's probably useful to start with an
operating frequency not too far above the quench frequency, perhaps a
couple dozen times f(quench).
Cheers,
Tom
John Crabtree
>Bill Higdon (w.hi...@attbi.com) writes:
>> I need some references on crystal controlled super regens, I know they
>> have been replaced in most new designs. But I have a idea that would be
>> easier to do with a super regen as opposed to the new studd.
>> Thanks,
>> bILL hIGDON
A good pair of recent articles was written by Eddy Insam:
i) E. Insam, Designing Super-regens, Electronics World,
April 2002, pp46-53
ii) E. Insam, Applications for Super-regens, Electronics World,
May 2002, pp18-20
Electronics World is the UK publication, which was for many years Wireless
World and finally metamorphosed into Electronics World.
Insam covers most of the key ground, ie separate and self quenching, what
controls bandwidth (rate of change of the conductance as oscillation starts),
why crystals are rarely if ever used (see later), etc.
Insam mentions, but does not go into any real detail on the use of SAW delay
lines in super-regens. This was mentioned by Pat Hawker, G3VA, in a piece he
wrote in Electronics and Wireless World (Wireless World as was), Dec 1987, p
1275. He referred to a paper by Darrell Ash of RF Monolithics:
'A low cost super-regenerative s.a.w. receiver', IEEE Trans-CE, Vol 33, August
1987, pp395-404
where better sensitivity, reduced bandwidth and better stability were claimed.
On the RF Monolithics web site (www.rfm.com) you will also details of the
Ampilifier Sequence Hybrid (ASH) receiver architecture, which is an alternative
to using super-regens at VHF and UHF frequencies.
In his piece, G3VA also mentioned some work done by Nat Bradley, ZL3VN, on the
use of crystals in super-regens (also written up in two pieces in his column in
the RSGB's Radio Communication, ca. 1979 and 1980). He tried using crystal
controlled receivers, but ".... did not have any great success with them. None
would operate on Xtal fundamental, but a harmonic Xtal could be made to
super-regenerate, kept under some control, and feed a separate detector for
some results. Try that approach." The quoted part comes from personal
correspondence in Nov 1999. It would seem that he was using the crystal
super-regeneration stage as an amplifier, with a separate detector.
BTW the standard reference on super-regenerative receivers was and still is:
J.R. Whitehead, Super-Regenerative Receivers, Cambridge University Press, 1950.
This excellent book contains all the theory. Whitehead was at the
Telecommunications Research Establishment in the UK during WW2 and played a, if
not the, key role in designing the Mk.3 I.F.F. (Identify-Friend-Foe) equipment
which used a super-regen receiver. The allies built 170,000 of these units.
>In the September 1972 issue of CQ, Irwin Math in his Math Notes column had
>a circuit for a crystal controlled superregen receiver. The same column
>had that then novel transmitter in an IC, the LP-2000.
>
>It was simply a cross-coupled multivibrator made up of two transistors,
>running at 20KHz, with a tank circuit with a crystal added in. He said
>the one he built at 10MHz worked fine. He pulled the circuit out of
>an issue of EDN, but no date is given.
>
>> snip >>
>
>Admittedly three of those four are regen and not superregen, but that's
>hardly important since a superregen is just a regen detector which
>is oscillating and modulated with an above audible frequency. (And take
>note that a Q-Multiplier is just a specific application of regeneration.)
I would look at it differently. A super-regen is a sampled data system. The
word modulated conveys the wrong idea to me.
>Indeed, a regen is just an oscillator with controlled feedback, or
>to look at it another way, a regen is just an amplifier with positive
>feedback.
>
>So you can take just about any oscillator circuit, and by adding some
>sort of control so you can reduce things so it's on the verge of oscillation,
>you've got a regen detector. Coupling into it may be easier with some
>circuits than others. And once you've got a regen circuit, adding
>some scheme to modulate it at an ultrasonic frequency will make it
>a superregen.
