64MHz oscillator, ideas and propositions
(Source: hanssummers.com/images/stories/fast/circ.gif)
that uses a 64MHz module as a master oscillator. I would like to replace this with a homebrown oscillator.
Signal quality does not matter too much because it is converted to square wave from the ICs after all.
Searching at the web I found different solutions
1. http://www.circuitsonline.net/forum/view/48065
(look this image in the page
http://www.uploadarchief.net/files/download/butler.png
this is a FET butler oscillator.
2. Another circuit from Elektor mag
http://www.eeweb.com/blog/extreme_ci...ncy-multiplier
3. A no name one
http://www.seekic.com/circuit_diagra...scillator.html
4. Another noname one
http://www.next.gr/inside-circuits/c...mhz-l9117.html
5. An interesting multivibrator that uses a crystal!
http://www.eleccircuit.com/symmetric...lator-circuit/
All of these operate on the 3rd harmonic I think. It would be more convenient to find one that operates on the 4th harmonic, since a common 16MHz crystal can be used. I do not know if anyone of these could operate in the 4th harmonic as well.
Which way should I better go for my application? I need something fast and simple.
An update:
By having a second look, probably a good a simple circuit for my app is the elektor one
http://www.eeweb.com/blog/extreme_ci...ncy-multiplier
because the output level will drive nicely the following chips
But it says that:
"When, for example, an 8 MHz crystal is used to obtain an output frequency of 72 MHz (9 × 8 = 72), the frequency determining inductors and capacitors have to be adjusted by a factor of 10/8."
I guess that for 7.1MHz crystal the components must be scaled to 10/7.11 ?
Also what does this scaling mean, I mean how I will calculate the new values?
To verify if I do it right, consider the C2 with a 10/7.11 scaling factor. My calculation shows that it will have a value of 23.463pF, is that correct?
Scaling is most reliable with the 3rd harmonic. The harmonics at 2x or 4x won't work.
Your original (slow) frequency is the instigator for your output (fast) frequency. The input waveform transitions are what drives the 3x resonant action.
Every transition of the input waveform, must serve to reinforce your output waveform's transition. It needs to be in the same direction. This is only possible at odd harmonics.
Sorry , I have not read all your references but as I understand your problem is doing a X4 multiplication. How about a tuned circuit input with a centre tapped secondary and two diodes feeding the output, like a power supply rectifier circuit. A bit of amplification (4XXX cmos buffer?) then another doubler circuit.
Frank
That 64Mhz oscillator just gets divided straight down to 32Mhz.
Why not use a standard 32Mhz crystal oscillator module that is more or less a standard readily available part.
http://au.rs-online.com/web/p/crysta...ators/4719017/
64 MHz oscillators:
http://www.digikey.com/product-searc...=0&pageSize=25
http://www.mouser.com/Passive-Compon...=64mhz&FS=True
http://uk.rs-online.com/web/c/semico...4294957187&m=1