user6972

If my answer gets deleted for reasons I'm not clear about who/where do I ask about it? (physics.stackexchange.com/questions/75809/…) Not that it really is important to me, but I thought it was odd that it was removed.

no problem. later.

Ok, maybe tomorrow. I don't have the time now. Any other questions?

RF engineers commonly keep a kit of beads around along with cable clamps for a quick check against noise sources fair-rite.com/cgibin/…

In general if you can get 200-500 ohms of load with the ferrite the noise will drop significantly.

Good question. The coax ground and the ground plane for the antenna are the reference plane. the choke just keeps your PCB ground from becoming contaminated with that band of noise, there will be an impedance discontinuity (and slight loss of rx power) due to the ground choke however once the received signal wave passes to the ground reference of the PCB the loss of power from the choke will be more than compensated by the cleaner input signal.

I would start with looking at the effect of putting 1,2, 3, 4 then 5 on the receiver end and you'll get an idea of if the problem is cable related. If it is, find out the minimum number of ferrites you need, then try spreading them out, one on each end. Ideally you'll want to break up every section of cable that is ~ 1/4 wave of your highest frequency, but usually people just clip on a bunch of ferrites on each end and call it done.

The 50 ohm load is good but make sure it is a shielded load. How long is the cable run?

By the way, I just want to say that I'm only asking 50 questions to try to understand the problem. I am in no way insinuating you don't know what you are doing.

Here's a list of clamp on ferrites with material 31. You can probably find them for cheap on ebay, just make sure the material is the right type. fair-rite.com/cgibin/…

You might find the cable isn't a big contributor or you might find that you need 3 or 4 ferrites to knock the signals down.

You can buy clamp on ferrites just to experiment. If you don't have a way to modify your PCB. Also try connecting your ADC directly to your antenna or a much shorter piece of coax and compare signal levels to get an idea of the cable's contribution.

Looks like material blend 31 would work fair-rite.com/newfair/materials31.htm

You can put ferrite chokes on your ground connection between the PCB and the coax by breaking up the grounds. Did you look at the muratta link at all?

you have to put the right ferrite and the right amount of ferrites. Run your FFT, record the levels, clamp on a Ferrite for the 7-30Mhz band, look at the differences.

Actually it can be a huge problem for transmitters, the feed back can burn up the PA, distort the signal or even pull the oscillator off frequency.

your question relates to how to make your coax shield a "bad" antenna.

your common-mode choke helps more for load-mismatch reflections.

no, along the length of the coax

it's identical to what you've been asking about. the shield is picking up signals like an antenna.

the power can leak out, but most of it comes by being picked back up once the wave is transmitted by the antenna.

this goes for large RX signals too because you can buy 0402 ferrites thinking they will work but if the current is too strong they won't.

no it's not irrelevant. if the ferrite is saturated it won't work.

there are more effective ways to design in noise rejection on the pcb, which I included in my answer too.

but to use them properly you have to pick the right material for the answer. on transmitters you have to make sure it is big enough not to be saturated by the current flow. EM waves are reversible, so anything that works for tx em waves will also work for rx.

that's why you see so many of them on computer cords and equipment.

but you need to use a material that absorbs in your frequency band. ferrites are very effective.

I've told you that in my answer, and so have others...ferrites.

yes, but an antenna null is not zero either. how high to they rise?

knowing that you can do an FFT is helpful.

If you point at those signals how high do they rise?

The digital filter is critical to the radios performance--fyi

Do you have a noise problem right now or are you just trying to set something up?

ok, so you're applying the narrower IF filter in software. Part of the problem you're having with noise could be related to the adjacent channel filtering. Do you know what the filter specs are?

what receiver are you using so I can see its specs?

that doesn't matter your receiver can only receive what falls into its channel bandwidth.

because I need to know how the noise effects your receiver. I need to know the filtering scheme and modulation scheme. There is no 23Mhz wide broadcast signal in that band only narrow band audio.

ok, so the receiver filter is going to be around 10Khz, for AM. is this correct? does your radio have a signal strength meter?

are you also transmitting this signal? what is it you're receiving?

power level? modulation?

Part of this is also knowing the bandwidth of the desired signal...

This means I need to know something about the signal...modulation, filters in the receiver, dynamic range if possible.

Since you don't have a spectrum analyzer I need to know how the receiver works in order to kludge a test setup for you.

@PhilFrost One channel and it's a 23Mhz bandwidth signal that you are receiving? I really need to know how your system/bandwidth/channel scheme works to help. AM or FM?

@PhilFrost I am trying to help you but I need a lot more info. Is this a narrow bandwidth AM audio receiver or not? Do you have a spectrum analyzer with a preamp and filter for 7M-30MHz?

@PhilFrost this sounds like an AM ssb HAM band with narrow (<10Khz) bandwidth. Do you have a spectrum analyzer and a preamp for this band?

@PhilFrost ferrites absorb the energy not reflect it. What is your frequency and bandwidth of the signal you are transmitting? What is the characteristics of the noise? (frequency, bandwidth, how does it cause problems?)

@Andyaka I've had time to edit my post to try to answer your question.

Also you're missing I4, which is the noise on the center conductor. There are components of I4 and I3 which are in phase/frequency (common-mode) and those that are not (non-common mode). How you filter them depends on what harms your design.

@ThePhoton That wikipage isn't very complete. By reducing the capacitance there is a dramatic improvement in the quality of the received signal -- less group delay, less reflections, wider bandwidth, etc. It also dramatically reduces the amount of noise leaked to the center conductor. Measuring small accurate currents is just a side benefit of less noise and clean signals.