FM/AM radio doesn't get affected by electrical intefernece?

[flyingsliverfin]

I'm doing a school project which involves analog signals bla bla bla...
but one problem I'm having is dealing with the 60 Hz interference (USA) coming from the power outlets and electrical appliances that mess up my readings

I then started wondering why for instance car radios or portable radio players don't get interference from appliances etc, do they have sort of ingenious noise filters?

[Lucradia]

Because buildings are super thick :P Also, there's only one 60~120volt outlet in a car anyway.

[wirepuller134]

To answer your question yes they do have noise filters also cars have high suppression ignition systems to keep them from interfering with the radio. The most important thing with any analog signal is proper shielding both on the the control circuit and external devices. I've never had issues with line voltage but we have had a lot of issues in plants with soft started and inverted motors without proper shielding on the load wires. We also use EMI filters and reactors on long runs for inverted motors that we are directly controlling/running from one of our panels.

[halibaitor]

FM (frequency modulation) radio is kinda immune from electrical interference due to the type of signal it is. AM (amplitude modulation) does get quite a bit of interference, but not from the 60 cycle stuff generated by the power lines. That is much too low a frequency to directly effect the 0.5 megacycle or more radio signal.

I am really curious what kind of analog signal you are measuring that is getting messed up by the noise generated from the power lines. What kind of system are you using to measure your signals?

[flyingsliverfin]

Oh my project is to use flex (piezo) sensors to detect vibrations. I'm using an ADC connected to an arduino for the controlling, so i just made some software to handle the 60hz. The random small jumps are actually being more problematic right now.
Sample rate is ~25 microseconds for each sensor, of which I have two.

[halibaitor]

Interesting. With a sample rate that high, I wouldn't think you would be bothered by 60 cycle noise. I would have expected a sample rate closer to 1 millisecond when looking for vibrations, but that's just me...

[flyingsliverfin]

Well I'm trying to catch the very first spike measured along a pvc pipe. According to my oscilloscope i'd need a 25 us sample rate for 1 sensor (2 pts per spike) but i can't manage this with the arduino, its a little too slow. So for two sensors I'm doing 50 us. Generally works.

For the 60 hz, since one quarter of a period (the up section) takes about 4 ms (equating to 80ish samples) so it could affect my data.

Getting +-20 values from sample to sample sometimes which are a bigger problem.

[halibaitor]

For the 60 hz, since one quarter of a period (the up section) takes about 4 ms (equating to 80ish samples) so it could affect my data.

Getting +-20 values from sample to sample sometimes which are a bigger problem.

If you mean that the value returned from sample to sample is varying by +- 20 units, then the problem isn't a 60 cycle induced error. You noted yourself that 1/4 period covers 80 samples, so the problem is happening at a rate of 20,000 hz or so. (Maybe I didn't understand your response correctly.)

[wirepuller134]

The big thing sticking out to me with that high of a sample rate, is what is the quality and rate of the sensors, the capturing device and even the scope and it's leads. What type of cable are you using to connect the sensors to the capture device and how are the connections made. How are you protecting all of the equipment from environmental interference Or to put it another way did you do a base line reading to filter out normal vibrations of the area.

[flyingsliverfin]

If you mean that the value returned from sample to sample is varying by +- 20 units, then the problem isn't a 60 cycle induced error. You noted yourself that 1/4 period covers 80 samples, so the problem is happening at a rate of 20,000 hz or so. (Maybe I didn't understand your response correctly.)

Sorry I meant that the 60 hz isnt that much of a problem, and its also fairly predictable. It and the +-20 are not related.
The +-20 is random interference that I'm getting otherwise. In the picture I attached its the squiggly stuff at the beginning.

Otherwise... I'm using a coax cable about 1 meter long to connect to the sensor on the pipe (soldered leads to coax). All my wires from ADC to arduino etc are just on a breadboard.

This is just me playing around. My brother (in uni) tells me he's had to take whole classes on interference/design etc so I shouldn't worry about it yet too much

[juancarlospaco]

Its not the radio, but the signal that gets interfered