The Desrt Ratt HR regenerative receiver has a good reputation, and I'd
love to build one, but I am a quite mediocre homebrewer... If I can
get some potential problems out of the way I can better concentrate on
the many others that will surely crop up!

These are my doubts w.r.t. the DR2 as described in the .PDFs on Paul
Harden's page:
http://www.aoc.nrao.edu/~pharden/hobby/DR2.pdf
http://www.aoc.nrao.edu/~pharden/hobby/DR2descr.pdf .
Can anyone comment on any of these points?

1) Tuning Coil
--------------
The PDF mentions inductance, number of turns and wire gage,
but leave us in suspense apropos the coil former:
"T1 is wound directly on the PCB". Uh?
I guess anything 5-6uH will do, but I can't keep from wondering...

2) Tuning Varicap
-----------------
"R5 sets the current across D2-D3" The reverse resistance on such
rectifiers is huge, much greater than R5 (10kohm). I noticed that Paul
Harden N5AN once offered V/C measurements on diodes for use as varicaps
(http://www.kkn.net/archives/html/QRP.../msg00360.html)
There, one sees a 1N4148, forward biased. From the specs one gleans
that this diode has a sizeable current when forward biased at voltages
within a reasonable tunng range - but we're talking about something
quite different here. What gives?

3) Q2 (oscillator) biasing
--------------------------
The resistor between C6 and C7 is not rated. A base bias on an regen's
oscillator can be rather, well, critical....

4) IC Amplifier
------------
R12 is indicated as "10W".
What would a reasonable resistance rating be?

5) Voltage Regulator
--------------------
The text says that the drop across the forward biased LED1
is 1.3-1.8V, while across the Si rectifier D1 is 0.7V.
In series they'd sum up to 2 - 2.5V, but on the circuit
one reads a total drop of 3.0V, i.e. up to 1 V higher.
Charles Kitchin's design fed the oscillator just 2.1V...
Would 2.5V be enough? Or, do LEDs come with a wide variety
of voltage drops?


TIA to everyone!

I hope this will encourage others to experiment with this circuit.

Filippo (SpamHog)
N1JPR/I2

SpamHog wrote:

The Desrt Ratt HR regenerative receiver has a good reputation, and I'd
love to build one, but I am a quite mediocre homebrewer... If I can
get some potential problems out of the way I can better concentrate on
the many others that will surely crop up!

These are my doubts w.r.t. the DR2 as described in the .PDFs on Paul
Harden's page:
http://www.aoc.nrao.edu/~pharden/hobby/DR2.pdf
http://www.aoc.nrao.edu/~pharden/hobby/DR2descr.pdf .
Can anyone comment on any of these points?

1) Tuning Coil
--------------
The PDF mentions inductance, number of turns and wire gage,
but leave us in suspense apropos the coil former:
"T1 is wound directly on the PCB". Uh?
I guess anything 5-6uH will do, but I can't keep from wondering...

Here is a link to a few pictures of an older version of the Desert Ratt.

http://www.njqrp.org/qhbextra/2/2a.html

You have to scroll down a little bit, but this shows that the coil is
wound on a toroid. As long as it has good characteristics in the 3-30 MHz
range, any one can be used. The permeability may vary from one toroid to
another but you should be able experiment a little to get things right.
Maybe someone else could suggest a proper toroid number. The one there
looks like a T50-?? or T38-??. It's hard to tell.

2) Tuning Varicap
-----------------
"R5 sets the current across D2-D3" The reverse resistance on such
rectifiers is huge, much greater than R5 (10kohm). I noticed that Paul
Harden N5AN once offered V/C measurements on diodes for use as varicaps
(http://www.kkn.net/archives/html/QRP.../msg00360.html)
There, one sees a 1N4148, forward biased. From the specs one gleans
that this diode has a sizeable current when forward biased at voltages
within a reasonable tunng range - but we're talking about something
quite different here. What gives?

Yes, forward biased the 1N4148 will have a high current. If you read
further through that thread, someone else described a problem with using
it in that mode. The high current caused excessive heating and drift.
Not really recommended for use.

3) Q2 (oscillator) biasing
--------------------------
The resistor between C6 and C7 is not rated. A base bias on an regen's
oscillator can be rather, well, critical....

C6, C7, and R? are there to filter and stabilize the Base-Emitter
voltage and bypass the Base-Emitter junction from RF. A value of
100K should be sufficient. The value isn't really that critical.

Here is a link to a really old version of the Desert Ratt.

http://duke.usask.ca/~buydens/ham/ratt/index.htm

The coil is different but the biasing structure of Q2 is pretty much
the same. Also note the slightly different part arrangement on the
lm386. Get a spec sheet on the lm386 to see what each part is suppose
to do.

4) IC Amplifier
------------
R12 is indicated as "10W".
What would a reasonable resistance rating be?

This just looks like a misprint. I believe that should be 10 Ohms.
Just go to google and type in lm386 for all the information you can
stand on that chip. I didn't see any configuration that used it
in quite that way, but that doesn't mean it can't be used that wy.

5) Voltage Regulator
--------------------
The text says that the drop across the forward biased LED1
is 1.3-1.8V, while across the Si rectifier D1 is 0.7V.
In series they'd sum up to 2 - 2.5V, but on the circuit
one reads a total drop of 3.0V, i.e. up to 1 V higher.
Charles Kitchin's design fed the oscillator just 2.1V...
Would 2.5V be enough? Or, do LEDs come with a wide variety
of voltage drops?

