This page updated March 26, 2015.

Material that was here previously will reappear, so be patient.

I bought a TenTec Paragon 585 transceiver at an auction for a ridiculously low price. It has low output and I am going to refurbish it.

Here is the K7QO Code Course 3.0 in ISO format. Enjoy. Feel free to use and give away. Teach a class and make me proud. As of Jan 2015, more than 16,000 copies have been given away worldwide.

Here are some photos to show the progress and steps taken to fix a TenTec Paragon 585, circa 1985. I got this from an auction for $150 US dollars. As you can see from the photos, it is in excellent shape. I am going to make it better or at least try. Not failure, but low aim is crime. Since I have very little invested in the rig, I am not willing to spend $110 an hour and the risk of shipping across the country to have it looked at. If I can't repair it, then I am not as educated as I think I am. Educated beyond my intelligence level and spent a lot of money on books. :-) I do love a challenge.

Click on thumbnails to get a larger image.

Photo 01 |
Photo 02 |
Photo 03 |

Photo 1: Front of Paragon 585 with cover removed. Photo 2: Top View of the insides. Photo 3: Back Left section of transceiver.

Photo 04 |
Photo 05 |
Photo 06 |
Photo 07 |

Photo 4: 81340 board. Photo 5: Scratches on top cover. Photo 6: 500Hz 6.3MHz IF Crystal Filter Photo 7: Restored top cover.

Photo 08 |
Photo 09 |
Photo 10 |
Photo 11 |

Photo 12 |
Photo 13 |
Photo 13 |

Photo 8: P/S and Cover. Photo 9: P/S side view. Yikes. Water soaked at one time. Katrina? Photo 10: Rectifier board. Photo 11: Speaker. Corrosion definitely due to exposure to water. Photo 12: Crowbar circuit board. Definitely a layer of some crud. Photo 13: Power transformer with serious corrosion. But, the darn things works just fine. Photo 14: PCB cleaned with CRC Contact Cleaner from Home Depot electrical department area.

Had I known the shape of the power supply, I would have never plugged the rig in and turned it on. I assumed, and incorrectly so, that because the inside of the Paragon was clean that the power supply would be in the same shape and thus I plugged it in. Even worked a couple of stations on 20m at 11W output. After listening in another receiver to make sure the signal was clean.

I'm invoking the DFW rule. Don't Fool With it. It is working, the PLL stays locked, so damn the torpedoes and full speed ahead..... March 20, 2015. Hell of way to start Spring.

In 1961 I started at McMurry College working on degrees in physics and mathematics. The first day I went into the office of Dr Virgil E Bottom, chair of the department of physics, and was hired immediately as a research assistant to work on a quartz crystal research project that was funded by James Knight, HP and Collins Radio. This was due to the fact that I knew how to grind quartz crystals, having done so as a Novice to get World War II surplus FT-243 crystals moved from outside the ham bands into the novice bands. This was at the time (1957) when novices were required to be crystal controlled.

I had to derive two sets of equations for the resonant modes quartz crystal unit for resonant frequencies. One set of two equations for the crystal itself and one set for the crystal in series with a capacitor.

The equivalent circuit for a quartz crystal unit is:

The R, L and C are the motional parameters for the quartz crystal unit. There are no inductor, capacitor or resistors in the quartz crystal unit (except for C_0). The piezoelectric effect and resonance makes the system act like there is. I will use the subscript m to indicate the motional parameters in the following equations.

I set up the equation for the impedance between points A and B, crunched the numbers and then solved a quadratic equation for the two roots for resonance. I obtained two roots.

\omega_s = \sqrt{\frac{1}{L_m C_m}}

where

\omega_s = 2 \pi f_s

for the series resonant frequency and
\omega_a = \sqrt{\frac{1}{L_m C_m}+\frac{1}{L_m C_0}}

for the parallel resonant or anti-resonant frequency. Note that this frequency is always higher than the series resonant frequency.

I will come back later, after a paper I submitted gets rejected or accepted, and show all the math, for the inquiring minds in the Universe. In the meantime, you can try to derive the equations yourself.

