EMRFD Message Archive 390

Message Date From Subject
390 2007-01-29 19:14:36 Stephen Some basic questions
Hi all, new here and just picked up a copy of this great book. I
originally posted the message to below to the QRP-L list and it was
suggested that I come here (I didn't even know this group existed) I
am wanting to start building stuff from scratch, I have built plenty
of kits in the past, but never anything "truely" homebrew and figured
this book was a great place to start!

Here is the email I posted to QRP-L:

Hi everyone, name here is Stephen, N1VLV. I got interested in this
list from listening to the solder smoke podcast (RIP KL7R, sad sad
news) and also a host of other things radio. I've been a ham for 13
years, and I am recently returning to it full swing and loving every
minute of it along the way.

I just bought and received a copy of Experimental Methods in RF
design, an excellent book, even though a lot of it is over my head. I
do have an electronics/radio background, was a ground radio repairman
in the USAF, and now work in cellular as a switch technician, but
never did any board level repair and as such I find myself needing to
relearn basic electronics, especially the AC side of the house.

I am in the very first chapter of the book, and I want to build the
NE602 based receiver to start with as the book suggests. The circuit
looks fairly straightforward, but I don't quite understand how L1/L2
relate to it, and the associated capacitors surrounding them. I have a
basic understanding of whats inside the NE602 and how some of it
works, but still a little lost as to the rest of the support circuitry.

I am thinking that L1 in conjunction with the caps surrounding it
connected to pins 1&2 is some kind of parallel resonant circuit? Could
I safely assume the same for L2 and associated components for the LO
on pins 6&7? Or am I off course altogether?

If someone would be willing to explain this in a little more detail to
enhance my understanding that would be a great help!

73's de N1VLV
Stephen
Johnson City, TN
395 2007-01-30 13:07:02 Mike Brainard Re: Some basic questions
Stephen,

You are correct that both L1 and L2 form parallel resonant circuits.

L2 and the associated components form the resonator or "tank circuit"
for a Colpitts type oscillator operating very close to the frequency of
the incoming signal which is to be received. The connection of these
components to one of the internal transistors (the oscillator
transistor) in the NE602 is shown in figure 1.10(A). Additional
information on the NE602 is included beginning on page 5.10 in the
Mixers chapter. Additional information on how such oscillator circuits
work is included in Chapter 4. The two 680 pF capacitors are frequently
referred to as the Colpitts capacitors. Don't let the fact that the
transistor is internal to the NE602 bother you. It will work like any
other NPN.

L1 and the capacitors connected to it form the input filter, a single
tuned circuit. This circuit is adjusted to resonate at the desired
signal frequency. Desired signals pass through it from the antenna
connection to the input of NE602 with minimum attenuation. Other
signals which are far removed from the desired signal frequency will be
attenuated more by the action of this filter. These undesired signals
may be very strong at the antenna connection and may tend to overload
the NE602 input and cause various distortions and other problems. The
attenuation provided by the input filter helps to protect the NE602
from strong out of band signals and therefore minimize these problems.
The input transistors in the NE602 are arranged as a differential
amplifier (see figure 5.26). The 0.1 uF capacitor connected to pin 2 of
the NE602 provides an AC ground on this pin (a transistor base lead)
while allowing the internally generated DC bias voltage to remain. The
signal is then connected to the other input at pin 1. This is one
possible way to connect the unbalanced signal which exists at the
output of the filter to the differential (or balanced) input of the
NE602.

Mike
AD5RJ

> I am thinking that L1 in conjunction with the caps surrounding it
> connected to pins 1&2 is some kind of parallel resonant circuit? Could
> I safely assume the same for L2 and associated components for the LO
> on pins 6&7? Or am I off course altogether?
398 2007-01-30 14:37:56 jr_dakota Re: Some basic questions
399 2007-01-30 16:29:54 Stephen Brown Jr Re: Some basic questions
Awesome! This makes much more sense to me now, I have spent a good part of the day with my nose buried in an ARRL handbook and performing variousĀ  calculations just to see how much I could "relearn" on basic electronic theory. After calculating resonance, learning how AC reacts to caps/inductors etc, I now also remember who "ELI the ICE man" is hehe. Been a just a few years!

