Up and down converting oscillator frequencies. Also LC question.

[Chris Wilson]

If you build say a 500 MHz oscillator and down convert its output to say 50 MHz by means of a frequency divider there never seems to be a filter in the divider circuit, yet if you up convert from say 50 MHz to 500 MHz by frequency multiplication, circuits usually feature filters. Why is this please? Why is the current trend to down converting to a lower IF in receivers from a high frequency LO, whereas historically they seemed to up convert in HF / VHF receivers? Is it the cheaper availability of high frequency, stable and clean oscillators?

Finally, if I build a basic LC circuit with a toroid wound coil and a ceramic cap, and want to then shift its resonant frequency, I can do it by either capacitance or inductance changes, but is there an optimum and harmonious balance in their values? How do you judge whether to alter the cap, or rewind the toroid, or if you need to change both values? All three options may result in the desired resonance, but I suspect that there's a preferred way to get the "balance" of the component values right?

Thanks.

[ejeffrey]

I'm not sure what you mean.  To convert a 50 MHz oscillator to 500 MHz or vice versa you generally don't use up/down conversion (mixing).  You use a PLL.  For a square wave oscillator you can do the division with a straight up divide by 10 counter, but if you want to generate a sine wave you need a PLL.  In principle you could go from 50 MHz to 500 MHz by a series of frequency doublers and mixers.  This is sometimes used in extreme circumstances like THz sources.

Up and down converters always require filters, but if the sum and difference frequencies are widely separated it could be very simple: as little an an AC coupling cap, or just relying on the bandwidth limit of the next stage to reject the sum frequency.

[vk6zgo]

If you build say a 500 MHz oscillator and down convert its output to say 50 MHz by means of a frequency divider there never seems to be a filter in the divider circuit, yet if you up convert from say 50 MHz to 500 MHz by frequency multiplication, circuits usually feature filters. Why is this please? Why is the current trend to down converting to a lower IF in receivers from a high frequency LO, whereas historically they seemed to up convert in HF / VHF receivers? Is it the cheaper availability of high frequency, stable and clean oscillators?

Finally, if I build a basic LC circuit with a toroid wound coil and a ceramic cap, and want to then shift its resonant frequency, I can do it by either capacitance or inductance changes, but is there an optimum and harmonious balance in their values? How do you judge whether to alter the cap, or rewind the toroid, or if you need to change both values? All three options may result in the desired resonance, but I suspect that there's a preferred way to get the "balance" of the component values right?

Thanks.

Direct Frequency multiplication  by a factor of 10 is a bit messy,because you would need to have two multiplier stages,
one at 5X,& one at 2X.
Odd harmonic multipliers above 3X are very rarely used,as they are not very efficient,so you would be more likely to  multiply by 8,from a fundamental of 62.5 MHz,or by 12, from 41.667MHz to obtain 500Mhz.
Generally,in practice you would use a greater multiplication factor to allow more economical choice of crystals.

The reason for a filter is that multipliers don't only produce the wanted harmonic,so the unwanted ones need to be suppressed.

Direct multiplication is seldom used commercially in modern designs even at HF,since the development of PLLs,which are not only use less components,but allow greater frequency agility.

Multiplication to UHF was quite rare,even in the old days,except for the production of local oscillator signals for use in an converter.
These devices are commonly used to convert modulated signals from a lower frequency to a higher one or vice versa.

You can't normally multiply  a modulated signal without destroying the modulation. * 
There is a special technique which was used in the early days of microwave SSB work,which allows such multiplication, but it  is hardly ever used these days.

On the receiver question:
The current trend is a reversion to an earlier idea,where the IF in HF or VHF receivers was  lower than the received frequency,except,for example ,the three low bands,160m,80m,& 40m in a  HF Ham radio with a 9MHz IF.
The high IF designs were very popular in the late '70s though to the  late '90s,but seem to have fallen out of favour in more recent years.
Why?
Your guess is as good as mine!

Re the LC circuit:
There is a wide range over which you can select components,as it is pretty much "select on test",but capacitors are less lossy than inductors which have inherently larger series resistance
Also larger capacitors tend to "swamp" any stray capacitance,so a "high C" circuit would be advantageous in many cases.

There is a limiting factor to the C/L ratio in a tuned circuit,but I can't quite recall  it off hand.

 *  see next posting

[vk6zgo]

But wait,there's more!
*I told a bit of a fib about frequency multiplying modulated signals.
What I said applies to the various types of amplitude modulated signals,such as full carrier AM,DSB,SSB,etc.
Angle modulated signals such as Frequency Modulation & Phase modulation can be multiplied-----In fact,that was how it was done for many years.
The fundamental crystal was modulated to a very small deviation,& the deviation automatically increased with frequency multiplication until at the final frequency,it was the desired deviation.