Now. For the circuit:

the impedance equation yields two roots

\omega_c = \sqrt{\frac{1}{L_m C_m}+\frac{1}{L_m (C_0 + C_x)}}

and

\omega_a = \sqrt{\frac{1}{L_m C_m}+\frac{1}{L_m C_0}}

The subscript c is to show that the equation is for the case with the crystal in series with a capacitor. You will also note that the parallel resonant frequency did not change and is not affected by C_x. Once again note that the series resonant frequency is lower than the parallel resonant frequency.

It is also important to note that C_0 \mbox{ and } C_x MUST include stray capacitance in the fixture. This was not an issue in the true series resonant frequency for the first case without the series cap.

Here is the schematic for my fixture.

Here is a photograph of the fixture I built using the muppet board PCB technique and showing the input cable from a NorCal FCC-2 DDS RF generator and an output cable going to an HP RF voltmeter.

The values for R3 and R4 should be around 10 ohms or slightly lower. It is not critical. If you make them too low you will require an extra sensitive RF voltmeter. With 0.5V input from the FCC-2, the output voltage at resonance is on the order of 15mV at the peak. If you make R3 and R4 larger, you will broaden the peak and increase the voltage and make it more difficult to find the resonant point within 1 Hz. BTW, you must do this with a frequency counter or DDS with 1Hz resolution to get accurate results. My Cx value was around 47pF. The fixture has an approximate input and output impedance of 50 ohms dependent upon the Rm of the crystal at resonance. Once again, not critical.

Too large of a value for R3 and R4 will give you a value for Rm that is going to be higher than the actual value of the crystal due to the loading factor. Also, the lower R3 and R4 values will swamp additional distributed capacitance of the fixture.

Note that I used an Augat machined socket for the crystal. I soldered the center pin to ground to reduce the distributed capacitance of the socket. I use 0.10" headers for the shorting of Cx and a Berg connector. I even cut a portion of the header off to reduce the distributed capacitance. The cap is a NPO mono at 47pF.

It is important to use a good L/C meter and measure the capacitance across the socket and across the jumper without the cap before building the rest of the circuit. Accurately measure Cx before installing. This will determine just how well your results are going to be

OK. How do we do the measurements and the calculations?

Now you are ready to do some math. You have f_s and f_c. You have C0 and Cx and you have the stray capacitance across the crystal socket and the jumper. Calculate C_t as the sum of all four cap values. Plug into the following equation.

L_m = \frac{1}{4 \pi^2 (f_c + f_s) (f_c - f_s) C_t}

OK, now plug and play. Note. A lot of people may have a formula that looks similar to this. If they have a term that looks like:

2 \cdot \Delta f \cdot f

Then they made an approximation to the difference of two large square numbers. That kills off some of the precision and accuracy in the calculation. Use my formula. Also, use some mathematical software like Mathematica, Maple, Wolfram Alpha or other software to do the calculation. You need at least 10 or more significant digits to get good results, even though the results are only good to 4 or 5 places depending upon how well you can measure capacitance.

You get Cm from the first formula

C_m = \frac{1}{4 \pi^2 f_s^2 L_m}

and you get the Q from the formula

C_m = \frac{2 \pi f_s L_m}{R_m}

and there you have it.

OK. I know you doubt my sanity. So, to prove to you that this stuff works. I sent off a set of crystals to the great state of Colorado months ago. Tom Thomson, W0IVJ, and Larry Benko, W0QE, measured the parameters for the crystals using their own AIM 4170 VNAs. Here are their results.

