Friday, April 17, 2020

Walter's VFO

Main VFO board under plastic -- buffer board off to the right
There is a lot of tribal knowledge in this short article by Walter KA4KXX.  (The most recent issue of SPRAT has a picture of a SPRAT cover from more than 30 years ago, announcing an article by Walter.  OM has been at this for quite a while.)  Walter's advice is similar to that of Doug DeMaw, especially on the air-core coil.   FB Walter,  thanks for sharing.  

Regarding the last line in Walter's article, please direct any complaints about drifting analog VFOs to Peter Juliano.  He will be delighted to receive these complaints, and will promptly provide all correspondents with very succinct advice on how to overcome the instability.  

My Way to a Low-Drift Analog VFO, by Walter KA4KXX

I recently built a digital VFO (from a QRP Labs Kit) to see what everybody is talking about, and it has been quite handy for my first two-band homebrew transceiver, but by the time one adds the low pass filter, a low-level RF amplifier to boost the weak output, a 5 VDC regulator, and a hefty +12VDC power line filter to the beast to keep the digital noise out of the rest of the radio, for a single band project I still prefer a low-drift analog VFO.  

My approach to minimizing drift is very simple and works well for a VFO range of 3 – 7.3 MHz, which is all that is needed for the 20 to 80 Meter bands, either direct conversion (adding a doubler for 20M) or single conversion with approximately a 10 MHz IF.  

First, see my schematic (adapted from Small Wonder Labs 40+ transceiver, original BITX40 Analog VFO, and other sources) and wind an air core inductor with stout magnet wire such as 24 or 22 AWG.  Use a thick, rigid plastic form of a diameter so that you need about 12 turns, and single coat with water-based sanding sealer, Q-dope, or similar.  Use hot glue to mount the coil firmly to the single-sided circuit board, and build a cover of some type, especially if the radio is to ever be used outdoors.
   
Second, all the VFO capacitors (except power supply bypass) should initially be the modern C0G type, which can be obtained from Mouser (such as TDK FG28C0G1H681JNT06 or the like), which I trust more than those labeled NP0.

Then, power up the VFO and tweak the coil, tuning arrangement, and range capacitor to get the frequency range you desire.

Next, monitor the drift from a cold start to see how fast it is moving as it warms up, and whether it stabilizes nicely (my goal for SSB Phone use is less than 20 Hz drift during any 10 minute period) after 10 – 15 minutes maximum.  If it does not stabilize to your satisfaction, then start substituting polystyrene caps for the C0G units one at a time until you are happy with the performance.

[Also remember that a stable BFO is important as well, and if you use the BITX 40 crystal oscillator design, I recommend installing a dedicated 78L09 power line regulator.]

If you build this VFO at the higher (7 MHz) frequency end, just change the inductor value to about 0.5 uH, with everything else about the same, but expect to do more tweaking to get stability, and the warm-up time may be closer to 15 minutes.  After years of experimenting, this is what has worked for me.  If it works for you, please send me an email (see my QRZ page) of thanks.  If not, file a complaint with Soldersmoke. 



4 comments:

  1. I bought some NP0 caps some 10 yrs ago from Mouser for the far off VFO, once my courage was up, to build a VFO. All I had read about was how much a pain they were to stabilize. I built one with many small caps, all NP0, mounted the voltage regulator and J310 in a separate enclosure to keep their heat out. As hard as I tried, I could not find the dreaded drift at 7mhz. I tried to find it, I let it run for hours, it was stable as rock! Right from start up! It is my MIRACLE VFO! Ed KC8SBV

    ReplyDelete
  2. Good advice. You can make a rock solid VFO, just takes a few precautions. Ill add that I try to stay away from using an air variable with a lot of plates, such as say a 100pf with a padding capacitor. Id much rather use something along the lines of 35pf air variable with no padder.

    ReplyDelete
  3. Walter is famous! He has published in SPRAT!!

    Ed KC8SBV

    ReplyDelete