Hello MICRO1700
You may be the one to steer this project back on course. Since I was unable to modulate audio at the collector of Q2, I tried injecting at the emitter, as I've seen some circuits that do this. No result with a matching transformer. No result with direct injection from the audio amp without a transformer.
Reading the Pixie2 Document as it is, it sounds to me like the author expected the circuit to have a modulation problem, and we can deduce from this that the author never actually got a working version, especially as all of his remedy suggestions are speculative and evidently not based on actual results.
My big fat book from The American Radio Relay League only shows tube circuits and no solid state.
Also, the solder smoke really gets to me so I'm going to put everything in a box for a while and do nothing but think.
Good luck!
Carl,
I don't want to pull you too far off track but I looked at the Pixie2 circuit and was struck by a few things. ( http://www.indianapolis.net/QRPp-I/talking_pixie2.html)
R3 at first glance appears to bias Q2 but since L1 shorts the base to ground to DC, R3 serves no purpose.
Since the base of Q2 is at DC ground, it is most likely operating Class C which also calls into question the need for R4 unless its function is to prevent over current in Q2. R4 will also reduce the gain of the stage. Class C RF amps usually use a tuned tank in the collector circuit (cap in parallel with L2) with loop coupling to the output. This would be a rather simple thing to try. Calculate the L2 and C variable in parallel needed to resonate at your frequency and tune for peak output. A RF bypass cap. will need to be added from the junction of L2 and T1 to ground. It should be large enough to short RF but small enough to pass audio.
I mentioned earlier the rather poor results in my electronic comm. class lab where a similar circuit was used. I think a more traditional Class C transformer modulated circuit would be an improvement.
Neil
Neil - Your comments are very helpful and point out some new things to try.
Here's a big question about L1. Since it's grounding the base of Q2, is Q2 able to amplify or or even pass RF from Q1?
In the section on "Variables" in the Pixie2 docs it suggests replacing L1 with a 5k variable resistor, which in light of your comments seems worth doing.
On a related note, I've got the modulation amp (Velleman K4001) beautifully balanced for the typical input from "consumer" lines, in this case a C.Crane Radio Plus tuned to the STL at 101.9. Such outputs tend to be -10dB, 1k. At the input is a 200 ohm variable feeding a 600 ohm to 50k ohm potted transformer, terminated into a 1k resistor and driving the input. The amps output is series fed across a 10 ohm resistor into the modulation transformer (8 ohms/1 k) as seen in the Pixie2 illustration (Radio Shack part). The range of control is smooth and continuous from very low level to the maximum headroom of the device.
Back to work.
Hi Carl and Neil and everybody!
I think R4 is a current limiting resister
that limits the amount of energy that
can go through Q2.
I like the idea of trying a different way
of coupling the RF to the antenna.
I don't know much...
But I know one thing. There are a lot
of RF designs out there that aren't quite
right but work anyway.
I think that, together, we will figure this out.
(I really like being part of this board, it is
a lot of fun!)
Best wishes to all!
Bruce, MICRO1690/1700
P.S. If Q2 was made to run at a lower power,
would it be easier to modulate?
Yes, MICRO1700, this board makes hobby experimentation fun all thanks to SCWIS and the other owners of this site. I also think that SCWIS, based on emails of his, enjoys it when his members get involved in very active threads like this one. Therefore, part15.us is a socially healthy website.
Now, I will test the idea posed by MICRO1700 of running Q2 at lower power, and my question is, do I need simply to increase the value of R4, say, to 100 ohms or higher?
Carl asked: "Here's a big question about L1. Since it's grounding the base of Q2, is Q2 able to amplify or or even pass RF from Q1?"
Yes, remember that an inductor presents a high impedance to RF so as far as the signal is concerned, it is not there.
Try placing a diode (1N914, 1N4148, or some other small signal diode from the base of Q2 to ground with the diode cathode (banded end) connected to the base. If this is intended to be a class C circuit then the idea is to have the transistor saturate on a small part of the signal cycle. The diode forms a voltage clamp which automatically makes this happen. Without the diode then the reverse base to emitter diode junction does this but the voltage is probably too high to be effective.
I suspect that this circuit did not start out to be Class C and perhaps I am bending it too much. If you use the pot in place of L1 then it will probably operate Class A which may have been the author's original intent.
Neil
Hi everybody!
I have the oscillator section of the transmitter
running on a protoboard at 13.560 MHz. Again,
because of eye problems this stuff takes a while
but I am getting there.
Best wishes to all!
Bruce MICRO1690/1700
P.S. I am using 2N4401s instead of
2N2222s because that's what was
available. More to follow...
