this is how breadboard looks inside... note the large area of the individual metal contacts that are arranged in parallel with the next one. two of those will have much higher capacitance than two wires or PCB tracks that are same distance from each other:

It looks like the parasitic capacitance ADDS to the total capacitance in my circuit. Is that always the case? Anyway, I rebuilt my circuit on a small piece of perfboard. BIG difference! The range is MUCH greater. Still have to make a few adjustments, but you pointed me in the right direction. I learned more from you in minutes than from weeks of searching and experimenting. I'm blown away. A big thank you!

 

It looks like the parasitic capacitance ADDS to the total capacitance in my circuit. Is that always the case? Anyway, I rebuilt my circuit on a small piece of perfboard. BIG difference! The range is MUCH greater. Still have to make a few adjustments, but you pointed me in the right direction. I learned more from you in minutes than from weeks of searching and experimenting. I'm blown away. A big thank you!

 

What I have rad about but not actually tried is expanding the voltage variation across the diode, including polarity reversal. Polarity reversal will have some effect and so changing the polarity as part of the sweep is a possible simple option. But limiting the tuning voltage range is rather a solid limit.

Of course, changing the performance of the circuit without any modifications will be a real challenge. I once worked at a place that did demand that on a number of occasions. When I succeeded in doing that a few times it was taken as a matter of course that it was always possible, simple, and cheap.

 

this is how breadboard looks inside... note the large area of the individual metal contacts that are arranged in parallel with the next one. two of those will have much higher capacitance than two wires or PCB tracks that are same distance from each other:

I took your advice and rebuilt my VCO on a copper-clad board with much success, thanks to your help. Also, I used the circuit you mentioned as a guide. Here's where I am now with new questions:

What’s the best way to couple my VCO to a buffer amplifier? I have enclosed circuit diagrams of the latest incarnations of my Vackar VCO and a voltage follower used as a buffer amplifier. Both circuits work great, but I’m confused about how to hook them together. Initially, I coupled the two circuits with a small capacitor, around 10pf. I wasn’t getting much output, around 150mv RMS, so I tried progressively larger caps up to 1000pf. That gave me more output. I reasoned that larger caps have less reactance, so no reactance would be even better. So I eliminated the coupling cap and connected the circuits directly. Voila! Way more voltage: Here’s what I measured:

347mv RMS @ 68.7MHz
460mv RMS @ 88.0MHz
254mv RMS @ 98.0MHz
215mv RMS @ 108.0 MHz
195mv RMS @ 110.9MHz

I am happy with the wide tuning range and level, but there are a couple questions. From 88MHz to 108MHz the output voltage decreases by more than 50%. How can I “flatten” the output voltage?

Also, it seems that connecting the circuits directly as I have, should NOT have worked, but it does. It works great. I noticed that a direct connection puts the 360 ohm resistor in the VCO in parallel with R1 in the Buffer. I thought that would mess up the bias of Q2. Instead, it was an improvement, so where did my thinking go wrong?

 

i like deferent strategy, where every stage of the circuit should really do one thing only... any time you ask for more than that, you are compromising it by introducing tradeoffs.

when it comes to oscillators it is a good idea to let the oscillator do one thing and one thing only - oscillate. getting correct waveform, frequency, range, stability etc. is already plenty of requirements. do not compromise any of those by trying to get large output. large output is something that next stage (or more) should take care of.

at 100MHz, 5pF capacitor has only impedance of some 300 Ohm.
and... that is rather low value. i probably would not put anything larger than that.

for comparison 1nF is only about 1.5 Ohm which is not much different from acting as a short circuit.

if you look at oscillator shared in post 6, oscillator coupling uses 6.8pF and 33 Ohm.

 

Been there, done that. And no the breadboard will work against you at these frequencies. Before I later created real PCB's I started to use what is called Dead Bug wiring PCB's. It has one large groundplane and you glue islands of copper where you need it. Keep all wires real short and I think your project will be much more succesfull.

 

What I have rad about but not actually tried is expanding the voltage variation across the diode, including polarity reversal. Polarity reversal will have some effect and so changing the polarity as part of the sweep is a possible simple option. But limiting the tuning voltage range is rather a solid limit.

Of course, changing the performance of the circuit without any modifications will be a real challenge. I once worked at a place that did demand that on a number of occasions. When I succeeded in doing that a few times it was taken as a matter of course that it was always possible, simple, and cheap.

I did that by adding a 9v battery in series with the main battery. It worked! Thank you!

 

i like deferent strategy, where every stage of the circuit should really do one thing only... any time you ask for more than that, you are compromising it by introducing tradeoffs.

when it comes to oscillators it is a good idea to let the oscillator do one thing and one thing only - oscillate. getting correct waveform, frequency, range, stability etc. is already plenty of requirements. do not compromise any of those by trying to get large output. large output is something that next stage (or more) should take care of.

at 100MHz, 5pF capacitor has only impedance of some 300 Ohm.
and... that is rather low value. i probably would not put anything larger than that.

for comparison 1nF is only about 1.5 Ohm which is not much different from acting as a short circuit.

if you look at oscillator shared in post 6, oscillator coupling uses 6.8pF and 33 Ohm.

I agree that each stage should only do one thing. So, I guess I shouldn't worry about increasing the output since it's gonna be followed by another amp stage. I still don't understand why using a 1nf coupling cap (basically a short) or direct connection actually works. I think you're suggesting that there should be some reactance and resistance like in the circuit in post #6 where they use a 6.8pf cap and a 33ohm resistor. What's the purpose of the resistor? I tried that and it works, but it decreases the output. What is the advantage?

