manmade sources of interference multiples of 18.5-23.5Hz up to 1KHz? (non random

[coppercone2]

Are there any likely high power interference devices that tend to make these frequencies on 60Hz grid?

Idea being that they would not get handled by a high Q bootstrapped notch filter network that is set to 60Hz multiples.

[T3sl4co1l]

Maybe some of treez's lamp fixtures were installed nearby?

Tim

[David Hess]

Anything which uses a thyristor for pulse width modulation will produce subharmonics of the power line frequency.

[vk6zgo]

Anything which uses a thyristor for pulse width modulation will produce subharmonics of the power line frequency.

The frequencies quoted don't look like sub harmonics of 60Hz, or 50Hz, for that matter.

[coppercone2]

Ok, I was kind of imagining messing around with some experiments on www.vlf.it and I think this merits building a adjustable filter notch network might be useful.

Otherwise you would used a fixed network I guess because it costs less then using some kind of LTC/Maxim filter blocks.


The idea is you can make notch filters for 60, 120, and so on, I think the power companies deals with up to the 7th harmonic near their substations with all the distortion from various things... but a high Q notch filter has high Q, so I made up the range of multiples between 18.5-23.5Hz because those might slip by not so high high Q notch filters, if you make them better this range increases.

You would get 74Hz-94Hz and so on in dead zones. I guess sub harmonics are important too because you don't want to get hit with some 30Hz or something. You might kinda get away with a few of the higher ones when the Q of the notches starts being shitty because of higher frequencies I think, but it still can pose a problem to where you might wanna tune the notches with a frequency controlled filter. I want to know if this complicated contraption should ever prove as even a marginally useful piece of test equipment.

I thought the notches could offer better response then any kind of preselector sweeping or whatever, especially if there is only a few offenders, I did not imagine the band to be totally chowdered.

[Ian.M]

With a 60Hz grid, cycle skipping power control of heating loads, at 1/3 power can produce 20Hz :  +ve half-cycle, skip whole cycle, -ve half cycle, skip whole cycle.

[coppercone2]

With a 60Hz grid, cycle skipping power control of heating loads, at 1/3 power can produce 20Hz :  +ve half-cycle, skip whole cycle, -ve half cycle, skip whole cycle.

I can't do a fft in my head but that might pollute 80Hz, if it hits a rectified even though its not odd?

Guess there are no safe frequency ranges lol

Maybe you can build some kind of chopper or whatever you wanna call it to only take readings near the 0 crossing when all the garbage is minimized, so long your not interested in any garbage in particular.

[coppercone2]

oh yea by non-random I mean signals that will be repetitive at least for minutes, I am not trying to make a filter that rejects lighting noise, atmospheric related etc

I figure you would probobly want to monitor the stopbands with a antinetwork of passband filters that have some AGC to keep an eye on random transients