Hey Good work MICRO 1690/1700!
And your mention of using 2N4401 came at the perfect time because I blew my Q2!
When first starting out Q2, 2N2222A, checked out fine with my "Lil Bitty Tester" so when I began having failures I eventuall found Q2 had blown out, for reasons unknown.
The circuit lit also suggests 2N3866 and 2N7000 (FET), but in my spare parts I have found many other numbers which I need to explore in case any of them will work.
I also found a metal "hat" that I should have added to Q2 for heat dissipation, but it's too late now.
And they're neck and neck, MICRO1700 taking the lead!
Hi Everybody!
Hey, Carl, this is fun!
Here's a question about Q2 destroying itself.
First of all, I know you are a really experienced
broadcast engineer. So I'll just ask this:
Did your dummy load become disconnected?
If Q2 is blown, then the transmitter wouldn't
modulate at all. (Is that right?)
Best wishes,
Bruce, MICRO1690/1700
Dear MICRO1700:
I'm not able to re-construct a logical theory as to why Q2 blew, because I don't know at what point along the way it actually blew. I may have been testing for days with a blown Q2. The 50 ohm dummy load was on, then I tried a long-wire antenna, and at the early stage I got strong modulation with sine wave tones but very distorted modulation with speech. Then there was no modulation. That's probably when the Q2 blew, but I kept trying things like a dunce, such as changing R4 to a higher value, changing L1 to a 5K variable (as mentioned in the doc), and about 5 different modulation transformers.
I was operating on a dead patient! Now I have to repeat a few of those experiments with a good Q2. I wish I was the great engineer you were kind enough to attribute.
Carl,
As you know once the smoke is let out of a part it is not possible to put it back in.
Please take some friendly advice. Clean off your desk and get that oscilloscope set up and have at it. At least you will know when the patient dies. Even if you don't know exactly what you are seeing you will know when it changes.
Neil
Dear Neil:
The only way to save face is to get that scope hooked up. And I have a question about your suggestion of 11-19... try a diode 1N914 or 1N4148 from Q2 base to ground...
well I found a 1N34AS... but my main question is, should the diode parallel L1 or replace L1 ?
Hi Carl!
You are still the great engineer that you
always were! You're just multitasking
like everybody else.
I'm going to buy a few more parts tomorrow
(don't tell my wife) and get that second
stage running.
Best Regards!
Bruce, MICRO1690/1700
Hey Guys I just thought of one more
thing!
We really don't even know what the
duty cycle of this circuit is. Even with
everything running perfectly,
(good output power, correct dummy
load, good modulation) Q2 may just
get too hot after a while and decide
to destroy itself. If that happens after
all other problems are solved, well,
we can just figure out how to lower
the output power.
As we know, the original Pixie 2 was
meant to be a morse code transmitter,
and in that mode, the carrier is on way
less than 100%
of the time. So if the thing is running
at 100% duty cycle at 100% modulation,
at the design power level, Q2 may just
blow anyway.
I am really enjoying thinking about this stuff.
Bruce, MICRO1690/1700
P.S. Of course I haven't looked up the specs
on this kind of transistor, but I know you guys
know what I mean.
Carl and Bruce,
Love your questions and am trying to give useful help. At least I hope so so keep the ???s coming.
Carl,
The 1N34 diodes are germanium point contact devices and are quite puny when it comes to being punched with peak currents as in this application. You can always try one but I don't recommend it. The diode should be in parallel with the the inductor. I will be happy to explain this in great detail but not right now just to keep this reply manageable.
Bruce,
You have raised a crucial issue with this circuit with your question about the power dissipation in Q2 and the duty cycle. In a Class C amplifier the duty cycle is 10 percent or less and even so since the transistor is either full off (0 amps, lotsa volts, power dissipated = 0 X lotsa = 0) or lotsa amps and O volts (lotsa amps X 0 Volts = 0 watts) then this is not an issue. I over simplify since there is always a Vce(sat) which is not zero but if Q2 is popping then this thing is not operating Class C. I have built RF amps with 40 watts output with minimal heatsinking and the thumb thermometer on the final can't even tell it is running.
I fear my suggestions are not appropriate for this particular circuit and maybe you should not use them. It appears that both of you have the oscillator portion working but the failure is in the final stage. Here's my bottom line on this: The final stage needs to be operated class C with the modulation transformer in series with the collector and bypassed to RF. I see no reason that you should be having problems with the final Q2 overheating unless something is wrong with the biasing or perhaps it popped because the transformer secondary voltage was too high.
You two have what it takes to get this working and when you do your feedback will be valuable. Keep going and I will try to help when and if I can.
Neil