 

Been there, done that. And no the breadboard will work against you at these frequencies. Before I later created real PCB's I started to use what is called Dead Bug wiring PCB's. It has one large groundplane and you glue islands of copper where you need it. Keep all wires real short and I think your project will be much more succesfull.

I started using dead bug construction recently and it does work much better. I'll stick to breadboards for the low frequency stuff. Thanks.

 

if it works for you, so be it... mission accomplished. i am not sure what you goals are. originally i though you are building some oscillator for test purposes but now i see plans to add amplifier and maximize output...

well... in that case, connect that puppy to a large RF amplifier and you can transmit on many frequencies simultaneously and interfere with pretty much anything. for maximum effect also modulate it with something like square wave (or two).

but if goal is stability and signal fidelity (sine wave), the idea is that load or whatever the following stage is, should have no or very little impact on oscillator... this means changes in load (impedance of next stage) should have no effect.

for example, temperature is one of key things affecting oscillator. so oscillator should use better quality capacitors with zero temperature drift (NP0). capacitors in other stages (unless tuned circuit) are not really important. also to minimize thermal drift, currents in oscillator may need to be small. and this affects driving of load (next stage).

you are using different approach... 12V supply, probably not even regulated... reduced values of resistors in emitter and collector circuit mean larger transistor current. together this means more power (and heating) in the oscillator, etc. capacitors C2,C3 and C4 as part of the resonant circuit. those are small value capacitors. even slight change affects resonant frequency. but then you use huge capacitance of 1nF to next stage. and the capacitance values in the next stage are neither precise, nor controlled nor insignificant...

Vackar is relying on 180deg phase shift in pi-filter and it easily starts oscillating. But as with any oscillator, if overdriven, the output form is distorted. and that means harmonics... your broadcast will be all over the FM band.


as mentioned, since the plan is to add additional stages to amplify the signal, i would prefer to see oscillator produce clean signal (even if weak), buffer it, then let the amplifier do its job. if the following stages contain tuned circuits, they can help removing unwanted frequencies. RF amplifiers are normally C class with tuned LC circuits so distortion if the signal is not an issue but signal drifting to another frequency due to oscillator shortcomings still is..


the next question is do you have license for this...?

 

... in the circuit in post #6 where they use a 6.8pf cap and a 33ohm resistor. What's the purpose of the resistor? I tried that and it works, but it decreases the output. What is the advantage?

capacitors block AC signals. also their impedance is frequency dependent. circuit in post #6 is meant to be very wide frequency range so most likely capacitor value is purposely large 6.8pF (low enough impedance at lower frequencies). resistor is probably added to make output more uniform across entire range.

 

My goal is a decent signal and fidelity good enough for voice. It's partly a hobby experiment, but anticipate a use for it if I can make it strong enough to work in a very hilly terrain. No, I don't have a license, but I won't be around any humans so I'm not worried about interference. This thing only has to be strong enough to get over some hills, maybe a few hundred feet. The signal I get seems pretty pure. If there are any harmonics, they're way above what I can see on my scope. A low pass filter seems easy enough to add. Anyway, some of this is still a mystery to me, especially when I'm trying to learn and my predictions turn out wrong or things behave counterintuitively. Maybe I should learn to use a simulator. Every change I make involves de-soldering and re-soldering. I have one more question if you have time. Where do I insert the audio so that it modulates the RF? Thanks again for all your help.

 

My goal is a decent signal and fidelity good enough for voice. It's partly a hobby experiment, but anticipate a use for it if I can make it strong enough to work in a very hilly terrain. No, I don't have a license, but I won't be around any humans so I'm not worried about interference. This thing only has to be strong enough to get over some hills, maybe a few hundred feet. The signal I get seems pretty pure. If there are any harmonics, they're way above what I can see on my scope. A low pass filter seems easy enough to add. Anyway, some of this is still a mystery to me, especially when I'm trying to learn and my predictions turn out wrong or things behave counterintuitively. Maybe I should learn to use a simulator. Every change I make involves de-soldering and re-soldering. I have one more question if you have time. Where do I insert the audio so that it modulates the RF? Thanks again for all your help.

 

you already have varactor.so that is the simplest way... just mix it with biasing voltage used to tune the oscillator frequency.
add capacitor so that audio is only AC coupled.

 

you already have varactor.so that is the simplest way... just mix it with biasing voltage used to tune the oscillator frequency.
add capacitor so that audio is only AC coupled.
View attachment 332227

That makes sense. I see that you moved the position of the RFC from the tuning pot to the varactor. I assume that it won't attenuate the audio very much since it's low frequency. And of course that cap insures that no DC gets through. You have been a great help for my learning process. Thank you for sharing your knowledge. Folks like you who are willing to spend time helping newbies like me are what makes this such a great hobby. And you have spent a lot of time helping me. I don't know how, but I wish there was some way I could return the favor. Cheers from Philadelphia! Allen.

 

By the way, you are increasing the tuning range, not the bandwidth. A stable oscillator is supposed to have a very narrow bandwidth, no matter what the tuning range is.

 

By the way, you are increasing the tuning range, not the bandwidth. A stable oscillator is supposed to have a very narrow bandwidth, no matter what the tuning range is.

Thank you for pointing that out. I was using the term "bandwidth" incorrectly. I understand the difference. Newbie mistake, but you guys are helping me to learn. Thanks again!