The voltage drop across the diode D1 will be as specified
for just about any 1N914 you use, however, the voltage across
a LED depends on exactly which one you choose. Here is another
link to a page that describes the kind of voltage drops you
can expect from LEDs.

http://www.theledlight.com/LED101.html

It seems that this value can be any where between 1.7 and
4.6 volts depending on the type you choose. I seem to recall that
even when you get a LED from Radio Shack the specs are printed
on the back. Personally, I would shoot for the 3.0 volts, listed
in the schematic.

TIA to everyone!

I hope this will encourage others to experiment with this circuit.

Filippo (SpamHog)
N1JPR/I2

--
Martin E. Meserve - K7MEM
http://www.k7mem.150m.com

Thank you Martin! Great pointers! {grrr... I hate it when Google's
interface gives u no feedback & u end up reposting n times....
apologies for not using a serious news client}

I also got very interesting replies directly from Paul Harden.

Paul ran into some personal matters that required much of his
attention for a while, but he's well aware that the latest PDF on his
site has a few issues.


On Wed, 2 Feb 2005, Filippo Cattaneo wrote:

1) Tuning Coil
--------------
You mention inductance, number of turns and wire gage,
but leave us in tantalizing suspense w.r.t. the former:
"T1 is wound directly on the PCB"

I did the webpage about a year ago when I began designing a kit version
for the DR2. Since then, my xyl broke her foot, an extremely heavy
work load, and recently, a QRP club trying to kit the DR2 without my
permission. So I temporarily removed the coil winding info. I am
hoping to fire things back up to get the DR2 kitted this summer.

The problem with a kit, to make things consistent, is winding the coil.
For one-of-a-kind homebrew, you can use anything, such as the 35mm film
can suggested, etc.

- as if there was some secret advantage
in winding "on the PCB".

Nothing secret. For the kit, the coil will be wound directly
on the board, that is, there will be two "slots" tooled into the board
for winding the coil. It works surprisingly well and a fairly decent
Q. It is an approach much easier than other first-time kit methods of
winding a coil.

I guess anything 5-6uH will do,
but I can't keep from wondering:

That is a good range if you are using an air variable in the 200-300pF
range.

With using back-biased diodes as varicaps, such as the 1N4004, you only
get about 60pF of tuning range, and the new prototype I'm working on
about 80pF. For that reason, you want to have the coil inductance in
the 2-3uH region for wider tuning.

- p-Chloro-alpha,alpha,alpha-trifluorotoluene?
Yes, that's my secret, but keep it
under the rug for right now :-)


2) Tuning Varicap
-----------------
"R5 sets the current across D2-D3"
The reverse resistance on such rectifiers is huge,
much greater than R5.
I noticed your V/C measurements on diodes,
(http://www.kkn.net/archives/html/QRP.../msg00360.html)
and also that a 1N4148 has a sizeable current when forward biased
at voltages within a reasonable tunng range -
but we're taling about something quite different here.
What gives?

I've never found 1N914/1N4148 to be very consistent, and in fact, some
batches I've tested have the maximum capacitance when forward biased.
I prefer the 1N4004/4005 series, as the V/C response is very
consistent. Much of the idea of the DR2 is to build it with very
common components. Obviously, a 100-200pF range varicap would be
ideal, but they are hard to find and expensive. R5 can always be
tinkered with to get the maximum tuning range for the diode that is
actually selected.


3) Amplifier
------------
R12 is indicated as "10W"

R11 and R12 are 10 ohms. I never noticed that before on R12 -thanks.
That happens when you convert to .pdf and the translater fails to
convert the special symbol. The omega symbol in the symbol font is the
capital letter W. It just didn't get converted. 1/4W is fine.

Both resistors are optional. R11 helps those LM386 that like to
motor-boat when the capacitor (C19) is not enough. Just forms a low
pass filter to keep audio off the Vcc line. R12 helps cross-over
distortion, but frankly, the regen action alone adds far more -hi.


4) Voltage Regulator
--------------------
You say that the drop across the forward biased LED1 is 1.3-1.8V,
while across the Si rectifier D1 it's 0.7V.
In series they'd sum up to 2 - 2.5V, but you also say that
the voltage in the circuit is 3.0V, up to 1 V higher.
Charles Kitchin's design fed the oscillator just 2.1V...
Anything more we need to know? Would 2V be enough?

I have built several of these circuits with Kitchin's
3-diodes-in-series scheme with good results. The LED was added several
years ago when several asked how would they add an LED to indicate
power on. So I added the LED, using it as a "poor mans regulator" and
the power on indicator to kill 2 birds with one stone. D1 raises this
voltage a bit, providing the lower voltage as the bias to the emitter
follower, and affording some isolation between that bias and the regen
bias, which knocked down some objectionable oscillations.

I apologize for the confusion in voltages. The DR2 has been documented
and written about several times over the years, so discrepencies appear
to exist. The batch of LED's I purchased for the kit drops 2.2v,
making the total voltage almost 3v. Your mileage will differ.

One of the secrets of this regen stage, however, is the relatively low
operating voltage. This is what keeps it from going into a wild regen
oscillation like other circuits. The smoother regen action makes it
much easier to control, holding in there over a wider frequency range,
and makes it much nicer for AM detection. For this reason, this
detector is not well suited for CW reception, as you never really get
that "howl" type of regen oscillation.

Let me know if there are any other details I can help you with, and of
course, I'm always interested in how others have built theirs. Due to
having to postpone my plans to kit the DR2, the documentation never got
completed and there are a few "holes" in it. I am hoping to get the
prototype for the kit built before long, and when finalized, I will
update the schematic, exact parts list and circuit description,
including some drawings on how to build it "Manhattan Style."

I thank you for your attention
and hope you'll share some wisdom with us...

Likewise, let me know how you build yours on your end.

Best Regards,

Filippo Cattaneo
N1JPR/I2

72, Paul NA5N