Yeas, thank you John WDCX, it will be necessary to be as exact as possible before calling this project ready for market, so to speak.
Right for now, would you agree with me that having changed R4 from 10 ohms to 330 ohms the output power has been reduced?
My 10-volt AC meter shows 1.9 volts at the Q2 output following C5, but after the low pass filter it only reads 1/10-of-a-volt going into the antenna
The previous output readings were made with a supply voltage of 12 VDC. Although previously I dropped to 9 VDC at the time Q2 emitter was directly grounded as a way of avoiding over-heating Q2, now with 330-ohm on Q2 emitter the circuit easily handles 12 VDC.
But I just now went to 6-VDC and the following was observed: the modulation from the Sangean receiver became much louder, (an inverse relation between power and modulation level is also observed in the Ramsey AM25) and the AC reading at C5 is now 1/5-of-a-volt and at the output of the low-pass filter the reading is negligible.
In the opinion of the readers, is the voltage supply a fair way of adjusting the output power, or do you believe that it should be done with an attenuator circuit (if needed)?
Ok. I have transferred the oscillator from
the protoboard to the perfboard. It has
been soldered together and it is running.
Carl, I am not an expert but referring to
your last comment about going down to
six volts and getting more modulation-
I guess you have to find a way to get
100% modulation (if possible). Then you can
go over to the RF output section, and
measure the power out into a dummy load
and calculate how much attenuation you need.
After that I would work on the RF output
filtering. But I'm not an expert. You are way
ahead of me on this thing.
Best wishes to you and everybody out there
Bruce, MICRO1690/1700
MICRO 1700
Even though lowering the voltage makes the modulation become louder, the modulation can also be made louder by turning up the audio input level, so the main focus of the DC voltage is to arrive at the correct output power.
I tested the radio reception at 100-feet with 6 VDC and it was very poor, so I think that's too little power. I dared to go up to 15 VDC for the first time, and the 100-foot reception was better, but as far down as 9 VDC it's also about the same, so that's where it is now.
What if we discover the output is well below what the FCC allows? Then we'll need to boost it instead of attenuating it.
Hi Carl!
Your modulation tests are interesting.
I have to think about that.
Right now, I plan to do everything you
have done. When I get the transmitter
done, I plan to try it with the modulation
transformer and your mods that made
that configuration work.
I'll measure the power out at that point
and see where things are.
I have a watt meter that measures down
roughly to 5 or 10 milliwatts at 50 ohms.
That should be good enough for now.
Here's my question:
What did you use for L3? That's one part
I don't have yet.
Bruce, MICRO1690/1700
MICRO1700
Your question about L3 made me find something I hadn't noticed, even though it's in plain sight. I don't have your visual handicap, but there are many things I don't see.
In his posting on 9/18 SCWIS placed a reference to a 13 MHz 5-pole Lowpass Filter with 50-ohms in/out. Unlike the L3 in the schematic, this filter has 3-inductors rather than one. So my answer is, for L3 I got all three of the inductors: 1 uH/1.5 uH/1 uH, all from jameco.com
I do not know what L3 would be in the version shown in the Pixie2 Diagram.
You said you were going to think about the modulation issue, and I also have been thinking about it, considering the bits and pieces of what I know. Amplitude modulation is achieved by a form of mixing, similar to audio mixing, but in this case the carrier is mixed with the audio until their proportions are correct (100% modulation). I guess that increasing the carrier strength probably requires a corresponding increase in audio. When the carrier is reduced, the audio has less signal to mix with, and becomes louder because it is correspondingly stronger than the carrier. Like I said, that's a guess.
Hi Carl!
I'm going to set the modulation question
aside for a moment.
I bought a Pixie 2 kit from a vendor that
I can't even remember right now, but they
sell the whole tranceiver. So that's where
I got all the parts. The circuit board was tiny-
too small for my eyes. So I'm using the perf-
board instead.
Now here's my point-
the kit is made to operate on two different
ham radio bands- 80 meters and 40 meters.
It is supposed to deliver several hundred
milliwatts of output. After C5, a .01 uf cap
that isolates the output circuit from the
power supply (I think), there
is what I think is a pi network that matches
Q2's impedance to the 50 ohm load.
It uses two 820 pF caps. For 80 meters L3
is a 2.2 uH coil. For 40 meters L3 is a 1.0 uH
coil. For 13 MHz, L3 has to be a different
(smaller) value. I think that Pi network has
to be in place for this thing to work. Then after
that, you can place the low pass filter.
So L3 has to be around half of 1.0 uH,
because 13 MHz is almost twice of the fourty
meter band, which is 7 MHz.
I'm typing this as fast as I can, because I have
to go in a minute. I don't know if I am
making myself clear, but I think you have
to have the correct Pi network, and THEN the
low pass filter, or the thing won't match to
50 ohms, which is what we all want it to do.
Am I making sense? Or am I not understanding
what you have done?
Bruce, MICRO1690/1700
Carl, this is getting complicated.
I just remembered that Neil mentioned
a different kind of output coupling after
Q2, a tuned circuit, which would transfer
the RF into a 50 ohm load. I think he
said that Q2 would have an easier time
operating Class C, which is what we want.
If we are modulating Q2, I think it has
to be operating in Class C, although I
could be wrong.
Unfortunately, I have to go, but that's
another item we can think about.
Best Regards
Bruce, MICRO 1690/1700
It's good, MICRO1700, that you have access to different information than I have, because it can serve to throw additional possibilities into the design process, even though it is complicated.
With the direct modulation circuit changes I've made I ask you, Neil, to comment how close this puts us to a Class C operation?
The pi-network PRIOR to the filter confuses me, because I thought the filter WAS or IS a pi-network (?)
