EMRFD Message Archive 2468

Message Date From Subject
2468 2008-12-07 22:26:31 Stephen Wandling EA3FXF & EA3GHS and the Pulga
Hola Joan and Eduardo,

I continue to be mesmerized by your little Pulga transceiver. I have
your Sprat Word document and the Spice files with related Word doc, but
have not found too much more on this clever little rig.

All of the Pulga circuits I've seen have been for 40M. I am interested
in trying other frequencies, and I would like to know your design
criteria for the three windings on T1. If you have LC specs for other
frequencies, that will be great as well.

It is my understanding that you are able to vary the crystal oscillator
frequency somewhat with the variable resistor V1? Is this because D1 is
acting as a varactor responding to the shifting voltage on the base of Q1?

Have you or anyone tried using VXO or maybe even Super-VXO with the
Pulga? Intuitively it seems like it would work and should still allow
the receive crystal filtering to function?

Thank you for supplying the Spice models. The feedbackfilter.asc output
shows a peak at approximately 7.196 Mhz, whereas the pulga40g.asc shows
an output frequency of approximately 7.160 Mhz. Am I correct to assume
that the filter model is not related to the one used in the transmitter?

I have some 2SC2988 transistors on order and will build the 40M version
as soon as they arrive. It is nice to live remotely in the British
Columbia mountains, as I do, but I sometimes miss being able to go to a
local store and pick up that one component I forgot to order in my
Internet shopping!

Thank you for this intriguing design.

72
Stephen
VE7NSD

--
"You never change things by fighting the existing reality.
To change something, build a new model that makes the
existing model obsolete." -- Buckminster Fuller

"First they ignore you, then they ridicule you, then they
fight you, then you win" -- Gandhi
2469 2008-12-08 11:20:56 Eduardo Alonso Re: EA3FXF & EA3GHS and the Pulga
> Hola Joan and Eduardo,

Hello,
Here Joan and Eduardo taking a tea and answering your questions together.

> I continue to be mesmerized by your little Pulga transceiver. I have

Thanks your your words.... small circuits can hide complex effects.

> All of the Pulga circuits I've seen have been for 40M. I am interested
> in trying other frequencies, and I would like to know your design
> criteria for the three windings on T1. If you have LC specs for other
> frequencies, that will be great as well.

Sorry, We don't calculate this circuit for other frequencies.
Currently we don't have antenna for other bands.

The Rloss of the resonant circuit is 14kohm. L3:L1 transforms the 50ohm
of the antenna to 14kohm.

L2 and L1 has 1:1 turns ratio. The collector sees a 50ohm load impedance.
You can reduce L2 to get more power output.
(but the sensibility of the receiver could be lower)

C1 C2 is a capacitive transformer, from 14kohm to base-emitter
resistance rpi=220ohm

Use this calculator
http://www.rootshell.be/%7Eea3ghs/ctrafo.php to recalculate new values.

> It is my understanding that you are able to vary the crystal oscillator
> frequency somewhat with the variable resistor V1? Is this because D1 is
> acting as a varactor responding to the shifting voltage on the base of Q1?

Yees!! This is a "second order" behavior not described anywhere. Works
like a RIT.
I think that we are changing the base-emitter capacitance.

The diode is not working like a varicap. Don't put a varicap in //
with C3 because QL fall down.
The diode maintains the current of the xtal in the negative
semiperiode of the oscillation.

> Have you or anyone tried using VXO or maybe even Super-VXO with the
> Pulga? Intuitively it seems like it would work and should still allow
> the receive crystal filtering to function?

No, please report your experiences..
In our tests, we connected the receiver to a computer with KGKSDR software
and we received signals +-24khz...

> Thank you for supplying the Spice models. The feedbackfilter.asc output
> shows a peak at approximately 7.196 Mhz, whereas the pulga40g.asc shows
> an output frequency of approximately 7.160 Mhz. Am I correct to assume
> that the filter model is not related to the one used in the transmitter?

Well... it's very difficult to test the circuit in receive mode.
I think that the uncontrolled capacitances changes the resonant frequencies.
We went directly to breadboard to test it in RX mode.

> I have some 2SC2988 transistors on order and will build the 40M version

You can get more power, but the reception with 2SC2988 is worse than 2N3866.

