"Jock Cooper" wrote in message
...
Hi everyone,

I've been interested in radios and how they work since I was a kid --
I had a crystal radio kit and a RF-2200. Recently I started building
radios and have been having a blast. Please keep in mind that I
am very newbie.


[tale of woe deleted]

I'd start again with a simple Hartley oscillator using a tapped toroid (you
don't need a choke). Get the basic oscillator working (a scope helps) then
add a buffer stage.

Forget about getting the oscillator and front end tuning to track properly
at this stage, use separate capacitors. When you get the thing working
properly you can sort out the tracking.

73, Leon
--
Leon Heller, G1HSM

http://www.geocities.com/leon_heller

"Jock Cooper" wrote in message
...
Hi everyone,

I've been interested in radios and how they work since I was a kid --
I had a crystal radio kit and a RF-2200. Recently I started building
radios and have been having a blast. Please keep in mind that I
am very newbie.


[tale of woe deleted]

I'd start again with a simple Hartley oscillator using a tapped toroid (you
don't need a choke). Get the basic oscillator working (a scope helps) then
add a buffer stage.

Forget about getting the oscillator and front end tuning to track properly
at this stage, use separate capacitors. When you get the thing working
properly you can sort out the tracking.

73, Leon
--
Leon Heller, G1HSM

http://www.geocities.com/leon_heller

Hi Jock,
there is nothing you should apologize for. My recommendations:
1/ Do not use breadboards for RF. They are perfect for AF, good for
digital citcuits, but RF circuits sometimes work, more often not. And
keep the groundplane as big as possible. The "ugly", "rats-nest",
"dead bug" or "manhattan" style really seem to be far the best for RF
prototyping. Look at
http://w1.859.telia.com/~u85920178/blocks/deadbug_0.htm for more info.
2/ Get an oscilloscope covering at least three times your frequency
range. An add-on unit for your PC can turn it into an oscillscope and
spectrum analyzer for a fair price.
3/ Use the necessary tuning capacitors and their configuration to keep
the oscillator and input frequency near to each other at more than one
point. The configuration and formulae can be found in many books.
4/ Try the oscillator from
http://w1.859.telia.com/~u85920178/blocks/osc7m00.htm ,it should work
reliably in a wide frequency band.

BR from Ivan OK1SIP

Hi Jock,
there is nothing you should apologize for. My recommendations:
1/ Do not use breadboards for RF. They are perfect for AF, good for
digital citcuits, but RF circuits sometimes work, more often not. And
keep the groundplane as big as possible. The "ugly", "rats-nest",
"dead bug" or "manhattan" style really seem to be far the best for RF
prototyping. Look at
http://w1.859.telia.com/~u85920178/blocks/deadbug_0.htm for more info.
2/ Get an oscilloscope covering at least three times your frequency
range. An add-on unit for your PC can turn it into an oscillscope and
spectrum analyzer for a fair price.
3/ Use the necessary tuning capacitors and their configuration to keep
the oscillator and input frequency near to each other at more than one
point. The configuration and formulae can be found in many books.
4/ Try the oscillator from
http://w1.859.telia.com/~u85920178/blocks/osc7m00.htm ,it should work
reliably in a wide frequency band.

BR from Ivan OK1SIP

Jock Cooper writes:

Hi everyone,
[snip]

Thanks to everyone for the great and helpful suggestions. This sure
is a friendly group.

Since I just recently started all this tinkering, my junkbox is a
little low in some areas, notably NP0 capacitors.. I have a few in
values from 3.9 up to 47 and a few 100s but not else. So I've used
various other types when higher values are called for.. I'm planning
on picking up an assortment soon.

Also I have some copper clad non drilled pc boards, I'm going to try
to rebuild the oscillator (and radio) using the 'ugly' method;
although I'm a little afraid I'll make myself a tangled mess of
useless and stuck together parts. The main reason I use breadboards is
so that I can take parts in and out of the circuit quickly (ie I'm at
the stage where I'm curious about what happens when I take such and
such part out, or change its value etc).

I'd love to get a oscilloscope but my 'play' budget is mostly spent
for awhile. So I've been making due with an LCR and a little cheapo
DMM.