What's amazing to me about this book as that I am only in the first few pages and already I am wanting to start experimenting with this, good job authors :)

I did have one question on parts procurement, I am having a hard time locating some of the variable and trimmer caps, I may not have been looking hard enough or in the wrong places. I already have an NE602 equivalent, I believe it's an NTE part but it escapes me as to what the part number is at the moment, I bought it from a local electronics house and it cross references to an NE602. The mica trim caps I have been able to locate are a little on the expensive side ($6 on avg) and I think that's a wee too much for something so small so I'm looking for an alternative or a cheaper source, the only place I have checked so far is ebay and digikey. Anyone know of a good source for things like this?

One other question on C1/C2. Just for clarification, as these are turned they in effect move through their ranges of effective capacitance which in turn changes the resonance on the circuit as a whole? With that being said, these are essentially the "tuner" of the receiver, which goes directly to the LO in the '602? If I remember correctly(and I'll reference the book as well), mixers work by taking the sum/difference of a received signal and outputting an IF signal or in this case an audio signal?

I want to build this on a breadboard to start with, then I will probably attempt my first ever ugly style construction and build an enclosure out of PCB material. If anyone knows of a good source to get the rest of the parts needed(trim caps, variable caps, toroids, wire, etc) please let me know.

Thanks to all that replied to me today, I have gained some valuable insight and I can't wait to start playing with some of this stuff!!!!!

73's
Stephen
N1VLV


On 1/30/07, Mike Brainard <mbrainard@prodigy.net> wrote:

Stephen,

You are correct that both L1 and L2 form parallel resonant circuits.

L2 and the associated components form the resonator or "tank circuit"
for a Colpitts type oscillator operating very close to the frequency of
the incoming signal which is to be received. The connection of these
components to one of the internal transistors (the oscillator
transistor) in the NE602 is shown in figure 1.10(A). Additional
information on the NE602 is included beginning on page 5.10 in the
Mixers chapter. Additional information on how such oscillator circuits
work is included in Chapter 4. The two 680 pF capacitors are frequently
referred to as the Colpitts capacitors. Don't let the fact that the
transistor is internal to the NE602 bother you. It will work like any
other NPN.

L1 and the capacitors connected to it form the input filter, a single
tuned circuit. This circuit is adjusted to resonate at the desired
signal frequency. Desired signals pass through it from the antenna
connection to the input of NE602 with minimum attenuation. Other
signals which are far removed from the desired signal frequency will be
attenuated more by the action of this filter. These undesired signals
may be very strong at the antenna connection and may tend to overload
the NE602 input and cause various distortions and other problems. The
attenuation provided by the input filter helps to protect the NE602
from strong out of band signals and therefore minimize these problems.
The input transistors in the NE602 are arranged as a differential
amplifier (see figure 5.26). The 0.1 uF capacitor connected to pin 2 of
the NE602 provides an AC ground on this pin (a transistor base lead)
while allowing the internally generated DC bias voltage to remain. The
signal is then connected to the other input at pin 1. This is one
possible way to connect the unbalanced signal which exists at the
output of the filter to the differential (or balanced) input of the
NE602.

Mike
AD5RJ

> I am thinking that L1 in conjunction with the caps surrounding it
> connected to pins 1&2 is some kind of parallel resonant circuit? Could
> I safely assume the same for L2 and associated components for the LO
> on pins 6&7? Or am I off course altogether?


406 2007-01-30 18:49:35 michael taylor Re: Some basic questions
On 1/30/07, Stephen Brown Jr <stephen.brown75@gmail.com> wrote:
>
> I did have one question on parts procurement, I am having a hard time locating
> some of the variable and trimmer caps, I may not have been looking hard enough

For air variable capacitors (air is the insulator), look at surplus
dealers, e.g. <http://www.oselectronics.com/ose_p96.htm>.

