EMRFD Message Archive 12556

Message Date From Subject
12556 2016-03-30 23:04:25 gamma_function Stray capacitance in oscillator circuit for IC-based DC receiver.
Hi all- I'm new to homebrewing and I've just picked up this very interesting book. I'm building the IC-based DC receiver in chap. 1. As you know, Mr. Hayward encourages using whatever is at hand to build the oscillator circuit. I don't have 50pF and 80pF variable caps but I do have a 30-410pF variable cap. A bandspread calculation for the 40m band yields a 585pF trimmer cap, a 200pF series cap, and a 3.15mH inductor (about 25 turns on a T50-2). Does this seems reasonable and do I have to worry about stray capacitance?  If so, how do I find that out? Even small amounts of stray capacitance seem to alter the calculation dramatically. There is no mention of it in chap. 1 so does it matter?

Thanks very much,

Brian




12557 2016-03-31 05:06:14 Nick Kennedy Re: Stray capacitance in oscillator circuit for IC-based DC receiver
I've recently been following a similar path, building a VFO, although it's not IC based.

Stray capacitance can have a large effect as you say, and it doesn't all have to be 'stray', since some of your capacitance may not be exactly as marked or as measured.

I think the key is to have your trimmers positioned so you can adjust for strays while monitoring frequency. It's also necessary that your stray C will "stay put" and not vary greatly with temperature, so you minimize drift. But that's your next phase after building.

73-

Nick, WA5BDU

12558 2016-03-31 07:09:56 Mike Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Brian:

The emrfd NE602/LM386 receiver is a great choice if you are looking for an easy 40M receiver build, particularly if you have not previously built up a receiver from scratch.  However: the 30-410 pf capacitor that you wish to use for the oscillator has a rather wide range (lots of capacitance change), in my opinion, if you are trying to get it to "sweep" from the low to the high end of the 40M band, but not too far beyond these limits.  (You did not indicate what the tuning range is that you'd like to cover...)  The emrfd description deals with restricting the effective tuning range by providing a series capacitor (10 pf in emrfd) in series with the "fine tuning" (C2; 50 pf) capacitor to achieve a limited tuning range that (I assume) does not extend too far beyond the low and high frequency limits of the 40M band.  Another approach (and one that I recommend for you, since you want to use a 30-410 tuning capacitor for tuning, is to "tap down" the 30-410 pf cap at a low tap point along L2.  You would still need capacitance (a fixed capacitor, or a parallel combination of fixed and variable capacitor across L2) to get the LO to oscillate at a bit above the center of the 40M band, and then by adding the 30-410 pf capacitor will lower the LO frequency into the 40M band where you want it to oscillate.  "Tapping down" the 30-410 pf capacitor connection on a tap of L2 (maybe somewhere between 10 and 20 percent, measured from ground) will greatly reduce the frequency range over which the LO will operate as the 30-410 pf capacitor is adjusted- and I suspect that this is what you will want.  You haven't mentioned if you are using a vernier dial of any kind, but if you are just putting a knob on the dial of the 30-410 pf capacitor to do the tuning, you will find that the tuning is very "fast" and it will be difficult to tune in a signal unless you have an extremely steady hand.

I've built several analog LC VFO's for receivers- including one using the neophyte circuitry- and in all cases I have tried to used all fixed capacitors other than the "tuning" capacitor in the LC circuit.  But in each case, this has resulted in a need to wire up the circuit and energize it, measure the frequency range covered, and then modify the circuit a bit (changing fixed capacitors, and/or rewinding toroids) to get the desired tuning range.  By contrast, the emrfd article describes use of two knob-adjusted variable capacitors in the LC oscillator circuit.  So the emrfd circuit allows for a single "build cycle" with subsequent use, but the presence of two variable capacitors makes it very difficult to apply any kind of frequency/dial calibration.  So this is something for you to consider.  I hope that your 30-410 pf capacitor is an air variable rather than a polystyrene; it will give somewhat better temperature stability.