I did order the book _Experimental Methods in RF Design_ .. I've also
got a lot of good info from an old (well 1998) ARRL handbook I got
on Ebay.

Jock

Jock Cooper writes:

Hi everyone,
[snip]

Thanks to everyone for the great and helpful suggestions. This sure
is a friendly group.

Since I just recently started all this tinkering, my junkbox is a
little low in some areas, notably NP0 capacitors.. I have a few in
values from 3.9 up to 47 and a few 100s but not else. So I've used
various other types when higher values are called for.. I'm planning
on picking up an assortment soon.

Also I have some copper clad non drilled pc boards, I'm going to try
to rebuild the oscillator (and radio) using the 'ugly' method;
although I'm a little afraid I'll make myself a tangled mess of
useless and stuck together parts. The main reason I use breadboards is
so that I can take parts in and out of the circuit quickly (ie I'm at
the stage where I'm curious about what happens when I take such and
such part out, or change its value etc).

I'd love to get a oscilloscope but my 'play' budget is mostly spent
for awhile. So I've been making due with an LCR and a little cheapo
DMM.

I did order the book _Experimental Methods in RF Design_ .. I've also
got a lot of good info from an old (well 1998) ARRL handbook I got
on Ebay.

Jock

You can series and parallel any number of capacitors to get the right
value. A given oscillator type works best within some range of reactance
values for L and C. If you go too far outside the ideal range, stability
will suffer, noise might increase, and if too extreme, it might not
oscillate. The ideal range depends on the oscillator design. That being
said, you can often vary L and C quite a great deal and still get
adequate operation.

Real inductors are, in general, a lot more different from their
theoretical ideal counterparts than capacitors are. Real inductors have
considerable loss and stray capacitance, as well as being temperature
sensitive. So you'll often find that a circuit works quite a bit
differently if you substitute an inductor for one with different
physical construction. For a VFO tank, I use only toroids wound on type
6 powdered iron material. (As I mentioned earlier, type 7 is probably as
good, but I've never used it.) For VFO tank capacitors, I use only NPO
ceramic. The combination produces low drift, dominated by the inductor.
If desired, the drift can be compensated by replacing part of the C with
negative temperature coefficient parts. Most other types of L and C have
substantially greater temperature sensitivity. You might get lucky and
have poor parts drift in opposite directions and more-or-less cancel,
but that's poor practice and hard to duplicate. Better to design for
minimal inherent drift, then compensate what's left if necessary.
Toroids have the big advantage of producing a relatively small external
field. That makes them much less sensitive to mounting, the influence of
nearby objects, and microphonics.

Jock Cooper wrote:
. . .
I wasn't sure if this was even OK to do (if it is too lossy or
someting), I figured I could experiment around with the result and
find out But maybe now I can just ask.. how much mucking around
can I do by adding in parallel and serial caps to tweak the variable
cap range? Also, Can I just find any old values of L and C that when
plugged in would give me the F I'm looking for? Or are there only
certain values of L and C that are appropriate for certain ranges of
Freq.

. . .

Roy Lewallen, W7EL

You can series and parallel any number of capacitors to get the right
value. A given oscillator type works best within some range of reactance
values for L and C. If you go too far outside the ideal range, stability
will suffer, noise might increase, and if too extreme, it might not
oscillate. The ideal range depends on the oscillator design. That being
said, you can often vary L and C quite a great deal and still get
adequate operation.

Real inductors are, in general, a lot more different from their
theoretical ideal counterparts than capacitors are. Real inductors have
considerable loss and stray capacitance, as well as being temperature
sensitive. So you'll often find that a circuit works quite a bit
differently if you substitute an inductor for one with different
physical construction. For a VFO tank, I use only toroids wound on type
6 powdered iron material. (As I mentioned earlier, type 7 is probably as
good, but I've never used it.) For VFO tank capacitors, I use only NPO
ceramic. The combination produces low drift, dominated by the inductor.
If desired, the drift can be compensated by replacing part of the C with
negative temperature coefficient parts. Most other types of L and C have
substantially greater temperature sensitivity. You might get lucky and
have poor parts drift in opposite directions and more-or-less cancel,
but that's poor practice and hard to duplicate. Better to design for
minimal inherent drift, then compensate what's left if necessary.
Toroids have the big advantage of producing a relatively small external
field. That makes them much less sensitive to mounting, the influence of
nearby objects, and microphonics.

