Hi
I have a Colpitts osc that is giving me a headache in radiated EMI
If I could pull the frequency at 100Hz so the energy is spread out, it could solve my problem
Sort of like this:
It looks like they use the inherent properties of the transistor to make a chaotic signal
I am using a standard BC847, so cannot do this
Anybody have an idea how to pull the frequency, cheap?
Sort of like an varicap across the tank caps, but varicaps are not cheap?
Cheers
Klaus
Crystal or LC?
You might be able to FM a bit by changing V_CE.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
What's your frequency? How far above spec limits is your EMI? In what measurement bandwidth? In other words, how much do you need to sweep?
As Phil says, frequency will depend on supply voltage.
And all diodes are varicaps. Most ceramic caps are varicaps too.
Either way, you'd need a low frequency oscillator or equivalent to sweep the oscillator frequency.
Is there a uP port pin available? You can have fun lowpass filtering a pseudoramdom bit stream.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Especially if the uP is clocked by the oscillator which is being jittered by the pseudorandom bit stream...
John
It's interesting to consider making truly random bit streams from a small uP. It needs an external source of entropy.
One idea is to pull down a tri-state pin and let it charge back up into the same, or another input pin, logic or an ADC input. That generates a sorta random time delay or number. Stir that into the guts of a pseudorandom sequence generator or several.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
If you can tolerate AM and the colpitts is high enough frequency you can modulate the oscillator transistor capacitances by injecting some wobble into it's base. Or instead of a costly varicap across the tank use a second vanilla BC847 across the tank either varicapping C-B or B-E junctions or as an old school reactance modulator (should be findable online).
piglet
He hasn't stated if he's using a xtal or just LC. If it's the former, there's much less latitude to pull the frequency.
-- "The BEST Deal is NO DEAL"
What about taking the audio output of a radio tuned off-station as an input to a processor for random numbers?
it's LC
Yes. I also tried loading the center cap point with variable resistance, and I can pull the frequency by doing that
Cheers
Klaus
The frequency is 20MHz. The test receiver for radiated emissions has 120kHz bandwidth, so to move it out of band, I need about 1% deviation, just to get 6dB reduction
Right now I am about 15 dB above the limits
Yes. I have a microcontroller on board, so I can use that to modulate the frequency. Although I would like a self running device, I have 2 IO pins that could be used for that purpose
Cheers
Klaus
I like the idea to inject into the base. I guess it should be an ac coupled cap. I have mains input voltage, so I could even feed mains ripple from the DC link capacitor into this point.
Cheers
Klaus
I do not have a radio receiver in this product
No shielding. It is leaking via a capacitance across a barrier. I have no clear way to reduce that leaking. I need to try other tricks
Cheers
Klaus
You don't need anything fancy for spread-spectrum. I've used a sloppy triangle from a schmitt-gate-RC oscillator. If you have a uP output pin, lowpass that and square wave drive to make a triangle. Fancy pseudo-random the pin to make a gaussion spectrum. But a triangle is fine, to make a rectangular spectrum.
Any diode is a varicap. It could be as simple as a schmitt gate, one resistor, one cap, one diode. Possibly simpler, depending on your oscillator circuit.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
The name is important. Octpart shows 174 hits for varicap:
and 1,000 hits for varactor:
The capactance range is wider and probably better controlled than with a random diode.
1N4001 shows typical values:
Then don't use a random diode! I think Klaus is super price sensitive, so some under-one-cent diode might be appropriate.
We don't know his frequency or the circuit, so we can't recommend anything.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Good luck finding a diode under a cent with controlled capacitance characteristice that can be used in his circuit.
Perhaps shielding might work better, or adjusting the operating parameters of the oscillator to reduce EMI.
For examples of properly adjusted Colpitts, see
Follow the instructions in the Readme.txt file.
Yes, as always :-)
I wrote that in another post, about 20MHz
Cheers
Klaus
Yes, I am working on this. I have a couple of pins free on the micro, so could use these for the function.
About the receiver comment, it's just that another post suggested a radio receiver, and as I wrote I do not have that luxury ;-)
Cheers
Klaus
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.