A few other suggestions- use NPO capacitors for any fixed capacitors in the LO circuit.  As much as possible (entirely, if you can tolerate the build/measure/modify cycle that I describe above), eliminate any trimmer capacitors in the LO cycle.  Or minimize the amount of capacitance that the trimmer capacitor is contributing in the circuit, compared to the fixed NPO's.  (But if you can get a NPO trimmer capacitor- I've had little luck finding them- then this might be a moot point).  Use a -6 toroid (or better yet, -7 if you can get one) in the LO circuit rather than the -2 toroid that you are proposing; the -6 and -7 cores provide better temperature stability.

As Nick has stated, stray capacitance can affect the frequency.  But trying to figure it in from the start is difficult (for me, at least).  I've often gotten "close" with designs (in the 3-30 mhz range) by guessing at a value of 10 to 20 pf of total stray capacitance for the purpose of initial design/hookup/test.

If this is your very first receiver build, you are in for a treat.  And you will learn a tremendous amount.  Keep us informed as to progress.

73
Mike N4MWP

12559 2016-03-31 17:12:20 kg4rll Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Welcome Brian,
The first DC receiver I built was from another book , Solid State Design for the Radio Amateur. Great book as well very old school by today's standards but loaded with practical examples.
Used an MC1496 as the product detector and two NPN transistors for audio out.
VFO was a Hartley if I remember correctly. Covered 80 meters and 40 meters by switching in additional capacitance for the LO.
Thing drifted like crazy! But it worked! Sounded great while it was on frequency.
VFO stability is a combination of several things...
Mechanical stability and component selection are important even when I use NPO caps throughout and good mechanical construction they almost always drifted down in frequency. Some faster than others. The cure for my vfo stability was using a hand selected Negative temperature coefficient cap of the proper value.
Building a stable vfo is an art and a science have fun building and keep us updated....

kg4rll
12560 2016-03-31 18:07:42 Nick Kennedy Re: Stray capacitance in oscillator circuit for IC-based DC receiver
I appreciate the discussion of "tapping down" the variable cap on the inductor. Sounds like a good tool to have in the box. I guess you do this sort of empirically as opposed to calculating? Playing with it in LTspice seems like a good way to get in the ballpark.

My VFO was drifting a bit more than I'd like so after reading a bit about compensation in chapter 4 of EMRFD, I dug though some of my older junk boxes and found some funky old N150 temperature coefficient disks. Without trying to calculate closely how much I needed, I took a shot at adding a couple and reduced the drift to about a fifth of what it was, which was a great improvement, obviously.

I made a mistake in doing the board for my VFO - I didn't etch away the copper on the back side beneath the frequency determining components and so I have some circuit board C which is going to have a poor temperature coefficient. I think most of my drift was coming from my type 6 core inductor though.

73-

Nick, WA5BDU



12563 2016-03-31 21:42:29 Tayloe, Dan (Noki... Re: Stray capacitance in oscillator circuit for IC-based DC receiver
The tap ratio is simply the square root of the smaller desired C to the C you have.  It is just using the VFO inductor as a transformer.  The impedance transforms with the square of the turns ratio, I.e., the ratio of 1 (i.e. the whole coil) to the tap point.

For example, tapping half way down causes the reactance of the variable cap to be 4x, at the top of the VFO coil which means it "looks" like 1/4 the real C value.

I have used this before and it does work.  However, there is some risk that the circuit might not react well to this arrangement.

- Dan, N7VE

Sent from my Windows Phone
12564 2016-03-31 23:42:07 gamma_function Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Thank you for these comments. They've enlightened me.

Actually, my main variable cap (with a 6:1 reduction drive I'm hoping will work for tuning 6950-7300MHz) is nominally supposed to be a 365pF AM BCB cap, but I've measured it as 410pF! And that was with a Fluke 87-v, so who knows what it really is! I'm inclined toward the common denominator of the advise I've received and will try to keep it down to a small value trimmer cap to adjust for strays. All my fixed caps are, in fact, NPO types.

kg4rll, by 'mechanical stability' I take it you mean short leads, use of copper ground plane, etc.?

