TheRadioBoard

I'm playing with Bob Weaver's excellent superhet tracking calculator and am surprised at the large effect on tracking of tiny changes in L or C, and am wondering how one actually aligns such a tracked set in practice since small errors will have large effects on the tracking.

For example, my SW1 range will be 3000-6000 kHz, with an IF of 2000 kHz. My varactors go up to 530 pF. Plugging this into the calculator gives a nice tracked response:



OK, so I try to substitute the capacitance values with some caps I have on hand, and the tracking gets way off:



Even a change of 0.01 uH in the inductance makes a huge change in the tracking. Here the LO coil is increased by 0.01 uH.



Here the LO coil is decreased by 0.01 uH.



So I'm wondering, how do you do the alignment in practice if tiny changes in L or C can throw the tracking so far off?

One idea I've considered is to make the front-end aerial tank oscillate at its resonant frequency by attaching a JFET and a tickler coil to the aerial coil. Then as the LO tank capacitance is tuned (ganged with the aerial tank capacitance), the front-end aerial tank's oscillation frequency will also change, above and below the IF of 2000 kHz. Assuming I have a working 2000 kHz product detector attached to the mixer output, I should be able to audibly "hear" the tracking as an audible heterodyne and hear as it goes above and below the IF of 2000 kHz. The LO tank L can be adjusted (in 0.01 uH increments!) until the tracking is reasonably correct.

I feel there is probably a better way to do the alignment. Any suggestions?

Also, I was considering hand-winding toroidal coils for the LO and aerial tanks, but seeing as how changes of 0.01 uH make such a large change in the tracking, I don't see how to adjust the L in such small increments.

Maybe it's just not feasible to homebrew accurately-tracking LO/aerial tanks over such large frequency ranges due to the precision required. Thoughts?

Ideally, the trimmer, padder and inductor should be adjustable, so you shouldn't have to worry about having the correct fixed values on hand. Of course, you can use fixed values to get into the right range, and then use trimmer caps to do the final adjustment. For example, if you need a padder of 836pF, you could parallel a 680pF fixed cap with a 200pF trimmer.

As for the general alignment method, it's an iterative approach. I forget the exact order, but IIRC, the inductance is adjusted for the lowest tracking frequency (3138 kHz in your example), the padder is adjusted at the middle tracking frequency (4219 kHz), and the trimmer is adjusted at the upper tracking frequency (5719 kHz). Then, the process is repeated as many times as necessary until there's no further change in the tracking. I may have the order of the inductor and padder backwards, but the trimmer is always done last and is done at the upper frequency.

When I built my homebrew superhets, I found that I was able to use the closest fixed value cap for the padder, and then compensate with the trimmer and inductor adjustments. A feature that I should add to the program at some point, is the ability to specify certain fixed values for some components, and then have it calculate the best values for the remaining ones. Hmmm, maybe a good project for when I get bored while visiting the relatives next week.


BTW, what operating system are you using? I don't recognize the window style.

Ah. I have trimmer caps on hand, but no adjustable inductors. I initially assumed I could "squeeze" toroid turns to do the inductor adjustments, but seeing as how 0.01uH or more precision is needed for adjustment, looks like I need to find (or make?) some slug-tuned inductors.

Also, how is "change in the tracking" measured? By ear, judging for max volume?


While we're on the subject of new features... how hard would it be to add low-side injection calculation as an option?

I'm running on Ubuntu GNU/Linux version 10.10. I'm pretty happy to see that there's a lot of good Linux software available for radio enthusiasts. Thanks for making your program work on Linux as well.

Been thinking about how to make adjustable inductors that can be adjusted with 0.01uH precision as is apparently needed for tracking. I'd prefer to avoid buying special slug-tuned inductors whose future availability may be uncertain.

One idea: wrap double-sided tape on a PVC pipe then close-wind the coil on top of that. The tape keeps the turns in place but allows removing turns. During alignment, remove turns until you're in the ballpark. For very fine alignment near the end, adjust the spacing between the last turn and the rest of the coil. I'm guessing this would allow fine-enough adjustment for tracking purposes. When finished, affix the turns permanently in place with Q-dope or similar.

Thoughts, or other ideas?

Yikes. You could use a second, smaller, adjustable coil in series with the fixed coil, and tweak that one. I've never met a circuit that didn't want to be tweaked down the line. Once you glue down that winding you're going to want to change it. You can make a decent slug-tuned coil with a brass-screw slug. Check out the Tin Ear receiver, which uses a brass screw in a plastic straw as the VFO tuning element.

http://www.amqrp.org/kits/tin_ear/index.html

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

qrp-... re: KR1S' suggestion


In a similar vein, was going to say that for a small adjustment, the coil leads could be coiled in a "pigtail" of a few turns which could then be stretched (or compressed) for micro adjustment? (I'm assuming the coil wire would be relatively stiff.)

ie. a "gimmick" inductor.

..................................................

About the smaller pigtail coil and stretching/squeezing it - according to some rough calculations, it seems that a compression of 1mm will cause an inductance change of more than 0.01uH, making tweaking difficult.

Regarding the brass screw slug idea - is the brass sufficiently non-lossy at RF (3-30 Mhz) for this usage? I'm concerned about the use of potentially lossy cores in the aerial coil, where our precious signal might get attenuated.