Jock Cooper wrote:
. . .
I wasn't sure if this was even OK to do (if it is too lossy or
someting), I figured I could experiment around with the result and
find out But maybe now I can just ask.. how much mucking around
can I do by adding in parallel and serial caps to tweak the variable
cap range? Also, Can I just find any old values of L and C that when
plugged in would give me the F I'm looking for? Or are there only
certain values of L and C that are appropriate for certain ranges of
Freq.

. . .

Roy Lewallen, W7EL

Roy Lewallen writes:

When you get _Experimental Methods. . _, look at the very first few
pages -- "Getting Started". You'll find a nice drawing and some other
information about "ugly construction".

Incidentally, the method was first described and the term coined in
the article "The Ugly Weekender" by Roger and Wes Hayward, in August
1981 QST. If you can get hold of a copy of this article, it has more
about the method, as well as some other good information.

If and when you do eventually decide to get a scope -- I've been
surprised to see Tek 465 scopes going on eBay for about $200. That's
an awful lot of bang for the buck.

Roy Lewallen, W7EL


Well last night I took some double sided copper clab board and cut off
a small square. (By the way, what is the best way to cut this stuff?)
I rebuilt the oscillator on there in 'ugly' style. And it was in fact
quite ugly. It looked like something I would have done as a kid. But
it worked! Now I'm sold on it. Hopefully I can get some pointers on
improving my technique from the book.

Now I have a new question though. The circuit is a Clapp based on the
MPF102 JFET. When I compute the frequency range, it is close to the
measured range but not the same; but there appears to be some
capacitance that I can't account for, and its value seems higher than
would be 'stray'.

The circuit I'm using is basically like the one at
http://www.electronics-tutorials.com...scillators.htm
in figure 2, but without the extra parallel stuff on the variable cap.
So I calculate the cap value by adding the var cap value in series
with the caps labelled on that website as cfb-a and cfb-b. (By the
way, what happens if those two aren't the same value?) Do any of the
other capacitors in the circuit factor in?

How much capacitance could my 'pill bottle inductor' be contributing?
I was careful, but not extra careful when winding it.

Roy Lewallen writes:

When you get _Experimental Methods. . _, look at the very first few
pages -- "Getting Started". You'll find a nice drawing and some other
information about "ugly construction".

Incidentally, the method was first described and the term coined in
the article "The Ugly Weekender" by Roger and Wes Hayward, in August
1981 QST. If you can get hold of a copy of this article, it has more
about the method, as well as some other good information.

If and when you do eventually decide to get a scope -- I've been
surprised to see Tek 465 scopes going on eBay for about $200. That's
an awful lot of bang for the buck.

Roy Lewallen, W7EL


Well last night I took some double sided copper clab board and cut off
a small square. (By the way, what is the best way to cut this stuff?)
I rebuilt the oscillator on there in 'ugly' style. And it was in fact
quite ugly. It looked like something I would have done as a kid. But
it worked! Now I'm sold on it. Hopefully I can get some pointers on
improving my technique from the book.

Now I have a new question though. The circuit is a Clapp based on the
MPF102 JFET. When I compute the frequency range, it is close to the
measured range but not the same; but there appears to be some
capacitance that I can't account for, and its value seems higher than
would be 'stray'.

The circuit I'm using is basically like the one at
http://www.electronics-tutorials.com...scillators.htm
in figure 2, but without the extra parallel stuff on the variable cap.
So I calculate the cap value by adding the var cap value in series
with the caps labelled on that website as cfb-a and cfb-b. (By the
way, what happens if those two aren't the same value?) Do any of the
other capacitors in the circuit factor in?

How much capacitance could my 'pill bottle inductor' be contributing?
I was careful, but not extra careful when winding it.