> as soon as they arrive. It is nice to live remotely in the British
> Columbia mountains, as I do, but I sometimes miss being able to go to a
> local store and pick up that one component I forgot to order in my
> Internet shopping!

OK, if you have broadcasting stations near, you will need a high-pass filter to
attenuate MW stations.

> Thank you for this intriguing design.

> 72
> Stephen
> VE7NSD

73/DX
Joan and Eduardo
2471 2008-12-08 16:38:23 cwfingertalker Re: EA3FXF & EA3GHS and the Pulga
Stephen,

Where can I find information on this rig?

Bill N7EU
2477 2008-12-08 19:26:45 Stephen Wandling Re: EA3FXF & EA3GHS and the Pulga
Hi Bill,

All of the information I have can be accessed through the last item in
the Links section of this list:

/the flea, a 40m minimalist transceiver/

I find this little circuit to be very interesting, especially the
receiver selectivity that is demonstrated in one of the Spice files, and
the RIT effect that is still somewhat puzzling. I would be eager to
hear more on this from some of the 'old hands' here.

72
Stephen
VE7NSD

cwfingertalker wrote:
> Stephen,
>
> Where can I find information on this rig?
>
> Bill N7EU
>
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>
>
>

--
"You never change things by fighting the existing reality.
To change something, build a new model that makes the
existing model obsolete." -- Buckminster Fuller

"First they ignore you, then they ridicule you, then they
fight you, then you win" -- Gandhi
2484 2008-12-10 10:01:40 Stephen Wandling Re: EA3FXF & EA3GHS and the Pulga
Hola Joan y Eduardo,

Thank you for your detailed and enthusiastic responses to my questions.
I enjoyed my cup of tea while reading them.

I have many questions, but will limit it to 2 for now.

Thank you for your capacitive matching program link. To be sure that I
was using it correctly, I tried to enter values that would result in the
capacitor values found in the posted circuit for the Pulga, but I have
not been able to do this. I used Rin = 220; Rout = 14k; F = 7030; L = 7
uH; Q = 100. The Sprat circuit shows C1/C2 = 270/56, and the
calculator shows 411/86, with the above values. Can you please show me
what I am doing wrong?

Further to your mention below, I have installed the KGKSDR software. I
am assuming that you were able to connect the output of the Pulga to
this software? I am not familiar with SDR, but have assumed that I
could put the audio output of my Pixie (no parts yet for the Pulga) into
either the Line In or Microphone inputs of my sound card and see a
spectrum display of the passband. I have tried this but do not get
anything in MGKSDR. Can you point me to a guide for doing this?

I have ordered the 2N3866 so I will be able to see its effects on
improving reception.

Thanks again for your interesting minimalist transceiver.

