unwanted oscillator synchronisation

Question

unwanted oscillator synchronisation

Wedwin (Electrical)

(OP)

5 Mar 04 12:45

Hi all!

I want to create a few oscillators using Schmitt inverters and R and C to set frequency. I want them to run almost but not exactly at the same frequency. This because I want to conduct som interference experiments.

When I get close, bam, they get synchronised. The same thing have been rebuilt twice and several values have been tested

Has anyone else come across this phenomenon? Do anyone have any suggestions?

I am around or below 100 Hz, but am not afraid to try other ranges. It just "feel" alright to handle frequencies that fit into my mind.

Best regards, Wed

cbarn24050 (Industrial) · Answer 1 · 2004-03-05 14:00:07

you are using the wrong sort of oscillator, try a LC type.

nbucska (Electrical) · Answer 2 · 2004-03-05 14:03:13

The cause is one osc. signal getting coupled into the
other. When this is close to the switching threshold,
its noise sensitivity is diminished so it gets triggered
prematurelly.

Solution:

Separate shielding buffering and decoupling, Run the osc.-s
at relative prime-times multiples ( e.g. 11*F1 and 13*F2 ) and divide them down.

<nbucska@pcperipherals.com>

GOTWW (Industrial) · Answer 3 · 2004-03-05 14:47:32

I have done this on purpose many times to power combine oscillators. Another good one is even if the (multiple) oscillators are far enough away in frequency, they will remain at their frequency, but will become syncronized in (phase) called mode-locked condition, Producing a repeatable interfernce pattern.

Warpspeed (Automotive) · Answer 4 · 2004-03-07 00:45:40

Another way you can do it is to use two oscillators on quite different operating frequencies, and then digitally divide each frequency down by an appropriate division ratio.

There should then be no interaction or frequency pulling effects, because the oscillators are on totally different frequencies.

cbarn24050 (Industrial) · Answer 5 · 2004-03-07 02:11:01

Hi warpspeed, relaxation oscillators will quite happily lock onto any hearby harmonic, back in the old tube days a multivibrator was used to divide by any number up to about 7.

Wedwin (Electrical) · Answer 6 · 2004-03-07 06:00:33

Thanks a lot all of You for the information.

An LC solution would probably bring up the frequency a notch or two, but foremost in size.

The whole idea was a pretentionless experiment. That is why a common RC inverter osc would fit so well.

Since it doesn't work, I am happy to know why. And I will try to apply this newfound knowledge.

Best regards, Wed

Warpspeed (Automotive) · Answer 7 · 2004-03-07 16:30:50

cbarn24050

hehe, yes, but if one oscillator is divided by seven, and the other by eleven, there will be a continuously varying phase relationship between both oscillators.

The worst that can happen is a bit of pulling and jitter as they momentarily pass through instants of simultaneous switching. But the dividers will eliminate most of that from the outputs.

I can remember tweaking tube circuits with relaxation oscillators used for frequency division in commercial equipment many years ago.

nbucska (Electrical) · Answer 8 · 2004-03-08 16:24:18

Cbaarn: Yes, you R right, simpler circuit == more interaction.
Warp: Thanks for the support. Larger #s == less coupling.

<nbucska@pcperipherals.com>

Come Join Us!

Posting Guidelines

unwanted oscillator synchronisation