IN REVIEW
Two main unknowns from my point of view:
1. How to design the output to maintain 50-ohms and provide needed harmonic filtering;
2. Determine whether the power output is above, at, or below FCC specs for Part 15.
Hi Carl!
I'll be back later with more but here's a few brief
comments.
I bought my Pixie 2 kit from HSC electronics.
If we didn't have to worry about amplitude modulation
and wanted to operate in the 3.5 or 7 MHz bands
we would be all set. The values for the Pi network
I mentioned above are for the parts that are
actually supplied with my kit. With these parts
the builder can get several hundred milliwatts of
RF into a 50 ohm load on 80 for 40 meters. The
caps in the Pi Net are the same for both bands,
but the inductor isn't. The caps are both 820 pF.
For 80 meters, the inductor is 2.2 uH. But for
40 meters, the inductor is 1.0 uH. Knowing
this, we can sort of figure out what to use for
13 MHz, I guess. Assuming we figure that out,
then if we had say, 200 milliwatts of output
at 13 MHz (which is the rough output power
for the thing) then to get to the Part 15.225
field strengh ballpark, we would need about
2 milliwatts of RF out (or maybe a little less)
but anyhow a 20DB pad would bring the output
down to that level. This is assuming that the
transmitting antenna is a dipole. The actual
output probably should be less, but I am
just using these numbers to make the example
easier for right now. Most people who are
on the 13 MHz Part 15 band have said that
the RF out should be somewhere between
0.5 millwatts to about 2 milliwatts into a
dipole to be compliant with the Part 15.225
field strength regs.
Then we could put your circuit on the end if
we wanted to, that is, the low pass filter that
you are using, which I THINK is a low pass filter,
but I could be wrong - but I'm going to assume
for a second that I'm right and that the low
pass filter is 50 ohms input and 50 ohms out.
Then,, assuming all the above is working right,
we would have 2 clean milliwatts into a 50
ohm load. Some people have mentioned
that the output power is so low that we
don't need the low pass filter. If the Pi net
is working right, maybe the output would
be cleaner than you have experienced, I
don't know.
After that, there is the modulation problem,
which I think you have either already solved
or at least you are close to solving it.
I hope I'm making sense of this thing.
I'll be back later. I do think that this is a
lot of fun.
Best Regards
Bruce, MICRO1690/1700
Carl,
I want to acknowledge that you asked "With the direct modulation circuit changes I've made I ask you, Neil, to comment how close this puts us to a Class C operation?" and let you know I am interested in what you are doing but right now some personal things are distracting me from spending much time and thought to the board (temporary, I hope).
I will draw out your circuit from your description and try to make a better answer but just off the top of my head I think your circuit is operating either Class A or Class AB. I know the class makes little difference to what you observe and I only mention it in case you want to look them up and read more about the topic. It essentially depends upon how the circuit is biased.
Hope to be back soon, meanwhile, carry on.
Neil
It's a good thing you are around. Hope
everything is OK.
Bruce, MICRO1690/1700
I'll review my literature regarding the amplifier classes, as I didn't take away with me what was found the first time I looked in the books.
Today I wandered outdoors with a Grundig in hand and now I'm convinced that the present circuit, although the modulation is perfect, is very weak because of the emitter of Q2 being connected to ground through a 330-ohm resistor. The transmitter this circuit came from was a 100 mW part 15 am transmitter which had three transistor stages, and I think that middle stage, which would be a "driver," is probably needed in this version of the Pixie2, to compensate for the lowered amplification of Q2.
The original circuit, with its modulation problems, had a far stronger signal outdoors. I may return to the working model with the transformer modulation and a grounded Q2 emitter. This first shortwave transmitter is stuck with two transistors because the circuit board is cut to size.
Transmitter # 2, later on, may have three stages.
Also, regarding power, I love the idea of a power control built into the circuit for more flexibility than a fixed- attenuator. Built-in power controls are found in the Rangemaster 1000 (as I understand it) and the Ramsey FM25B.
On the Pixie2 I think a variable-resistor on the emitter could allow for power control, but comments on the idea are welcome.
Replying to myself, I returned to a previous version of the experimental Pixie2, based on my realization that the "direct modulation (transformerless)" version had noticably weak RF.
Once again we are transformer modulating with Q2 emitter directly grounded. Indeed the outdoor field strength on the Grundig over a 200' walk around is substantially stronger.
There is a difference in tonight's version over the previous version, and that has to do with how the base of Q2 is configured. Originally the schematic diagram was followed exactly, with 33k R3 to base and L1 from base to ground. But L1 prevented modulation from happening, so it was removed and not replaced with anything.
Then, for the direct modulation method, borrowing an idea from another circuit, we replaced R3 with 10k and replaced L1 with 10k. This arrangement has been left in place.
I do not have enough experience with transistor base wiring to have any notion of what difference might exist between these two ways of configuring Q2.
What I do know is that now the output power is strong enough to worry about, so that needs to be the next area for analysis. Otherwise, this is the "keeper" circuit.
I think now we're getting some meaningful numbers.
First, looking back in time to the recent direct-modulation (transformerless) version of Pixie2, output at C5 coming off of Q2 was about 1-VAC and the output after the filter was between negligible and 1/2-VAC, looking into 50 ohms with 9 VDC B+
Now, see recent post, back to transformer modulation, Q2 emitter grounded, definite strong signal out in the field (using 1/4 wavelength indoor antenna, within 200' on Grundig radio, look at these numbers:
6-VDC B+, 5-VAC OUTPUT AT C5, 2-VAC FILTER OUTPUT;
9-VDC B+, 6-VAC OUTPUT AT C5, 1.9-VAC FILTER OUPUT.
Looking into 50-ohm dummy load.