multiconductor cable problem (stray capacitance)

Another option would be to install the amplifier directly adjacent to the sensor. For low level signals it is a very common approach.

To start towards answering your question:

Why 100 MHz? Where does the 100 MHz come from? Is that a power supply (asking because the 60Hz 120V seems to be...).

Although 100 MHz would probably couple much more from wire to wire than 60 Hz, it might be cleaner and (being higher frequency) easier to filter out. It depends.

the 100Mhz signal is a digital signal originating from a computer. I figured it would be a good approximation for coupling to say that its a sinusoid.

I took just 1 class on instrumentation 2 years ago, and can't remember much... I only remember reading a couple of paragraphs on stray capacitance... (i'm a biochemical engineer) The design is too complicated to explain here... But the only option I got is to use the cable to carry two signals...One is fixed, and the other, I can either use the power line or the digital signal.... It would be great if you could explain why...

So you have a cable with 120Vac 60Hz on one set of wires, and some sort of 100MHz digital signal from a PC on another set of wires, and you need to move a small amplitude analog signal using some of those same wires?

If this isn't it, then explain about the cable again.

I have a varying (slowly) small amplitude signal (still didnt figure out its properties) on one set of wires. For the second set, I have 2 options...as mentioned above (a power signal or a digital signal).... Which one would be a better option...? the power one or the digital one?

I see. Sorry I'm slow on the uptake. My excuse is that I still can't really envision how one would have a system architecture choice between AC power or a 100 MHz digital signal (but I do believe you).

Your answer depends on many things. But I'd instinctively try the higher frequency digital signal because it would seem to be much easier to filter out 100 MHz (plus and minus sidebands and who-knows-how-much noise), than to try to filter out all the low frequency rubbish found on the typical AC power line.

You'd probably want to use a differential input at the amplifier end so that much of the noise (being common mode) would be ignored.

On the face of it, it would seem to be a fairly simple exercise to remove the high frequency from the slowly varying. It's a spectrum separation approach.

Stand by for other opinions - they're all free.

Never try to carry a low amplitude signal over a cable.

(Never)^2 try to carry a low amplitude signal over a shared cable.

Connect an amplifier directly to your sensor. Use the cable to power the amplifier, and an ADC, and a small RF transmitter to send the amplified and sampled signal wherever you need it to go.



Mike Halloran
Pembroke Pines, FL, USA

Thanks for the help guys..
VE1BLL: This is an idea that I came up with while I was doing my senior design... So I've just been putting together some stuff for that...

And Mike: "((Never)^2 try to carry a low amplitude signal over a shared cable)" Care to explain why? I'm assuming its due to the stray capacitance, but are there other factors?

Not just 'stray capacitance', but also inductive coupling. Unless the wire pairs are individually shielded, you have to assume that there will be plenty of coupling (depending on the overall length).

But, it is possible for various signal to share even just ONE PAIR of conductors by taking advantage of the frequency spectrum.

For example, satellite TV cables from the dish to the receiver have chaning 13 or 18 Vdc power, switched 22kHz audio, sometimes a digital code riding on the 22kHz audio, and L-band RF all running on the same cable.

Many microphones provide DC bias on the same wires.

Stray everything, and connector resistance.




Mike Halloran
Pembroke Pines, FL, USA