Flying PigRig (2): Assembly

 The PigRig assembly is coming along.  I've installed resistors, toroid and binocular transformers, ICs and diodes.  Next up is the crystals. 

I need to figure out what I'm going to use as a case for this....

Flying PigRig (1): Assembly

 I have a kit that Mike bought that I started assembling today.

The PigRig is a single channel 40 meter device for club use.  The frequency is 7.0307 MHz.   5W output, and CW.  

This shows how far I've gotten on the board assemble so far: 

My eyes are that good anymore, so I have to use a magnifier.  Also a good pair of tweezers helps a lot. I sorted the groups of parts into a plastic parts box I got from Harbor Freight. 

The instructions appear to be pretty nice.  I've been putting in resistors all morning.  I soldered one resistor in the wrong place and had a hard time getting that problem fixed.  I wish I could find the Gerber file for the board so I could look up the location of the each component before installing.  As it is, it's a crowded board with many components and it takes a few minutes to find the location of each component to install it. 

20-Meter High Performance Direct Conversion Receiver (5)

 I simulated the single balanced mixer circuit is LTSPICE.  I wanted to see if the LO amplitude had a large effect on the mixer gain. 

• I was able to calculate models for the L1 to L5 input circuit toroids.  I lumped the tuning network into the gate drive transformer and found the tuning capacitance Ctune by trial and error. 
• According to the 2N2222 datasheet the small signal input impedance should be between 2Kohm a nd 8Kohm. I used 2 Kohm, shown as RQ3 in the circuit. 
• The audio transformer T1 is a Calectro D1-711.  I searched the Internet for a couple of hours trying to find the specs for it,  but eventually I gave up and pulled T1 off the board and measured its parameters with my LCR meter.  I was surprised that something the size of a small acorn could be 2.5 Henry, but that's what it was.  See the circuit for the parameters.  I measured the parameters at 1 kHz.  The transformer had significant series resistances, which are included in the model. 
• There is a 1:2 step up from the antenna across the gate driver transistor.  I modeled a 1 uV rms input signal and scaled it accordingly. 
• The LO is simulated at 14.000 Mhz, and the signal simulated at 14.001 MHz.  The baseband frequency is 1 kHz. 
• The conversion gain from 1 uV RF in to the baseband 1 kHz detected waveform across RQ3 is  12 dB.  Doug DeMaw, is his "QRP Notebook"  states that this kind of single balanced mixer usually gets about 10 dB of gain.  So 12 dB feels reasonable
• I simulated with LO amplitude of 2V rms, given in the magazine article, and 0.87 V rms as actually measured on the board.  There was only a very slight, barely noticeable drop in conversion gain. 

Base band output waveform is given below. 

20-Meter High Performance Direct Conversion Receiver (4)

Link to schematic

I simulated the oscillator in LTSPICE in order to get an idea of the expected voltage amplitudes.

The LTSPICE simulation schematic is shown below. 

The inductor is a tapped coil made from a T37-6 core. 12 turns for L1, 7 turns for L2.  The coils are coupled at unity using the K1... SPICE command line.  V1, the 9V power, is turned on using the 1 us ramp up.  I originally tried the simulation with a simple DC source for V1, but the initial solution wouldn't converge. 

After the transient oscillator startup, the plot of oscillator waveforms is shown below. 

The oscillator frequency is 14.88 MHz, so there is some stray capacitance in the actual circuit that isn't included in the simulation.  This stray capacitance would bring the frequency down into the 20 meter band.  

1. The top trace is Vtank, which is take at the C2.  The voltage is 16V pk-pk. 
2. The center trace shows the the JFET gate voltage.  This is about 13 volts pk-pk, and is limited at the high end by forward biasing of D1.  I haven't checked,  but D1 probably cuts in before the gate-source junction of J1, so D1 protect the transistor junction from being forward biased. The bottom voltage is limited by the cutoff voltage of the transistor. 
3. The bottom trace shows the  voltage at the JFET source, which is the output of the oscillator that is coupled into the mixer.  This voltage is 6V pk-pk, which works out to 2.1V rms. This rms voltage is given as 2V rms in the original article,  which indicates the simulation is realistic.  

As discussed in the previous post.  The actual voltage here in the unit I'm working on is 0.87 volts.   So the oscillator appears to be a sick little puppy.  I suspect that the JFET is damages or just a variant far from the normal characteristics for a MPS102...  The MPS102 is infamous for being wildly variable in its characteristics.  I will replace the transistor with a new one and it will probably improve. 

A cross plot of Id as a function of Vgs indicates gm is about 5 mmho for the simulated JFET.

I may try to figure a way to plot the drain characteristic of the actual transistor.  

20 meter High Performance Direct Conversion Receiver (3). Oscillator troubleshooting.

Link to schematic

Did some more diagnostic work on the this unit.  

Notes: 

1. The QST article says that the cathode of D2 should be at 9.1V.  Actual voltage is 8.9 V.  Good.
2. The QST article says that the C21 should be 2 V rms.  Actual is 0.87 V rms.  So the oscillator voltage is down significantly from what it should be.  I checked all the resistor values in the oscillator circuit and they are as expected. 
3. Removed C22 to see if the mixer was loading the oscillator somehow.  No evidence of that.  C22 was nominal 10 pF, and measured 9.6 pF.

There are other components to check in the oscillator circuit, but I suspect the transconductance, gm, of the Q4 is down and may be the source of the problem.  

I will order some MPF102 JFETs to see if replacing the transistor helps.  Perhaps the transistor was toasted while soldering. 

I will pull capacitor C19, C20, and C23 to check, and will pull L6 to check as well.

I am planning to do some simulations to check the dependence of the oscillator output on JFET gm.