Practical information on RF oscillations....

[guitar_199]

I am wondering if someone might help "shine a light" on a gap in my understanding.

How does spurrious oscillation usually show up in an RF circuit? What leads one to suspect that a circuit is oscillating and could use some work on neutralization?

Is there any outward sign or symptom? Is it caught on an oscilloscope by random chance?

I would appreciate any comments that anyone might be willing to provide!

Thanks,
Bob

 

[Handy Andy]

Got a Cobra 29 or 25' play with the TX mixer L18/L21 that long twin coil tank.

To develop "spurries" or harmonics, you just need to mix two signals together. RF or Audio. You can generate problems with simple overdrive of one signal.

You hear these events all the time on a typical CB - just listen to a local using an amp - not too hard to hear the "bandwidth" products - or even use a simple Heterodyne experiment thru two - single tone oscillators and just adjust the frequency of one so that it can "Beat" resonate to the other or move off resonance and you'll hear a pitch change in tone as a difference a generation of a Third - and even further tuning away, can generate a series of complex fundamentals.

You asked about Oscillation though, and much of that is thru design artifacts, improper matching of impedance and or power levels between stages making the output on one appear unbalanced to another.

Gain of the stage can also play a role -

You can also make a stage with such high gain, the part can place itself in a loop so it oscillates - so you need to correct the errant effect by neutralization - design a loop of a negative feedback path to correct the negative error

The "random" chance is also a part can self destruct from the parasitic oscillation - a frequency either higher or lower than the original frequency - due to mixing products. Or one part, or combination thereof, can work to make a "resonate tank" that is similar to a relaxation oscillator - self induced.

Much of the "spurious" effort is wasted heating and extra power consumption - but can still be seen

Even the use of a typical Power meter - a 2nd harmonic from the 1st - ADDS into the power present for the meter to "see".

In some cases, the 2nd harmonic or any order higher, adds their power to the meter - until the meters band width ability to detect the frequencies in it's band pass - exceed it's detectable bandwidth.

It's why Spectrum Analyzers are one of many preferred tools of the trade, a tech, engineer or designer can see the effects of saturation, clipping and distortion. It is then their job - to force the stage to produce more power into the band it's designed for, not wasted on extra frequencies not desired.

 

[guitar_199]

Interesting! So oscillations CAN/WILL show up on a spectrum analyzer......

I have been doing some reading on this in Lou Franklins book.... and he refers to very small caps along the IF strip.... may be collector back to base... or it may stretch a little further across the stage. It just makes theoretical sense that a tiny value cap is providing just enough negative feedback to make things settle down. I was just wondering what tells you that you have oscillations!!!

 

[Handy Andy]

Take a look at this thread...
https://www.worldwidedx.com/threads/rci-69-base.253568/

You'll see this...

And in another post or two down...

This...

​

Many of todays radios are simple high-gain stage RF amps - neutralized to a point but still able to produce mixing products.

So the above graphic shows "open" cap (deleted part) that when installed, can attenuate(neutralize to a degree) many artifacts or REDUCE THEM so they can be easily attenuated in the PI-Filter (output) network

But in doing so, REDUCES power - which many say they won't buy a weak radio - they want a BIG radio - SIGH - so the self-degradation cycle of radio continues...

​

 

[guitar_199]

So they want all the power they can get.... no matter what part of the spectrum it shows up in!!!!!!

But in doing so, REDUCES power - which many say they won't buy a weak radio - they want a BIG radio - SIGH - so the self-degradation cycle of radio continues...

 

[Handy Andy]

Sadly the NFB on a bipolar - they never bothered to straighten out the effort by making the loop adjustable.

Some examples others used are...

Cobra (2000/even 148's)

Does not mean they always did it, but they did provision for it.

Cobra 146GTL

Unidens Grant

HR2510...
​

Each one uses some form of NFB - not unlike Texas Star Amps does, but in the above cases, the Open loop gain was saved for last. But to keep waveforms within limits, many tied NFB to Drivers' loop gain.

Go back and see what Galaxy does, they provide, but not always install, "X".

In the above examples, any cap value can be used, and a resistor can also be used in SERIES with it to reduce or "tailor" the level of feedback needed to correct a spurry or two...nothing truly too detrimental to the performance of the radio.

​

 

[guitar_199]

I can't tell you how much I appreciate all of this info.

Got some studying to do!!!!!!!!!!

 

[Handy Andy]

Much of the "spurious" effort is wasted heating and extra power consumption - but can still be seen

I was just wondering what tells you that you have oscillations!!!

That's where a spectrum analyzer can help - it can find the frequency of the oscillation let alone the MAGNITUDE and spread of the effect.

In the above example of the "screen grab" I got off of Smokin Joes YT vids about these radios - it shows a slice of spectrum - starting at XX.xxxMHz on up about 300MHz Band width wide. showing even 6th-order peaks.

But one thing a Spectrum Analyzer can also do is look from DC thru the ELF (Audio) as well, helping the repair move along by showing not just the oscillation - but where in the Audio to RF Spectrum it is - can help provide guidance in to to tame the problem - not just plugging in NFB values - but knowing where and how this even takes place, many power supply bypassing and even coupling networks from the low-signal to high-power crossover feeders can have their built in self-resonance detected and the designed/fixer-upper'er can tailor the interfacing to meet the demands better - then NFB can be applied to correct that which is self-induced distortion in the stage - not just applying a bandage to staunch the bleeding from a wound self -immolated by the stages own open-loop gains and wreaking havoc with everything else tied to it in the box.

When you think about it, it's more of a horror movie...runaway oscillations - you asked a good question but you also wanted to know how you find these...

​

Put an ammeter in line along with a fuse - before Murphy shows up knocking...

​

Well, by the time you can tell, all that magic smokes' been let out and what caused it, is wafting up into the shop lights...

I look at power demands - BEFORE the moment starts to - when it dims the lights ...

 

[guitar_199]

That last message "packs a lot of punch"!!!! AND gives me something to think about when doing ANY testing on a radio. Whenever I use a Variac I generally like both voltage AND current meters.... so I can watch the behavior as I spin the AC feed up.

I know this is a little different....but still... staying "in tune" with the current demand and "catching it" when you see something unreasonable..... can possibly save a few headaches.

Thanks!!!