It is also easy to end up with a super-regen cct by accident if trying to build
a reqen cct. :-)
>Since a crystal controlled superregen is relatively rare (who knows
>why), you might want to start things by going through this sequence
>and figuring out the best setup yourself.
Time to tackle this issue. The rise time of the oscillations and quench time
(when you want the oscillations to drop to zero) are both dependent upon the Q
of the oscillator, which for a crystal is very high. Insam discusses this on
p. 49 of his first article, and ends by saying that a 30MHz crystal may place a
quench rate limit of 400Hz. Clearly this could only be used for low bit-rate
data recovery.
>Someone once pointed out here that for a lot of uses you don't need ultra
>selectivity, and so a crystal in a regen is out of place. I'm not sure
>how this translates to a superregen.
>
>Be advised that while there has been some work done on using better
>tuned circuits in a superregen, it seems more likely that the reason
>for it's passband is because it is in effect an oscillator that is
>being modulated. Working on the quench circuit, both amplitude and
>waveform, would seem to be a better path. Charles Kitchin has done
>some work on this, and he claims he has gotten better results by
>fiddling with this. I've never seen the article, but he wrote
>about superregens in the Fall 1994 issue of Communications Quarterly.
>His article about regens the following year was an interesting mix
>of history and new circuits, and I suspect the superregen article is
>similar. He's also had a few superregen circuits in QST (and in
>the Handbook for an issue or two), but I don't have a bibliography handy.
>There was a superregen receiver article of his on the ARRL site,
>http://www.arrl.org at one point, and doing a search there would hopefully
>find it.
Here are the details of Charles Kitchin's. N1TEV, more recent work on
super-regens:
Dec 1997 QST 39 An Ultra-Simple VHF Receiver for 6 Meters
Article: http://www.arrl.org/tis/info/pdf/9712039.pdf (187,078 bytes)
May 2001 QEX p62 New Super-Regenerative Circuits for Amateur VHF and UHF
Experimentation (Letters to the Editor)
Sep 2000 QEX pp18-32 New Super-Regenerative Circuits for Amateur
VHF and UHF Experimentation
In the Sept 2000 QEX article Charles states that the sinusoidal quench waveform
is the reason that the receiver can detect NBFM. I believe that quench
waveform is the effect of the operating conditons where the rate of change of
conductance as the oscillations start is low. One should also realise that
this is a self-quenched receiver.
The other interesting effect is that for a logarithmic mode super-regen, the
selectivity improves as the signal to noise ratio DECREASES. Whitehead
discussed this in his book, and Charles Kitchin's resuts agree.
>Take note that most superregens use the same active element as both
>the RF oscillator and as the quench oscillator. Separating them,
>has certain advantages in trying to get the best quench, and these
>days it pretty much costs nothing to add the extra circuitry, unlike
>in the 30s when the superregen was pretty hot.
>
>I'd be curious about why you think you need a crystal controlled
>superregen. Most superregens have been at VHF (indeed, if you go
>by some rules of thumb of making the quench frequency a certain percentage
>of the operating frequency the quench will be at audio once you get below
>a certain frequency), and getting a crystal to oscillate up
>there may start being a problem especially the higher you go. Though,
>I think some work has been done with SAW controlled superregens.
>And whether it adds to selectivity is under question. Stability might
>be a factor, but then the traditional wide bandwidth has always offset
>any issue of frequency stability. Who cares about drift when a superregen's
>bandwidth would basically cover most of the 2meter band?
>
> Michael VE2BVW
So where does this leave us? For VHF/UHF applications, one has:
i) L-C controlled super-regen
ii) SAW controlled super-regen
iii) ASH receiver
iv) low-cost superhet.
For this last option, see:
I. Hickman, Super-regen or Super-replacement, Electronics World (UK), Feb 1999,
pp138-143 where he discusses the use of a low cost superhet chip from Micrel as
an alternative to super-regens for data transmission. See also the Micrel web
site at www.micrel.com, and look for the 'Qwikwire' product details in the
'Radio Frequency' section.
HTH and that I have not written too much.
73 John KC0GGH
Minneapolis, MN