Lab Crystal FSeries FParallel Rs Ls(mH) Cs(pf) Cp(pf) Qs Measuring Tech Number Instrument W0IVJ 1 3.578426 3.585154 49.822 139.418 0.0141885 3.780 65801 AIM 4170 VNA W0QE 1 3.578427 3.585256 49.700 142.449 0.0138866 3.638 64443 AIM 4170 VNA W0IVJ 2 4.193154 4.200966 16.943 111.185 0.0129572 3.484 180122 AIM 4170 VNA W0QE 2 4.193163 4.200894 17.159 115.719 0.0124495 3.376 177674 AIM 4170 VNA W0IVJ 3 4.031548 4.036547 40.962 309.640 0.0050332 2.032 211124 AIM 4170 VNA W0QE 3 4.031552 4.036428 40.046 340.051 0.0045830 1.895 215101 AIM 4170 VNA W0IVJ 4 4.193152 4.201202 18.176 107.122 0.0134487 3.509 165524 AIM 4170 VNA W0QE 4 4.193157 4.201100 18.432 112.633 0.0127907 3.376 160993 AIM 4170 VNA W0IVJ 5 4.094814 4.102963 23.238 134.404 0.0112398 2.830 147508 AIM 4170 VNA W0QE 5 4.094819 4.103052 23.469 134.440 0.0112368 2.794 147386 AIM 4170 VNA W0IVJ 6 3.998939 4.005005 22.484 132.716 0.0119351 3.940 153331 AIM 4170 VNA W0QE 6 3.998953 4.005015 22.619 136.166 0.0116326 3.837 151261 AIM 4170 VNA W0IVJ 7 11.055203 11.079818 7.407 11.640 0.0178059 4.007 109648 AIM 4170 VNA W0QE 7 11.055211 11.079788 7.337 11.755 0.0176319 3.965 111282 AIM 4170 VNA W0IVJ 8 4.094873 4.102956 24.034 130.270 0.0115962 2.943 144651 AIM 4170 VNA W0QE 8 4.094876 4.103023 24.476 134.745 0.0112110 2.818 141641 AIM 4170 VNA W0IVJ 9 13.499968 13.529170 4.063 5.074 0.0273900 6.345 100390 AIM 4170 VNA W0QE 9 13.499973 13.529200 4.129 5.064 0.0274465 6.339 104041 AIM 4170 VNA Crystal Identification Printed on Each Unit 1 HC-49U MPCO 3.579545 2. HC-49U HOSONIC 4.1943 B603 3. HC-49S 4.032 4. HC-49U HOSONIC 4.1943 B603 5. HC-49U MMD A18BA1 4.096JHz 9942G 6. HC-49U ABRACON 4.000 AB 0443 7. HC-49U FOX115-20 11.0592 8. HC-49U MMD A18BA1 4.096MHz 9. HC-49U 78941-1 13.500 KDS 5K

This morning, March 23rd, 2015, I took the above fixture and measured crystal 1-6 and here is what I got.

Nr f_s f_c C0 Lm Cm Rm Q 1 3573435 3578898 3.65pF 144.38mH 13.70fF 47.1 68,900 2 4193162 4193654 3.45pF 115.13mH 12.51fF 15.7 193,200 3 4031554 4031738 1.97pF 329.34mH 4.73fF 39.7 210,100 4 4193162 4193664 3.37pF 113.01mH 12.75fF 16.3 182,700 5 4094821 4095266 2.75pF 132.08mH 11.44fF 22.4 151,700 6 3998949 3999388 3.66pF 134.77mH 11.75fF 21.4 158,200

As you can see, excellent agreement. Resistance values a few ohms low and I think that is due to inductance in variable resistor and I'm going to return to try another variable resistor to see the Cermet variable does indeed have distributed inductance in it. Film at 11, meaning I'll get back on it PDQ.

March 26, 2015, 0230 UTC. OK, put the Cermet 25T variable resistor into the AADE L/C II meter and found 1.0uH. That means at 7MHz I've got 50 ohms of inductive reactance in the circuit, so my voltage out match is off due to the inductance. So, next step is to take an ordinary trim pot and make a small board with leads to fit the socket and remeasure all the R_m values. This will, in turn, change the Q values for the crystals. Stay tuned.

My email address is the usual hiding from Internet bots.

chuck dot adams dot k7qo at gmail dot comHaving just watched 'The Equalizer' movie, I thought it interesting to have a list of 100 books that one should read before dying. I, like every one else, got on the Web and googled for the list. It seems that there is not a list that every one can agree on. No surprise there. So, I just picked one and here it is: