Saturday, September 10, 2022

The Cure for Frequency "Bunching Up" in Analog LC VFOs -- It Is Not So Simple. But we have a good calculator. Comments sought!

Bob's calculator shows good tuning linearity with an ordinary SLC capacitor 

One of the complaints about analog LC VFOs is that they have non-linear tuning -- when you turn the dial (usually attached to a variable capacitor) the space between frequencies is NOT constant.  This is especially apparent at the high end of the frequency scale where frequencies (and stations) appear to be severely bunched together, making tuning difficult.  This problem contributes to the defection of some great homebrewers:  They give up on LC VFOs and they switch to digital VFOs.  Sad. 

But there is hope:  Not all LC VFOs tune this way.  Even on rigs from "back in the day," back when the Si5351 hadn't even been thought of,  some LC VFO rigs tuned linearly.  My beloved Drake 2-B and my almost equally beloved HT-37 are good examples.  How did they do this?  How did they escape the dreaded "bunching up?"  

For a while, I thought that it might have had to do with the use of the series tuned Clapp circuit.  But on further noodling, this didn't seem to make much sense. 

Then -- like others -- I thought that it  must be caused by the adroit use of special capacitors. You see, in ordinary variable capacitors when you turn the dial, the capacitance increases linearly.  But in the LC circuit, frequency changes as the inverse of the square of the capacitance. Thus the bunching up. So the solution must come from the use of the special capacitors that compensate for this, that -- because of the shape of their plates --  produce linear tuning.  With these variable caps, frequencies on the dial are spaced out nicely, there is no bunching up. Great right?  

From Terman, Radio Engineers Handbook, 1943, page 123

Over the years, many hams have jumped to the conclusion that rigs with good tuning linearity MUST be using these special caps. For example, in 2013 a ham posted in the Antique Radio forum this message: 

There are three types of open, variable plate caps;
SLC= straight line capacitance where the capacitance varies linearly,
these are the most common and have half-circle plates
SLF= straight line frequency where the plates are tapered to allow
for linear tuning of the frequency
SLW= straight line wavelength, you get the idea...

SLF and SLW caps have oblong plates.

The effect on tuning a receiver can be dramatic. One example is the
Hammarlund SP series of receivers where the ham bands are very
compressed at one end of the tuning range. They used SLC caps
in the VFO. On the other hand rigs like the Kenwood TS-520
and FT-101 series have linear tuning across each band. These use
SLF variable caps. Most old 1920's battery radios used SLW
where stations were identified by their wavelength.

Well, not really.  

-- I now have several VFOs from the extremely linear-tuning FT-101.  But when you open them up to look at the tuning capacitor, it is NOT a Straight Line Frequency capacitor.  

-- Many of us over the years have built VFOs that are quite linear in their tuning without resort to these special capacitors -- we did it with ordinary Straight Line Capacitance caps.

-- When you look at the "How to build VFO" literature in the ham radio books, you see a lot of good recommendations about using solid, brass-vaned caps with ball bearings at either end.  But never do you see circuits that require the use of SLF or SLH capacitors. If they were the key to tuning linearity, we'd see them mentioned in the literature. But we don't.   

So where does the linearity -- or bunching up -- come from?  

The answer comes to us from a really neat calculator from Bob's Electron Bunker: 

http://electronbunker.ca/eb/BandspreadCalc.html

This calculator allows you to select your frequency range, and the tuning range of your variable cap. It then displays for you what the tuning range will look like on your dial.  You can see if there will be bunching up, or if the frequencies will be nicely spread out.  And -- and this is the really cool part -- you can then specify if your capacitor is SLF, SLW, SLC or Midline-Centerline.  This really illustrates the effect of the different capacitor types. 

I used Bob's calculators to do some experiments with various types of capacitors, various frequency ranges, and various combinations of trimmers and padders.   You can see what I did here: 

http://soldersmoke.com/VariableCapsSLCSLF.pdf

One important thing to keep in mind:  The SLF caps were made for AM broadcast receivers that were tuning from 540 to 1600 kc.  That is a 3:1 tuning range.  Most of the time in HF ham radio, we are tuning across a much smaller range, say from 5 MHz to 5.5 MHz.   That is a 1.1:1 tuning range. In those cases where we ARE tuning across a wide tuning range -- for example with a receiver covering 3-9 MHz, the SLF cap can help prevent the bunching up. 

But we can have fairly good linear tuning without resort to SLF caps.  Bob and his calculator point out that by narrowing the frequency range of interest, and by using either smaller range caps (ordinary SLC caps), or SLC caps with trimmers and padders, we can achieve tuning linearity.  And sometimes, when you have achieved this nice tuning linearity with a plain SLC cap, putting a fancy SLF cap makes tuning linearity worse. 

One piece of VFO tribal wisdom that is confirmed by all this:  It is better to use a smaller variable cap with a maximum capacity of about 30 picofarads. 

I think we should spend as much time focusing on VFO tuning linearity as we do on VFO frequency stability.  Bob told me that in the old days, the calculations for various tuning linearity scenarios were difficult.  But now we have Bob's calculator.  When building a VFO, just use Bob's calculator, plugging in the numbers to get a preview of what your tuning linearity will be like. If it is bunched up, you can play with the trimmer and padder values to achieve the tuning linearity you desire.   