> from a local electronics house and it cross references to an NE602. The mica
> trim caps I have been able to locate are a little on the expensive side ($6 on avg)

The (silver) mica is for temperature stability, which can be an issue
for on-the-air radios, but if you are just getting started "playing"
more than wanting to build an all-singing, all-dancing radio, I'd just
go with the cheapest small trimmer for prototyping I could find.


> With that being said, these are essentially the "tuner" of the receiver, which
> goes directly to the LO in the '602? If I remember correctly(and I'll reference
> the book as well), mixers work by taking the sum/difference of a received signal
> and outputting an IF signal or in this case an audio signal?

Yes, that sounds like you have the right idea of what is going on.

> I want to build this on a breadboard to start with, then I will probably attempt my
> first ever ugly style construction and build an enclosure out of PCB material. If

You shouldn't try to build RF circuits with a solderless breadboard,
there is too much stray capacitance and will cause problems. This is
mentioned in chapter 1, section 1.2 subtitle Breadboarding.

Enjoy,
Michael Taylor, VE3TIX
407 2007-01-30 21:04:20 Wes Hayward Re: Some basic questions
Hi Stephen,

Great to have you join us in this fascinating and addicting game we
play. It appears that the guys are getting you moving in the right
direction with the comments offered. You are obviously on the right
track with your reading and analysis.

One comment that has probably been said before, but always bears
repeating is in regard to the policy that I used with the projects
that were published in the book. When I described a circuit in the
book, which was often just one of a kind, I was always careful to
list the exact parts that I used. For want of a better descriptor,
I call it "DeMaw's Rule," for this was one of the policies that Doug
DeMaw promoted when he was the Technical Department Manager at
ARRL. In QST articles or Handbook items, I usually try to look for
generally available parts that are well priced (if possible) and
avoid special items such as surplus junk box variable capacitors and
the like. This becomes important when a project is going to be
duplicated. EMRFD is more about the general design ideas and
methods rather than about specific projects. So we assumed that it
was generally fair to make just one model, measure it, and describe
it as it was built. Still, DeMaw's rule says that I don't call out
parts that are different than I actually used.

In the case of that mica compression trimmer used to tune L1 in the
little receiver on page 1.9, I merely used it because it was in the
junk box. Truth be known, I probably grabbed the variable,
measured it, and then designed the rest of the tuned circuit to fit
the variable. Moral to the story is to use what you have in your
junk box, or can find at the right price. Don't assume when you see
a specific component used that it is firmly specified and cast in
stone, for it rarely is. Some redesign may be required, but that
only adds to the fun.

This design procedure was fairly direct. The 180 pF trimmer had a
nominal C of around 100 pF with a lot of range around that value. I
wanted this to be just part of the C in the tuned circuit, perhaps 20
to 30 % of the total. A standard value (also in my junk box) was
270. If you add 100 (the nominal trimmer value) to this you end up
with 370. Figure another 30 or 40 pF for coupling to the outside
world and you end up with a total of 400. This resonates at 7 MHz
with 1.3 uH. The inductance constant for a T37-6 toroid is 2.9
nH/turn^2. So 1300 nH results with N=square root of 1300/2.9 = 21
t. If you use a turn or two less, you can compress turns just a bit
and end up with the desired L. While it is always good to measure
the L, I probably didn't in this instance. The toroids are
dependable enough with a modest trimmer.

Let's now take a look at the Q of this circuit. There will be two
external components that load the coil. One will be the NE602 while
the other is the antenna. The 602 has an input impedance of about
3K between the two bases at pins 1 and 2. This is both measured
data and is consistent with the standing current of 2 mA in a diff-
pair where the beta is modest and operating well below F-t. Let's
assume for the moment that we will have a similar load from the
antenna. So the net parallel R from the two is 1.5K. The
inductive reactance of the 1.3 uH coil is 57 Ohms. The loaded Q,
neglecting any coil loss, will then be 1500/57, or 26. This is low
enough that the whole band can be covered without retuning, but is
high enough that it offers some protection from signals away from 40
meters.