72
Stephen
VE7NSD

Eduardo Alonso wrote:
>> Hola Joan and Eduardo,
>>
>
> Hello,
> Here Joan and Eduardo taking a tea and answering your questions together.
>
>
>> I continue to be mesmerized by your little Pulga transceiver. I have
>>
>
> Thanks your your words.... small circuits can hide complex effects.
>
>
>> All of the Pulga circuits I've seen have been for 40M. I am interested
>> in trying other frequencies, and I would like to know your design
>> criteria for the three windings on T1. If you have LC specs for other
>> frequencies, that will be great as well.
>>
>
> Sorry, We don't calculate this circuit for other frequencies.
> Currently we don't have antenna for other bands.
>
> The Rloss of the resonant circuit is 14kohm. L3:L1 transforms the 50ohm
> of the antenna to 14kohm.
>
> L2 and L1 has 1:1 turns ratio. The collector sees a 50ohm load impedance.
> You can reduce L2 to get more power output.
> (but the sensibility of the receiver could be lower)
>
> C1 C2 is a capacitive transformer, from 14kohm to base-emitter
> resistance rpi=220ohm
>
> Use this calculator
> http://www.rootshell.be/%7Eea3ghs/ctrafo.php to recalculate new values.
>
>
>> It is my understanding that you are able to vary the crystal oscillator
>> frequency somewhat with the variable resistor V1? Is this because D1 is
>> acting as a varactor responding to the shifting voltage on the base of Q1?
>>
>
> Yees!! This is a "second order" behavior not described anywhere. Works
> like a RIT.
> I think that we are changing the base-emitter capacitance.
>
> The diode is not working like a varicap. Don't put a varicap in //
> with C3 because QL fall down.
> The diode maintains the current of the xtal in the negative
> semiperiode of the oscillation.
>
>
>> Have you or anyone tried using VXO or maybe even Super-VXO with the
>> Pulga? Intuitively it seems like it would work and should still allow
>> the receive crystal filtering to function?
>>
>
> No, please report your experiences..
> In our tests, we connected the receiver to a computer with KGKSDR software
> and we received signals +-24khz...
>
>
>> Thank you for supplying the Spice models. The feedbackfilter.asc output
>> shows a peak at approximately 7.196 Mhz, whereas the pulga40g.asc shows
>> an output frequency of approximately 7.160 Mhz. Am I correct to assume
>> that the filter model is not related to the one used in the transmitter?
>>
>
> Well... it's very difficult to test the circuit in receive mode.
> I think that the uncontrolled capacitances changes the resonant frequencies.
> We went directly to breadboard to test it in RX mode.
>
>
>> I have some 2SC2988 transistors on order and will build the 40M version
>>
>
> You can get more power, but the reception with 2SC2988 is worse than 2N3866.
>
>
>> as soon as they arrive. It is nice to live remotely in the British
>> Columbia mountains, as I do, but I sometimes miss being able to go to a
>> local store and pick up that one component I forgot to order in my
>> Internet shopping!
>>
>
> OK, if you have broadcasting stations near, you will need a high-pass filter to
> attenuate MW stations.
>
>
>> Thank you for this intriguing design.
>>
>
>
>> 72
>> Stephen
>> VE7NSD
>>
>
> 73/DX
> Joan and Eduardo
>
> ------------------------------------
>
> Yahoo! Groups Links
>
>
>
>
>

--
"You never change things by fighting the existing reality.
To change something, build a new model that makes the
existing model obsolete." -- Buckminster Fuller

"First they ignore you, then they ridicule you, then they
fight you, then you win" -- Gandhi
2493 2008-12-12 08:11:52 Kenneth Ketner Re: EA3FXF & EA3GHS and the Pulga
Greetings to Pulga XCVR inventors in Spain, and those on the list who
love experimenting with electrons. I have been having fun with the
circuit. Thanks for sharing it.

I have built a Cubic Incher/Pulga TX manhattan style, and following the
derivation of the RX from the original TX as explained in your article.
I first was able to get something like 2 watts from the TX alone. Then I
added a 30 K ohm rheostat between the emitter of Q1 and ground. From the
emitter I connected the filter mentioned in your article. After the
filter I connected a small high power audio amplifier block aslready
prepared. Then I injected a small signal at the crystal frequency into
the antenna socket and made alignment adjustments. If my test signal was
modulated with a 400 Hertz tone, I could hear the signal in the output
of my audio amp. But if I turned off the 400 Hz tone to use a bare CW
signal, I heard nothing but amplitude noise in the band. My conclusion
is that in RX mode, my junk Q1 is not oscillating.

Therefore, it seems critical to check the following things, perhaps.

I was using a junk transistor. Perhaps I need a more specific
transistor that has properties particularly well suited to this XCVR
circuit.
If this is a factor, please let me know.

Perhaps my C3 is not quite right. What is the approximate
Capacitance of C3 in a working circuit?

Perhaps my V1 is not right. What is the approximate Resistance from
Q1 emitter to ground in a working circuit?

If I am understanding the functioning of the circuit, the value of V1
Resistance is very critical. If V1 is out of the circuit because the
closed key shorts the emitter of Q1 to ground, Q1 oscillates strongly as
a TX.

But if V1 is between the emitter and ground, Q1 still oscillates a small
amount, and then the circuit operates as a direct conversion receiver in
a Colpitts oscillator format, almost like the old regenerative tube
receivers in which one tube acts as an oscillator, detector, mixer, and
amplifier. Perhaps this means that Armstrong was the inventor of the
Direct Conversion receiver.

