AM modulation index

How many bits is your ADC? That will determine how much of the carrier has to be removed.

My suggestion would be to pull a small amount of signal from the carrier and phase-lock an oscillator to it. The oscillator should have very low levels of amplitude and phase noise. The phase of the oscillator should be adjusted until it is 180 deg. out of phase with the carrier. If it is amplified until its voltage approaches that of the carrier, summing the two carriers will reduce the carrier to modulation ratio.

Rather than amplify the oscillator voltage to 250 V, it might be easier to go through multiple stages of summing.

John Fields, John Larkin and jgreimer (and others)

First of all, thanks for your moral support.

The circuit is for reading out an FDX transponder. The ADC is 16 bits to keep it within my budget, not only for the ADC itself but also the chain of digital hardware after it.

I thought of the PLL (and even a DLL) suggestion already, but it works out that it is very difficult to do. Certainly for 250 Volts, I have to attenuate that first, from there every little phase shift and pinch of noise ruins all the nice work, it simple isn't good enough.

Maybe the notch will work, I'll try that one.

B.T.W.: There's one thing I'm sure of, I know the exact frequency of the signal. The transponder is (faintly) modulating the carrier I'm transmitting by sucking up, now and then, a teeny little bit of the EM wave I generate myself. That gave me the idea of the DLL I mentioned. Does that ring any bells at your end?

Thanks so far, Rutger

You're generating the original 100 KHz carrier, and the transponder is essentially selectively reflecting it? If that's the case, bandpass the received signal (to improve gross s/n) and then sum it with a tweaked copy of your prime signal source. Tweak the phase and amplitude of the summed local signal to reduce most of the carrier in the received thing; some simple feedback loops should work, maybe even trimpots.

But the dynamic range and s/n situation still sounds nasty.

John

of

out

noise

transmitting

The transmitter coil is in essence just a loop antenna burried in the floor, one end is grounded, the other end connected to the transmitter output. The transponder is lying in this loop, detects the carrier and then, after picking up some energy for it's electronics, short circuits it's own (very small) coil in the rithm of it's unique (digital) ID, thus slightly modulating the EM field. I'm measuring over the same transmitter coil, so there is no separate receiving antenna. That's how the incoming signal originates. That clarifies how I know the frequency exactly.

As stated, the modulation index is very small. If anything metal in the neighbourhood moves, you get phase shifting and interfering AM (even a little FM) modulation, although (hopefully) not of the same frequency as the transponder's modulation. So there you are, it looks ghastly. That's why I want to filter and demodulate the signal digitaly, but then again, before going digital I have to boost up the transponder's AM index.

"Rutger Stoots"

" a software engineer and not really comfortable with electronics " ????

....... Phil

Hell, you're describing a metal detector. There are pretty much three kinds:

Oscillator, where the sensed object pulls the frequency of an oscillator which uses the sense loop as its tank. Usually detected by listening to the heterodyne.

Bridge: the loop is driven by an oscillator/amplifier and is part of an AC bridge. The bridge is tuned for null, and a detected object upsets the bridge balance.

All of these can be made sensitive in the ballpark of parts-per-million. The oscillator version is out for you, but a bridge arrangement might work. A slow servo could null the bridge, and a phase-sensitive detector could recover the faster data.

Still, there will be noise, including amplitude and phase noise in your oscillator/amplifier itself, and the signal levels you're considering are tiny.

Could you do the RFID thing, have the transponder passively double the excitation frequency? Or just use the stored energy to respond on a totally different frequency? That would change everything.

John

Too bad you're all insults and no ideas. But that's in character: ideas are fun, and you never let yourself have fun.

John

John, Ever hear the term "shunning"?

If EVERYONE would simply shun Phil, the problem will vanish.

But as long as you persist with the joust the rest of us have to put up with the crap.

...Jim Thompson

Don't get snippy with me, young Jim.

I'm only increasing Phil's injected crap level incrementally. Ignoring him will probably not help a lot, since he comes to newsgroups specifically to distribute insults.

I wonder why.

John

You've never heard of "Usenet Filters" ??

The only time I see Phil's insults is when you reply to them.

I guess I just need to kill "AM modulation index", rather than reading it to see if I can help ??

...Jim Thompson

Well, if the transponder stores up energy for its response, then just pulse the transmitter long enough to charge it up, then listen for its response at whatever frequency - it could even be the same frequency.

That's how they do radar and ultrasonic, you know. :-)

Good Luck! Rich

It's called "trolling", and he seems to be doing a pretty effective job at it.

Thanks, Rich

Nah, he's describing standard RFID technology. I wonder whether there aren't any canned solutions around.

Meindert

Don't RFIDs generally respond on a different frequency than the excitation? Often a harmonic.

John

Hey, bud, whatever happened to that SSB 1 GHz mixer you were going to post on ABSE?

Jim

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...