Okay so ferrite slug tuned coils lower in frequency as the slug is lowered into the coil, however not sure how I never realized this but the metal slugs (shiny silver ones) have the opposite raising the resonant frequency when lowered into the coil?
I am building a VCO for the FM broadcast band to cover the channel I use for pirating and aim to allow a high quality POT change frequency of the oscillator via simple varicaps. Of course that is also where modulation is inserted. The goal is to make this oscillator as stable as possible. Hope to get near the low drift that a crystal would provide. Of course that is next to impossible but I can wish and try.
Oscillator transistor will be heavily shielded, buffer stages shielded and I plan on increasing capacitance coupling between the output to any further stage to even make it more stable. All ARRL precaution crap they talk about in those books will be taken lol.
So far I am using all NP0 capacitors in the oscillator stage and buffers. Stability is excellent. Warm up drift is almost zero since I am feeding the oscillator with very little and highly regulated voltage. Drift as I have studied is caused by cheap capacitors, transistors, badly designed coils, and so on.
Now I got this oscillator plus buffers to fire up exactly (near) frequency. Drift is only around 50KHz. I start out at 96.95 and after 5 minutes I am right at 96.9MHz. It sticks to that for hours. The variable resistor POT allows me to swing frequency or fine tune from 95-97MHz just in case.
Anyways back to the main question... What is more temperature stable? Ferrite slug tuned coils or the ones with metallic slugs?
I notice in FM stereo receivers that predate PLLs and digital tuning that they seem to use metal core slug tuned coils in their stages. Also a great benefit I can see there is that they are unbreakable when tweaking them with a tuning tool unlike ferrite which can snap when jarring it around or being completely careless with improper tools.
I really want to design this sucker to be as stable and spot on as possible so that eventually I have a simple mono exciter that I can just power up and is almost exactly on frequency and requires VERY LITTLE adjustment to get it spot on frequency and remain there w/o the need for a PLL.
I know getting this stuff spot on for VHF and above is asking for a lot. Even classic FM stereo receivers drift a few tens of kHz over hours of use or drastic temperature change.
Any tips here are welcome. Building a good VFO/VCO is its own art. I also notice NP0 caps tend to have a negative drift with temperature so I may need to use a mix of positive/negative coefficient caps to get complete stability in the end.
Holy hell this should be a blog, and may eventually be moved into a new thread dealing with further experimentation. Forgive my brain dump lol.
Here are a few basic questions I have... What type of slug coil is best? Wire to make it out of, coil former, type of ferrite/metal material to tune it that is least heat sensitive? How to achieve best drift management at VHF besides mixing a low with a high frequency? (or is that the only proper way?) Are there VFOs for VHF that are almost near rock solid? Schematics or tips on how they function so well?
BTW I really wish I could implement a basic blog function here. Maybe something to look into in the future.
My goal here is to build a backup transmitter, or even possibly replace the current one. This is only the first block of the design that I am starting with. Obviously after making a highly stable oscillator, or even making provisions for PLL control which after all varicap designs can be easily converted to then comes the PA stages to get it up to some reasonable power and harmonic filtering.
After I get that far then comes building a basic stereo MPX circuit which so far looks like I will start out with a simple 2x oversampled circuit to replicate the BA1404 like functions, aka generated by a circuit somewhat similar to... Circuit
The oversampled versions I only have in jpg format at the moment and no links exist to provide.
Supposedly with proper filtering this simplistic way of generating stereo can work almost equally as well as a complete analog MPX circuit. Just requires some harmonic audio filtering and phase shifting to get good stereo separation. If the BA1404 and the like can do it, so can I much better I would assume.
I can't address the slug material, but do have a tip for the capacitors. Look up web info on polystyrene capacitors. I've read they are good for temperature stability.
Found out it really wasn't so much the capacitors that were hurting me for drift as much as the varicaps themselves. I tested all the NP0 (also called C0G) ceramic capacitors I have by blowing my breath on them, using a hot lightbulb and even putting a lighter near them while testing on my multimeter to find the ones with little to no drift. Instead of using two varicaps like the schematic calls for I cut it back to using just one and replaced the other with a fixed capacitor which helped tremendously.
I am pondering if sticking a thermal resistor into the circuit somewhere to have some kind of opposite effect on temperature to get near neutral drift. Drift so far is down to .05MHz for 5 to 10 degrees temperature change. Somewhat taking a guess on that since my frequency counter is not incredibly accurate. Hopefully with careful selection of caps and some other tinkering I will get it to an acceptable level when all is said and done
Will definitely look into other kinds of capacitors. Also found out that it is brass slugs that move frequency up as the slug is lowered into the coil, and ferrite that lowers frequency as it is lowered into coil. I read somewhere that brass slugs are better for VHF and up. Not sure exactly why.
I simply stuck a generic 100k (+-50k) thermistor in series with a POT to drag the VCO tune voltage up/down very lightly as ambient temperature changes to compensate for the varicaps temperature instability. Guess instead of a PLL type circuit you could call this a temperature locked loop
By varying the POT control and putting a source of heat near the circuit, then letting cool, repeat until I zero in I seem to be able to make the circuit almost insensitive to temperature, which after all is the main issue with free running oscillator drift.
The thermistor came out of a junked stereo amplifier that was used to bias the power transistors so that as its heat sink got hot it would compensate with the biasing to reduce part failure.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