The loaded Q is quite a ways down from the typical unloaded Q that we
might expect from this toroid core at 7 MHz, which is around 150.
With the Q dominated by the external loads, there will only be around
1 dB of insertion loss in this front end filter.

All that is left for us to do is to figure out what we need for a
capacitor to load the tuned circuit. We want a series cap that will
make 50 Ohms look like 3K when it appears across the inductor. We
find this in IRFD, p79, Eq 3.1-5, or derive it when needed. We will
want a cap with a reactance of (3000*50-50^2)^0.5=384 Ohms. This is
satisfied with an 59 pF capacitor. I used 47, which is close enough.

Loading changes as we back off on the 1K RF gain pot, but that does
not really matter, for we want reduced gain in that condition.

How else might we have done this? One method would have been to use
link coupling into L1. If we wanted a link from 50 Ohms to appear as
3K across L1, we would want an impedance ratio of 3000/50, or 60.
The turns ratio would then be approximated by the square root of
this, or 7.7. With 20 turns, a link of 2.6 turns is suggested.
Use 3 and things would be close. To really design this in detail,
you need to know the coupling coefficient. The square root
dependence assumes k=1, but that is only a good number for ferrite
cores. k=0.8 is a closer number for a powdered iron transformer on
a toroid with a permeability of 8, which is what we have with the -
6. Exact values don't much matter in this application though, so 2
or 3 turns would work well.

A similar procedure was used with the oscillator. You can follow
the resonance part of that evoluti
422 2007-02-01 08:21:15 jim.scott@scottel... Re: Some basic questions
** Reply to note from "michael taylor" <mctylr@gmail.com> Tue, 30 Jan 2007 21:49:32 -0500


FWIW - Air Variables - There is a source for the 365pf Air Variables single section and multi-ganged versions especially
popular with the hobbiests building MF, HF, grid leak, regens, and crystal based sets. All stuff for simple home
constructed radios. The organization has a newsletter too.

http://www.midnightscience.com

Alfred Morgan would be proud :-)

Later,
JIm Scott
WB0IYC
8657 2013-05-15 20:25:43 mm0gyx Re: Some basic questions
> All that is left for us to do is to figure out what we need for a
> capacitor to load the tuned circuit. We want a series cap that will
> make 50 Ohms look like 3K when it appears across the inductor. We
> find this in IRFD, p79, Eq 3.1-5, or derive it when needed. We will
> want a cap with a reactance of (3000*50-50^2)^0.5=384 Ohms. This is
> satisfied with an 59 pF capacitor. I used 47, which is close enough.

I'm a total beginner to rf design; this is my first post so please be gentle. I have EMRFD now and have stopped at the 602/386 dc rx. I don't own IRFD.

I don't follow the above quote from Wes relating to the 47pf capacitor in the band pass circuit, I see the formula but don't really understand how it's utilised. Is there a simple way to explain how the impedance matching works. Am I correct in assuming a nominal antenna impedance of 50 ohms is being referred to, which eventually has to match in to the 602 which is 3000 ohms?

I hope the question makes sense.

I've done some ugly construction, latest was a rockmite 30 from bits I gathered together. I'd like to develop a better knowledge of the design of equipment.

73,

Ian, MM0GYX
8658 2013-05-16 09:12:27 Nick Kennedy Re: Some basic questions
That little bit you quoted is pretty interesting and enlightening on the
subject of impedance conversion.

Consider that you have a black box with two terminals and you measure the
impedance at 50 -j384 at some frequency. Expressed in that way, it's
implying a 50 ohm resistor with a -j384 capacitor in series. But you could
also express that impedance as a conductance, or conductance plus
susceptance (I may get lost on my terms for inverse reactance here).

