The work continues on the infamous Radiokit debugging competition. The
pressure is on.
Started my test session today by reconfirming what I found yesterday with
the receiver. Still working OK. Don't know what the exact sensitivity is,
but it seems to be usable. Picks up a signal out of the air from the
generator set to -60 dBm. The frequency shifted a little overnight, which
is to be expected with the flying wires where the VFO cap should be. Won't
know what the stability will be until it is assembled in the box.
Learned a little more about the AGC oscillation. It seems to be very
dependent on the power supply and supply bypassing. Since this will change
radically when I remove the haywiring and put it into the case with the
regulator, I decided to leave it alone for now. I'll revisit the AGC when
things are cleaned up.
Started to work on the transmitter. Larry spent a lot of time preparing to
bring up the transmitter, but I don't know if he actually got around to
key-down smoke releasing.
OK. Power on. Key down.
Nada.
Look for the obvious. Yep, no power to the driver or PA. Why not? Trace
it out. Of course -- gotta connect the other wire! As my High School
Thermodynamics instructor always reminded us, "Do you have kinetic energy?
Do you have heat energy? Do you have electrical energy? Of course not!
It ain't plugged in!"
OK. Power on. Key down.
No smoke. No output. That's good. If there's no smoke, there's hope.
Checked the mixer output with the scope. The 4.95 MHz VFO gets mixed with
a 12 MHz xtal osc in an NE602 to produce the 7.050 transmit frequency and
the 16.95 MHz image. Remember that.
Looks OK. Peak up the trimmer. Feels right -- got a good peaking
response.
Just for the heck of it, I checked the current through the driver (2N2222)
by measuring the DC voltage across the emitter resistor, whose value I
noted was 15 ohms. 670 mV across 15 ohms is 44 mA. Too much! That puppy
must be getting a little warm. I'll just check it with my finger.. OUCH!
Key up. Power off. That was close; almost let the smoke out.
15 ohms....15 ohms... hmmmm. I recalled having read something about that
in the notes. Aha! "If using a 2N2222, use a 47 ohm emitter resistor."
So I changed it. Measured the quiescent collector current at the emitter
again. 14 mA. Much better! Checked the bias voltages. Looks reasonable.
I noticed when measuring the emitter voltage with the scope that there was
no RF signal superimposed on the DC. That's good. The driver uses an
emitter bypass cap so there is no degenerative feedback at RF. This points
out a good debugging technique. As you go along, confirm that everything
is consistent with the way the circuit should work. Check signal values.
Even if you get what you think is a good output, things still may not be
working the way the designer intended, and you may have an unpleasant
surprise later. (Repeat after me: Unpleasant surprise later... unpleasant
surprise later...)
Peaked the tuned circuit on the collector side. Seemed to peak OK, but I
only got 900 mV on the base of the PA. That's not even close for a class C
amp. Read Larry's notes. He calculated the resonant value of the tuned
circuit and figured that it wouldn't work without an extra cap, which he
soldered on the back of the board. Hmmm. Tried removing it. Now I can
peak it up to 5 Vpp to the PA. The power supply draw is now 250 mA (up
from 40 mA on receive) and the PA transistor gets comfortably warm after a
few seconds without a heat sink. Should be more that 250 mA at full power,
though. Checked the output level at the dummy load with the scope.
Nothing. Hmmm.
Look at the drive signal with the scope. Play with the tuning. Feels
right, but something is still wrong. Turn on the cursors to read out the
frequency. You idiot! (Me, not you) I peaked things up on the transmit
image frequency of 16.950 MHz! That's why it worked better with Larry's
cap removed. Larry, I never should have doubted you. (...unpleasant
surprise later...)
Back to the mixer. Can't peak it for 7 MHz. Values must be off here, too.
I think Larry had a suggestion about that in his notes, also.
Tacked a 100 pf cap across the mixer output circuit. Now it peaks up just
fine on 7MHz. Retune all the stages after putting Larry's collector cap
back in.
WE HAVE IGNITION! Got a beautiful 37 Vpp sine wave at 7 MHz on the dummy
load. That works out to 3.35 w. Power supply current is consistent.
Supply voltage is 12 volts.
Measured the frequency with the Radio Shack counter. Hmmm. About 20 kHz
off from where the receiver is receiving. Since the same VFO is used for
transmit and receive, this can only happen in one place -- the 12 MHz xtal
oscillator in the transmit mixer NE602. Tweaked the trimmer. Wow, this is
sensitive. It has a surprisingly wide frequency range for a trimmed xtal
oscillator, so it is hard to set exactly right, but I got it within 200 Hz
of where it should be. Aha, Now we have sidetone! The sidetone is
actually the receiver hearing its own transmitter.
Got the BFO, transmit offset, and filters all working within 200 Hz of 700
Hz. Good enough for useful work. Runs on lower sideband, single signal.
Raised the power supply to 13.8 volts and measured 42.8 Vpp into 50 ohms.
That works out to 4.6 W. I can lower that smoothly with the drive control
pot. Signal looks clean on the scope. Don't have a spectrum analyzer to
check the harmonics, but I don't see anything on the scope to worry about.
Draws 670 mA from the power supply. Just right.
This rig is starting to look like it might work pretty well. The main
problem so far seems to be inadequate documentation, especially concerning
changes necessary to put it on the selected band.
What next? All the circuits have been checked and seem to be doing
something useful on the bench, albeit with a few minor problems. Maybe
it's time to clean it up and put it in the box?
What do you think?
Mike K1MG
[
QRP-L Archive |
]
[
1993 |
1994 |
1995 |
1996 |
1997 |
1998 |
1999 |
2000
]