Brian
12566 2016-04-01 07:45:09 Mike Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Nick:

Not sure if this should perhaps be a new thread rather than "stray capacitance..." but here goes:

I think Dan Tayloe (and maybe others too) addressed the issue of "tapping down" on the inductor very concisely.  In response to your efforts to provide temperature compensation of your LC-based VFO, I strongly recommend that you read Wes Hayward's QST December 1993 article "Measuring and Compensating Oscillator Frequency Drift" (the article is available to ARRL members [QST subscribers] at the arrl.org website).  I must admit that I have probably re-read Hayward's article maybe 30 times; there's a lot of "meat" in it.  I also constructed a drift measurement chamber (using a styrofoam cooler) as described by Hayward, and used it to temperature-compensate a VFO.  But the whole process took me probably about 5 hours to "construct" the stability-drift-measurement chamber, and 20 or 30 hours for subsequent stabilization of a single VFO, so it's not something that I encourage others to follow casually.  My own suggestions regarding VFO construction to get temperature stability are as follows:

1.  Use NPO capacitors and air variable capacitors.  Use air wound coils (best), or -7 toroids (almost as good), or -6 toroids (still good).

2. Provide good mechanical stability for oscillator components.

3.  Don't put any heat-generating components near the oscillator components.  Get them as far removed from the oscillator circuit as possible.  Remember that any temperature rise (above the "starting temperature") in any of the oscillator components can result in some drift.  And this temperature rise can result from either internal heating of the individual VFO components themselves, or it can be temperature rise from the environment in which the VFO's components are located.

4.  Understand that providing temperature compensation over a wide temperature range is extremely difficult with an LC VFO.  Doing so at a fixed temperature is considerably easier but still takes some work (measure drift, modify circuit, retest...)

If you follow items 1, 2, 3 above, and if your VFO is used in a temperature-stable environment (e.g., in a room at your house) then I think you are likely to get very acceptable frequency stability.

73
Mike N4MWP

12568 2016-04-01 09:08:02 farhanbox@gmail.c... Re: Stray capacitance in oscillator circuit for IC-based DC receiver

i would add that using smd capacitors really makes a huge difference in the drift.

12570 2016-04-01 10:20:01 Mike Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Brian:

If you are going to use this capacitor with its built-in 6:1 vernier drive to tune over a 350 khz range as you indicate, it will provide an effective average tuning rate of about 117 khz per revolution.  I consider this to be very fast, and I suspect that you will find it difficult to tune in SSB signals (maybe even CW signals will be tricky, depending on filter bandwidth) at such a fast tuning rate.  Although a big knob will help somewhat.  You might want to reduce the range to a more limited portion of the 40M band.  Also, you are proposing to use a BCB variable, which has plates that are probably "cut" so as to provide (as much as possible) a "straight-line" correlation between capacitor rotation and frequency- but this is true only for the wide frequency spread (from 530 khz to 1600 khz) that is characteristic of the broadcast AM (MW) band.  When this capacitor is used for the much more compact frequency "spread" that you are targeting (6.95 to 7.3 mhz) the rotation vs frequency characteristic will be skewed considerably.

I'm not trying to discourage you, but just to give you some advance warning.  In any event, you will probably want to adapt and refine your VFO as time proceeds.

73
Mike N4MWP

12574 2016-04-01 14:31:41 kg4rll Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Hi Brian,
For me mechanical stability is put the whole thing in a box, variable capacitor included if possible. Mount the circuit board to the variable cap with screws and make a box around it with circuit board material soldered together. This provides mechanical stability, thermal stability and shielding the RF inside the box.( you don't want RF from the LO getting into the front end filter, causes microphonics)
The test for mechanical stability is tune in a station and then push in on the vfo knob and listen for any change in frequency.
If you can't do the box thing use solid conductor wire of a large diameter to provide stability. Definitely don't want the vfo wire moving about.
If you are using the oscillator on the 602 you could build the radio 602 , 386 combo on one board and mount the whole thing on the vfo cap and mount the cap to the front panel of the cabnet
Then build the front end and put it in a shield box to isolate it from the LO.

kg4rll
12576 2016-04-01 17:17:48 Gary Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Good or bad?  Why?
Thanks

Sent from my iPhone

12577 2016-04-01 23:33:04 Ashhar Farhan Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Here are my two bits. 