Take the inverse of 50 -j384 and you get the susceptance of the same
network as 3.334E-4 +2.56E-3j. Expressing these susceptance terms as a sum
implies a parallel network of a real value of 3.334E-4 and an imaginary
value of 2.56E-3j. If you individually take the reciprocal of each of
those terms, you see their values as resistance and reactance.
1/3.334E-4 = 2999 ohms and
1/2.56E-3j = -391j

So this demonstrates that 50 ohms with -384j ohms of capacitive reactance
in series is equivalent to 3000 ohms with -391j ohms of capacitive
reactance in parallel.

Or put another way, by putting -384j ohms in series with 50 ohms you have
transformed the 50 ohms into 3000 ohms in parallel with -391j ohms.

73-

Nick, WA5BDU


[Non-text portions of this message have been removed]
8659 2013-05-17 06:30:01 mm0gyx Re: Some basic questions
Nick,

I'm not following you, I see something that looks like j notation, which is not something I understand. I'll re-read it but...

73,

Ian

8660 2013-05-17 06:39:46 Thomas S. Knutsen Re: Some basic questions
This is basic Impedance calculations, I would reccomend reading up on
complex impedances.
Most College level engineering books covering basic DC and AC would have an
good explaination, and excerices.

This seems like an nice coverage, but no introducti
8661 2013-05-17 17:50:43 Nick Kennedy Re: Some basic questions
Yes, this approach might be difficult if you're not familiar with complex
number arithmetic. Of course, the j coefficient in electrical engineering
is equivalent to i in mathematics. Some manipulation of complex quantities
to help get familiar with them is possible with calculators like the HP48
series and probably many others, and Excel will also do complex number
calcs but requires some cumbersome functions to do so. The HP48 will mix
and match real, imaginary and complex quantities effortlessly.

The method of converting between resistance plus reactance (series version)
and equivalent resistance in parallel with reactance is extremely useful in
understanding impedance matching with reactances. I came onto it late in
my ham radio and engineering career and had quite a "wow" reaction.

If complex arithmetic is too daunting, a Smith Chart might tell you the
same thing visually. I've never gotten really handy with a Smith Chart, so
I usually don't take that route. Lots of nice Smith Chart software out
there now though.

73-

Nick, WA5BDU



8662 2013-05-17 23:31:17 Johan Bodin Re: Some basic questions
CalcEd by Wolf DL4YHF (the author of Spectrum Lab) is another tool that
lets you work easily with complex numbers. It looks like a simple text
editor for Windows but it has an advanced calculator built-in. Just put
some mathematical expressions in the text that you are typing, hit the
calculate key and the math is evaluated in-place in your text. It has a
curve plotting function too.

The first few lines from the readme file:
----------------
Introduction

On the first glance, CalcEd looks like NOTEPAD. In fact, it's a simple
text editor for plain text (*.TXT) and Rich Text files (*.RTF).
But the main purpose is: CALCULATING. You can use it for step-by-step
calculations, and save all steps in a simple textfile.
-------------
8663 2013-05-19 08:46:52 Dave Daniel Re: Some basic questions
The ARRL handbook would be a good book to read up on complex impedances.
If you don't have a copy, I'm sure you can find a used copy for a pretty
inexpensive price using bookfinder.com. It doesn't have to be a recent
edition, either.

73,
DaveD

8664 2013-05-19 08:46:57 mm0gyx Re: Some basic questions
Thomas, Nick, Johan,

Very many thanks for replying, I'll read through the messages and see where it takes me. I am quite keen to approach this mathematically if possible.

I'll get reading.

73,

Ian



8666 2013-05-20 20:55:51 mm0gyx Re: Some basic questions
Thanks Dave,

I'm reading up on complex numbers and expanding my basic electrics. Thanks to all for your advice.

73,

Ian.

8667 2013-05-20 20:56:04 Dave Hartman Re: Some basic questions
Maybe look at this site:

http://www.allaboutcircuits.com/

Dave - AC2GL
8668 2013-05-20 20:56:06 Dave Hartman Re: Some basic questions
Here's another site:

http://openbookproject.net/electricCircuits/

Dave - AC2GL