I am not an engineer, but one who enjoys learning from experimenting
with circuits. It is a way to be in tune with nature.

73 to all,

--
Kenneth Laine Ketner, KA5ELD
Texas Tech University
2494 2008-12-12 09:51:04 Eduardo Alonso Re: EA3FXF & EA3GHS and the Pulga
2495 2008-12-12 11:28:32 Stephen Wandling Re: EA3FXF & EA3GHS and the Pulga
Eduardo,

Eduardo Alonso wrote:
> Nothing magic here.
I am sure that much magic is to be found in this and other simple circuits.

I am wondering if Kenneth can benefit from comparing the specifications
for the 2N3866, which you recommend as superior for reception, and the
2SC2988 with higher power (1.8W), but which you say is inferior for
reception.

[My parts order which included the 2SC2988, arrived with the 2SC2988 on
back order. I am now awaiting some 2N3866 which I am getting from eBay,
item number 160123299730]

I discovered the Yahoo Group site for the M0KGK SDR decoder, so I have
sorted out my sound card problems and can make some use of this
interesting software. I thought I had no interest in SDR, but can now
see it, at the least, as an interesting evaluation tool when used as you
have used it with your simple receivers. I have been inspired to begin
reading Gerald Youngblood's "A Software-Defined Radio for the Masses",
available online, and find it to be a gentle intro to SDR.

After being introduced to Quadrature sampling mixers and their resultant
"I" and "Q" outputs, I am curious if you make use of the KGKSDR's
abilities to integrate I & Q, when using it with the Pulga and other
simple circuits? Or, do you just use the simple audio output of your
receiver?

I'm not sure if you saw my post regarding my failure to get your
Capacitive Transformer Calculator to generate the capacitor values used
in the circuit in Figure 5 in the Sprat - Flea.doc. I am interested in
an 80 meter version of the Pulga and want to be sure I am using your
calculator correctly.

72 & 73
Stephen
VE7NSD

"Small is Beautiful"
2499 2008-12-12 22:27:30 Glen Leinweber Re: EA3FXF & EA3GHS and the Pulga
Kenneth,
From a quick look at the cubic-incher, I'd guess
that you're just about right - in receive mode its
operating as an oscillating regen. It is reasonable to
suggest that a regen, when pushed into oscillation
is a direct-conversion receiver. Such a circuit is
properly called an autodyne - a self-oscillating
mixer.
If you were to build one with a 5 MHz. crystal,
or a 10 MHz. crystal, and trim the frequency very
very close to WWV, this circuit has the capability
to self-lock, similar to a PLL (if WWV is strong
enough). A frequency counter tied into this
oscillator could be calibrated this way.
This class of operation is sometimes called
an "injection-locked oscillator". Any regen can
be set up similarly.
For a crystal, injection locking has a very very
narrow lock-in range - perhaps less than a hertz
wide. Getting it to lock, and being able to tell
when it is locked, would be tricky.
An LC oscillator (having lower Q) has
a wider lock-in range. My experience is that
a strong signal might tend to lock with about
+/- 100 Hz. lock-in range. Much easier to tell
when its locked (just by listening).
2511 2008-12-19 14:55:31 Stephen Wandling Re: EA3FXF & EA3GHS and the Pulga
Hi Kenneth,

Based on your questions below, I am unclear as to the circuit you are
using to build your Pulga. I have considered the circuit in the
Sprat-Flea.doc to be the most usable, but I have now found problems
with it, that may relate to your issues.

I have still not received the parts needed to build my Pulga, but I do
have the Spice model that Joan and Eduardo provided. I have been
trying to grasp how to use the capacitive transformer calculator
http://tinyurl.com/4howd5 and could not get it to confirm the values
for C1/C2.

In the process of working with the calculator and reviewing the
circuit in the Sprat-Flea.doc, I note that in Figure 5 in that
document, the values of C1/C2 are reversed. I have confirmed this in
the Spice model, and it will not oscillate unless the "upper"
capacitor, C1 is 56 pF and the lower is 270 pF.

The Sprat-Flea.doc can be found on Joan's page, from the link in the
Links section.

I am currently trying to sort out the bandpass filter (band stop??)
filter used at the output in Figure 5.

I would be interested to know how you are making out with your
experiments