1. Try building the VFO with SMD components that are mounted across islands cut into copper clad board. the epoxy base in combination with the smd components ensures that the expansions are minimal.

2. Check the drift with and without the tuning cap. The wires to and from the tuning caps can be a source of drift. If possible, mount the tuning cap by drilling holes for it on the same copper clad board that has the rest of the VFO. ground the vfo's capacitor directly to the ground plane and use a very short, thick wire for the hot-end of the tuning cap.

after soldering/changing the vfo, wait for at least 10 minutes before starting out to measure the drift. VFO drift is specified after an initial time period of anywhere from 2 minutes to 30 minutes. 

my recommendation is that the VFO should be permanently connected to the power line, bypassing the power line. so when you want to fire up your rig, the vfo is already warmed up. 

- f


12578 2016-04-02 06:27:03 Tayloe, Dan (Noki... Re: Stray capacitance in oscillator circuit for IC-based DC receiver
I can imagine SMT would have a solid, low thermal resistance path to the ground plane, perhaps minimizing C change due to internal heating.

I was also impressed by the ultra low noise VFO circuit.  I looked up the transistor in the mouser catalog and found it was a 20w power device.  Not exactly a QRP sized. Again if it were mounted like a power device with a good thermal path, a few mA of current would cause almost no heating in that device, especially compared to our typical TO92 300mW transistor.

- Dan, N7VE

Sent from my Galaxy Tab® A
-------- Original message --------
12579 2016-04-02 08:40:18 Bill Carver Re: Stray capacitance in oscillator circuit for IC-based DC receiver
12580 2016-04-02 10:09:44 gamma_function Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Mike: I now must ask- If I did not already have a variable tuning cap and was hunting about for one (or two), what would you consider ideal?

Brian
12581 2016-04-02 10:10:40 gamma_function Re: Stray capacitance in oscillator circuit for IC-based DC receiver
I am indeed interested in exploring surface mount techniques. I recently stumbled upon this interesting post on electronics.stackexchange:
Making prototypes with high speed SMD components

 

12582 2016-04-02 22:27:03 vasilyivanenko Re: Stray capacitance in oscillator circuit for IC-based DC receiver

Greetings:


While the advice of that link article may hold true for some things –  it’s unlikely making SMD prototype circuits on PCBs will prove faster than building on a copper glad ground-plane for the HF experimenter.


You’ll only get good at making temperature stable VFOs by building them – talking, simulating, reading although helpful, won’t ensure temperature stability like bench experiments will. With respect to VFO temperature drift ---- only the first 5-15 circuits will prove difficult – then it gets much easier!


CG0/NP0 temp compensation caps versus through hole-hole CG0/NP0 temp compensation in L-C  VFOs :


Effects are frequency dependent – SMT proves more predictable at higher frequencies - SMT offers better

mechanical performance in the event of vibration but not under mechanical stress . They may also easily short during soldering by newer builders, plus add confusion by parts identification difficulties. I’ve not measured or seen better temperature drift in my transfer function graphs with SMT versus through-hole capacitors during my bench experiments.  The copper clad traces/pads needed for soldering the non-grounded end of for SMT parts may introduce “capacitors” of low Q and variable temperature drift into the resonator circuit.  


Board warping due to insufficient enclosure stiffness and/or slack mechanical copper board anchoring to this enclosure may intensify this effect and stress SMT caps more than hole-through types.


Wes and many others have  written ++ about techniques you may ply to help ensure low temperature drift in an LC VFO --- but at the end of the day, all your entire tank connected parts will contribute to temperature drift [up, neutral, or down]. Your job is to apply the techniques to minimize the drift in your initial build -- and then cancel drift in 1 direction with temperature compensation afterward. That’s best done on your bench through experiments. 

 

Best to you !

T/V

12583 2016-04-02 22:28:54 vasilyivanenko Re: Stray capacitance in oscillator circuit for IC-based DC receiver
Sorry for the typo ---  C0G/NPO temperature characteristics