Thanks to Bob of Bob's Electron Bunker for this great calculator. 

You can see another discussion of "bunched-up" tuning in the comments section of this article: https://www.nutsvolts.com/?/magazine/article/may2015_Whipple

What do you folks think of this?   Please put comments below. 

8 comments:

  1. Very interesting Bill,
    I do recall when I built my version of The Progressive RX
    (http://www.parelectronics.com/par-homebrew-projects.php) I had to do a lot of experimentation to achieve a linear dial. Bob's calc would have been very welcome. As I recall a varicap L-C circuit is just the opposite- bunching up at the low end where capacity changes rapidly with small changes in voltage. I recall using an audio taper pot to help that out- probably along with a series or parallel cap. Too many years back now.
    Love the articles,

    Dale W4OP

    ReplyDelete
  2. Great to hear from you Dale. I had my FIRST Barebones Superhet on the bench yesterday. I thought of the ones you sent me. FB! Thanks and 73 Bill

    ReplyDelete
  3. Hi Bill - the math lurks in amateur radio publications and articles form the later 60s into
    the 80s. The most popular available resource would be:

    A Progressive Communications Receiver by Hayward and Lawson
    QST for 1981. Page 21 This is a must have article for anyone who ponders home brew design.

    The whole premise becomes:

    Assuming you have a mechanically fit air variable capacitor with a desirable temperature coefficient (TC)
    For example, not a polyvaricon. Or you manage to find a rare, old adjustable TC capacitor
    with say for example, offers >= 200 pF delta C.

    You can make any value of capacitor work on your VFO by applying the correct series and parallel connections of fixed capacitors

    In some cases, I might even consider making 1 of them a variable trimmer capacitance for in-situ trimming and also work that capacitor's TC into my overall temperature compensation scheme

    The equations to calculate are not complex and are easily performed with a basic calculator.

    You can also use basic geometric mean analysis in some situations - something we use a lot in electronic design

    X = unknown fixed capacitor
    C2 = current target swing (delta c) of our tuning capacitor
    C3 = the target delta of our tuning capacitor that we seek

    X = C2 / (C2-C3/C3)

    Best!!!

    ReplyDelete
  4. Didn't I do a varicap in one of those? Pwrhaps that was the one where I made a compartment and filled it with tiny styrofoam spheres.
    Today's restoration is a long sought after Lafayette HA-410 10M AM rig. It is coming out better than new. Nuvistor front end and it hears my Adret sig gen down to -130dBm.

    ReplyDelete
  5. Dale: Yes, the first one has a varicap and a DC to DC voltage increaser to allow for wider freq range. I have had many different oscillator circuits in there. It currently has an LC VFO that allows it to tune 17 meters. And the tuning linearity is pretty good, even with an SLC cap! 73 Bill

    ReplyDelete
  6. Spasibo Vasily! I took a new look at the progressive. Nice RX and many great ideas in there. One thing I wanted to ask you about: The active AF filter in the audio chain. I can see the need in a DC receiver, but in a superhet? Farhan is using something similar in his Daylight transceiver. But do we really need this in a superhet? 73 Bill

    ReplyDelete
    Replies
    1. I agree with Farhan’s approach, however, unless, I’m making a discrete ‘transistor radio’, I’ll use 5532 op-amps to make the low-pass filters in the low-level part of my AF chain to reduce noise.

      I don’t make superhets — prefer DC architectures. I’m all too aware of the many DC deficiencies such as noise ---- DC offset from the LO and RF being so close together, poor second order intermodulation distortion + flicker noise all which vexingly gets into the baseband audio, but still prefer them over superhets and I'm thrilled to eliminate all the IF stage signal processing and associated cost.

      Still, when I did build superhets, I preferred to put some low-pass AF filtration in the AF chain. This actually goes for all of my receivers including my radio telescope receivers: I want to define the audio bandwidth with some high-pass, plus low pass filtration at a fairly low signal level in my audio chain – even if it’s just 1 pole for each.

      The high-pass filter could simply just be an AC coupling capacitor in series with the next stage’s input impedance to roll-off the rumbling lows – or it could be an active stage using 4 poles. For low-pass filtering, I normally choose a value between 2 and 6 KHz 3dB fCo ; probably the latter is chosen more often. This works to lower some of the high frequency noise going into the next AF preamp stage and then to the PA. I use low Q, Gaussian-response filters to avoid group-delay effects and because I still want to hear some AF harmonics. This is not a vigorous CW low-pass filter – that’s a whole other ball game.

      If you’re making a low complexity receiver you can just use a <= 1 µF series polypropylene coupling cap from your post-mixer AF stage (or from your active mixer output port) and then apply even a higher value such as .001 µF or so shunt to ground somewhere early on to do the same high-pass/low-pass thing, but this time, completely passively. The low-pass cap also bypasses any BCB AM detection early-on in your AF chain.

      I listen exclusively through loudspeakers so my ideas might not be palatable for some. To each their own.

      Delete
  7. For CW- yes it was beneficial. However, in my version I designed separate crystal filters for AM, SSB and CW, so the active filter board is here, but unused.
    Dale W4OP

    ReplyDelete